Discuss the types of errors that are classified in a Root Cause Analysis.

Discuss the types of errors that are classified in a Root Cause Analysis. Include a discussion of the accident models presented in Chapter 6 of the text and how these tools could be used to analyze the errors you identified.

Embed course material concepts, principles, and theories (which require supporting citations) in your initial response along with at least one scholarly, peer-reviewed journal article. Keep in mind that these scholarly references can be found in the Saudi Digital Library by conducting an advanced search specific to scholarly references. Use Saudi Electronic University academic writing standards and APA style guidelines.

Requirements: ……………………

123A Case-Based ApproachRahul K. ShahSandip A. GodambeEditorsPatient Safety and Quality Improvement in Healthcare

Patient Safety and Quality Improvement in Healthcare

Rahul K. Shah • Sandip A. GodambeEditorsPatient Safety and Quality Improvement in HealthcareA Case-Based Approach

ISBN 978-3-030-55828-4 ISBN 978-3-030-55829-1 (eBook)https://doi.org/10.1007/978-3-030-55829-1© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed.The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use.The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.This Springer imprint is published by the registered company Springer Nature Switzerland AGThe registered company address is: Gewerbestrasse 11, 6330 Cham, SwitzerlandEditorsRahul K. ShahChildren’s National Health SystemWashington, DCUSASandip A. GodambeChildren’s Hospital of The King’s DaughtersNorfolk, VAUSA

vEmily died in our hospital. She was 3 years old. She passed away following a preventable medical error. As recently as 20 years ago, an event such as this might only show up when a grieved family brings suit against the hospital and providers. Yet today, the national dialogue and focus on patient safety and transparent outcomes has dramatically changed. In most hospitals, not only would Emily’s passing be analyzed in meticulous detail, but the results would be promulgated within and across the hospital to ensure that providers and the hospital system minimize any chance of recurrence. Further, with resilience engineering and the growing concept of Safety II, hospital systems and indi-viduals may even learn to anticipate the circumstances that predispose to pre-ventable errors [1–3] and prevent them before they occur.A plethora of texts exist that are filled with theory and concepts intending to teach about making sure “Emily” never happens again—in any of our hos-pitals. In their text, Shah and Godambe have taken the conversation and teaching about quality and safety to a more practical level. They have not only challenged the talented group of chapter authors to discuss esoteric safety and quality theory, but also to bring these concepts to life through case-based scenarios. This approach brings important safety principles into stark reality as real clinical world events showcase practical approaches to imple-ment change and achieve results. Chapters such as Behavioral Economics by Jack Stevens, Workplace Safety by Joel Bundy, and Human Factors Engineering by Jon Gleason exemplify the innovation and creativity their text displays. Those chapters represent some of the most cutting edge and chal-lenging aspects of quality and safety.I applaud Drs. Shah and Godambe for compiling a different kind of quality and safety text. One well worth the read for both students and experts. There is something for everyone in this well-done epistle.ForewordColumbus, OH, USA Richard J. Brilli, MD, FAAP, MCCM References 1. Woodward S. Moving towards a safety II approach. J Pat Saf Risk Mgmt. 2019;24(3):96–9.

vi 2. Merandi J, Vannatta K, Davis JT, et al. Safety II behavior in a pediatric intensive care unit. Pediatrics. 2018;141(6). Pii:e20180018. 3. Hollnagel E. Safety II in Practice: developing the resilience potentials. London: Routledge, Taylor & Francis Group; 2018.Foreword

viiDo we really need another book about hospital safety and quality? There are journals, webinars, and myriad national conferences that help drive the field forward. The socio-political-legal environment in the United States has never been more focused on ensuring that American healthcare protects patients and drives quality. There are numerous safety and quality assessments, task-forces, and committees coupled with insurers, industry, and innovators work-ing towards the goal to create the best healthcare delivery system. So, do we really need another book about hospital safety and quality?The passionate authors of this text provide their insights as to where the field of improvement and safety science is with regard to the views and aspi-rations of the aforementioned healthcare advocates and customers. The authors are the top safety and quality leaders. We all have and continue to lead and participate in all of the aforementioned programmatic approaches towards hospital safety and quality. However, we still feel the void. We are inundated by theoretical frameworks, “what-ifs,” and extrapolations from one industry to another, all trying to help us drive safety and quality to new pla-teaus in our organizations. However, we still feel a void. The feeling can be summed up as such: “what about us?” A gap in the programmatic approach is that the materials, conferences, and teachings oftentimes fall short of provid-ing the audience with tangible, concrete examples, with direct linkages from a structure to measured processes to discrete outcomes.Additionally, our responsibility to train our teams and future leaders in improvement and safety science cannot be forgotten – “if the student has not learned, the teacher has not taught,” a phrase used often by our Toyota sensei (John Heer, Manager, Toyota Production System Support Center (TSSC) – Australia, personal communication). W. Edwards Deming eloquently said, “there is no substitute for knowledge” [1]. The lessons from healthcare are applicable to other work sectors and vice versa – some of our expert authors, not surprisingly, come from other industries.This textbook uses a case-based approach to share knowledge and tech-niques on how to operationalize much of the theoretical underpinnings of hospital quality and safety. We were fortunate to have the leaders in quality and safety embrace this concept as it resonated with their sentiments as well. Furthermore, they all stepped up to contribute to the 22 chapters in this edi-tion. We are confident that a case-based approach with vignettes through the chapters will help solidify the theoretical underpinnings and drive home the learnings. At the end of each chapter, there are comments by the editors which Preface

viiihighlight what we believe are important concepts or connections between the various chapters in the book.As we strive to reach zero harm to our patients and staff, we must embrace different ways of thinking. This textbook presents a novel approach towards hospital safety and quality with the goal to help us reach zero harm in our organizations. Reference 1. Deming WE. New economics for industry, government and education. 2nd ed. Cambridge: MIT Press; 2000.Washington, DC, USA Rahul K. ShahNorfolk, VA, USA Sandip A. GodambePreface

ixThis book is the result of the hard work of many dedicated authors with the support of their respective families. It has been a pleasure to work with them and make this dream concept of a case-based learning textbook a reality. We would especially like to thank the countless patients and families, trainees, and colleagues, past and present, whose thoughtful questions and expecta-tions of excellence have made us better improvement and safety scientists and clinicians. Finally, many thanks to our loving families, especially our wives, Banu and Libby, and children, Nisreen, Amir, Maya, Samir, and Riya, who have made sacrifices, yet have been there to support, entertain, and inspire us!We would like to remind everyone of our goal – to strive for and attain the goal of zero harm!Acknowledgement

xi 1 Introduction: A Case-Based Approach to Quality Improvement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1Sandip A. Godambe and Rahul K. Shah 2 Organizational Safety Culture: The Foundation for Safety and Quality Improvement . . . . . . . . . . . . . . . . . . . . . . . . . 15Michael F. Gutzeit, Holly O’Brien, and Jackie E. Valentine 3 Creation of Quality Management Systems: Frameworks for Performance Excellence . . . . . . . . . . . . . . . . . . . 37Adam M. Campbell, Donald E. Lighter, and Brigitta U. Mueller 4 Reliability, Resilience, and Developing a Problem-Solving Culture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55David P. Johnson and Heather S. McLean 5 Building an Engaging Toyota Production System Culture to Drive Winning Performance for Our Patients, Caregivers, Hospitals, and Communities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69Jamie P. Bonini, Sandip A. Godambe, Christopher D. Mangum, John Heer, Susan Black, Denise Ranada, Annette Berbano, and Katherine Stringer 6 What to Do When an Event Happens: Building Trust in Every Step . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117Michaeleen Green and Lee E. Budin 7 Communication with Disclosure and Its Importance in Safety . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143Kristin Cummins, Katherine A. Feley, Michele Saysana, and Brian Wagers 8 Using Data to Drive Change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155Lisa L. Schroeder 9 Quality Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173Michael T. Bigham, Michael W. Bird, and Jodi L. Simon 10 Designing Improvement Teams for Success . . . . . . . . . . . . . . . . . 193Nicole M. Leone and Anupama SubramonyContents

xii 11 Handoffs: Reducing Harm Through High Reliability and Inter-Professional Communication . . . . . . . . . . . . . . . . . . . . . . . . 207Kheyandra D. Lewis, Stacy McConkey, and Shilpa J. Patel 12 Safety II: A Novel Approach to Reducing Harm . . . . . . . . . . . . . 219Thomas Bartman, Jenna Merandi, Tensing Maa, Tara C. Cosgrove, and Richard J. Brilli 13 Bundles and Checklists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231Gary Frank, Rustin B. Morse, Proshad Efune, Nikhil K. Chanani, Cindy Darnell Bowens, and Joshua Wolovits 14 Pathways and Guidelines: An Approach to Operationalizing Patient Safety and Quality Improvement . . . . 245Andrew R. Buchert and Gabriella A. Butler 15 Accountable Justifications and Peer Comparisons as Behavioral Economic Nudges to Improve Clinical Practice . . . . 255Jack Stevens 16 Diagnostic Errors and Their Associated Cognitive Biases . . . . . 265Jennifer E. Melvin, Michael F. Perry, and Richard E. McClead Jr. 17 An Improvement Operating System: A Case for a Digital Infrastructure for Continuous Improvement . . . . . . . . . . . . . . . . 281Daniel Baily and Kapil Raj Nair 18 Patient Flow in Healthcare: A Key to Quality . . . . . . . . . . . . . . . 293Karen Murrell 19 It Takes Teamwork: Consideration of Difficult Hospital-Acquired Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309J. Wesley Diddle, Christine M. Riley, and Darren Klugman 20 Human Factors in Healthcare . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319Laurie Wolf, Sarah Henrickson Parker, and Jonathan L. Gleason 21 Workforce Safety . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335Joel T. Bundy and Mary M. Morin 22 Changing the Improvement Paradigm for Our Kids . . . . . . . . . 353Daniel B. Wolfson, Jeffrey Scott Warshaw, and Julianne C. Coleman Afterword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375 Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377Contents

xiiiDaniel Baily, MSHS Beterra Health, Newnan, GA, USAThomas Bartman, MD, PhD Nationwide Children’s Hospital, Division of Neonatology, Columbus, OH, USAAnnette Berbano, MSN, RN, CCRN Kaizen Promotion Office, Harbor- UCLA Medical Center, Torrance, CA, USAMichael T. Bigham, MD, FAAP, FCCM Akron Children’s Hospital, Department of Quality Services, Akron, OH, USAMichael W. Bird, MD, MPH Akron Children’s Hospital, Department of Quality Services, Akron, OH, USASusan Black, MSN, NP Kaizen Promotion Office, Harbor-UCLA Medical Center, Torrance, CA, USAJamie P. Bonini, MS Toyota Production System Support Center (TSSC), Toyota Motor Corporation, Plano, TX, USACindy Darnell Bowens, MD, MSCS University of Texas Southwestern, Children’s Health Dallas, Department of Pediatric Critical Care, Dallas, TX, USARichard J. Brilli, MD, FAAP, MCCM Nationwide Children’s Hospital, Division of Pediatric Critical Care Medicine, Columbus, OH, USAAndrew R. Buchert, MD Clinical Resource Management, UPMC Children’s Hospital of Pittsburgh, Pittsburgh, PA, USAGME Quality and Safety, Wolff Center at UPMC, Pittsburgh, PA, USAPediatrics, University of Pittsburgh School of Medicine, Pittsburgh, PA, USALee E. Budin, MD Driscoll Health System, Corpus Christi, TX, USAJoel T. Bundy, MD Sentara Healthcare, Virginia Beach, VA, USAGabriella A. Butler, MSN, RN Healthcare Analytics and Strategy, UPMC Children’s Hospital of Pittsburgh, Pittsburgh, PA, USAContributors

xivAdam M. Campbell, PhD Children’s Hospital of The King’s Daughters, Department of Quality and Safety, Norfolk, VA, USANikhil K. Chanani, MD Quality and Outcomes, Cardiac Service Line, Children’s Healthcare of Atlanta, Emory University School of Medicine, Department of Pediatrics, Atlanta, GA, USAJulianne C. Coleman, MA, EdD The CORE Districts, Sacramento, CA, USATara C. Cosgrove, MD Nationwide Children’s Hospital, The Heart Center, Columbus, OH, USAKristin Cummins, DNP, RN, NE-BC Children’s Health, Dallas, TX, USAJ. Wesley Diddle, MD Pediatric Cardiac Intensivist, Children’s National Hospital, Cardiac Critical Care Medicine, Washington, DC, USAProshad Efune, MD Children’s Health Dallas, University of Texas Southwestern, Department of Anesthesia and Pain Management, Dallas, TX, USAKatherine A. Feley, DNP, RN Indiana State Nurses Association, Indianapolis, IN, USAGary Frank, MD, MSEM Children’s Healthcare of Atlanta, Atlanta, GA, USAJonathan L. Gleason, MD Jefferson Health, Philadelphia, PA, USASandip A. Godambe, MD, PhD, MBA Children’s Hospital of The King’s Daughters, Norfolk, VA, USAMichaeleen Green, BA, Mathematics Ann & Robert H Lurie Children’s Hospital of Chicago, Chicago, IL, USAMichael F. Gutzeit, MD Children’s Hospital of Wisconsin, Milwaukee, WI, USAJohn Heer, BEng, MBA Toyota Production System Support Center (TSSC), Toyota Motor Corporation Australia, Melbourne, AustraliaDavid P. Johnson, MD Monroe Carell Jr. Children’s Hospital at Vanderbilt, Department of Pediatrics, Vanderbilt University School of Medicine, Nashville, TN, USADarren Klugman, MD, MS Cardiac Intensive Care Unit, Children’s National Heart Institute, Children’s National Hospital, Cardiac Critical Care Medicine, Washington, DC, USANicole M. Leone, MD Cohen Children’s Medical Center, Zucker School of Medicine at Hofstra/Northwell, Department of Pediatrics, New Hyde Park, NY, USAContributors

xvKheyandra D. Lewis, MD Drexel University College of Medicine, St. Christopher’s Hospital for Children, Department of Pediatrics, Philadelphia, PA, USADonald E. Lighter, MD, MBA, FAAP, FACHE Institute for Healthcare Quality Research and Education (IHQRE), and Physician Executive MBA Program, University of Tennessee, Department of Business Administration, Knoxville, TN, USATensing Maa, MD Nationwide Children’s Hospital, Division of Pediatric Critical Care Medicine, Columbus, OH, USAChristopher D. Mangum, CSSBB Children’s Hospital of The King’s Daughters, Department of Quality, Norfolk, VA, USARichard E. McClead Jr, MD, MHA Nationwide Children’s Hospital, Hospital Administration, Columbus, OH, USAStacy McConkey, MD Adventhealth for Children, Graduate Medical Education, Department of Pediatrics, Orlando, FL, USAHeather S. McLean, MD Duke Children’s Hospital, Department of Pediatrics, Duke University Medical Center, Durham, NC, USAJennifer E. Melvin, MD Nationwide Children’s Hospital, Department of Emergency Medicine, Columbus, OH, USAJenna Merandi, PharmD, MS, CPPS Nationwide Children’s Hospital, Pharmacy Department, Columbus, OH, USAMary M. Morin, RN, MSN Clinical Effectiveness and Employee Health Services, Sentara Healthcare, Virginia Beach, VA, USARustin B. Morse, MD, MMM Nationwide Children’s Hospital, Columbus, OH, USABrigitta U. Mueller, MD, MHCM ECRI, Plymouth Meeting, PA, USAKaren Murrell, MD, MBA Performance Improvement, TeamHealth, Knoxville, TN, USAKapil Raj Nair, MSHS Beterra Health, Newnan, GA, USAHolly O’Brien, MSN RN CPPS Children’s Hospital of Wisconsin, Department of Quality and Safety, Milwaukee, WI, USASarah Henrickson Parker, PhD Center for Simulation, Research and Patient Safety, Carilion Clinic and Fralin Biomedical Research Institute at Virginia Tech, Roanoke, VA, USAShilpa J. Patel, MD University of Hawaii, John A. Burns School of Medicine, Kapi‘olani Medical Center for Women & Children, Department of Pediatrics, Honolulu, HI, USAContributors

xviMichael F. Perry, MD Nationwide Children’s Hospital, Department of Hospital Medicine, Columbus, OH, USADenise Ranada, MSN, RN Kaizen Promotion Office, Harbor-UCLA Medical Center, Torrance, CA, USAChristine M. Riley, MSN, APRN, CPNP-AC Cardiac Intensive Care Unit, Children’s National Hospital, Washington, DC, USAMichele Saysana, MD Indiana University Health, Indiana University School of Medicine, Indianapolis, IN, USALisa L. Schroeder, MD Children’s Mercy Kansas City, University of Missouri-Kansas City School of Medicine, Department of Medical Administration, Kansas City, MO, USARahul K. Shah, MD, MBA Children’s National Hospital, Washington, DC, USAJodi L. Simon, MSHA, BS Akron Children’s Hospital, Department of Quality Services, Akron, OH, USAJack Stevens, PhD Nationwide Children’s Hospital, Ohio State University Department of Pediatrics, Columbus, OH, USAKatherine Stringer, BSPH Kaizen Promotion Office, Harbor-UCLA Medical Center, Torrance, CA, USAAnupama Subramony, MD, MBA Cohen Children’s Medical Center, Zucker School of Medicine at Hofstra/Northwell, Department of Pediatrics, New Hyde Park, NY, USAJackie E. Valentine, RPh, MHA Seattle Children’s Hospital, Department of Patient Safety, Seattle, WA, USABrian Wagers, MD Indiana University School of Medicine, Riley Hospital for Children, Department of Emergency Medicine and Pediatrics, Indianapolis, IN, USAJeffrey Scott Warshaw, BA, MS San Diego County Office of Education, Learning and Leadership Services, San Diego, CA, USALaurie Wolf, PhD Carilion Clinic, Clinical Advancement and Patient Safety, Roanoke, VA, USADaniel B. Wolfson, AB, MA, EdD San Diego County Office of Education, Learning and Leadership Services, San Diego, CA, USAJoshua Wolovits, MD UT Southwestern Medical Center, Cardiac ICU, Children’s Health, Dallas, TX, USAContributors

xviiAbout the EditorsRahul K. Shah, MD, MBA obtained a combined BA/MD degree from Boston University School of Medicine (2000), thereafter completing an otolar-yngology residency (Tufts University) and a pediat-ric otolaryngology fellowship (Children’s Hospital Boston, Harvard University). He joined the faculty of Children’s National Medical Center (2006), ris-ing to the rank of Professor (2017) at George Washington University School of Medicine and Health Sciences. Dr. Shah’s research interests include resource utilization and outcomes, patient safety, and medical errors; he has received numer-ous awards for his research. He is recognized as a leader in patient safety and quality improvement, and has chaired and serves on myriad national com-mittees related to patient safety and quality improve-ment. Dr. Shah was Executive Director of the Global Tracheostomy Collaborative, an interna-tional not-for-profit quality improvement initiative. He was the inaugural Associate Surgeon-in-Chief at Children’s National Medical Center and the Medical Director of Peri-operative Services from 2011 to 2014. Dr. Shah served as President of the Medical Staff at Children’s National Medical Center from 2012 to 2014. In 2014, he was appointed the inau-gural Vice President, Chief Quality and Safety Officer for Children’s National Health System and in 2018 was appointed the inaugural Vice President, Medical Affairs as an additional executive responsi-bility; he has served as the acting Chief Medical Information Officer (July–December 2019). Dr. Shah has authored over 130 peer-reviewed articles and has given hundreds of national and international presentations. Under his leadership, Children’s National has received numerous safety and quality distinctions and is a recognized leader in pediatric safety and quality.

xviiiSandip A. Godambe, MD, PhD, MBA is a phy-sician leader who obtained a combined MD-PhD degree from Washington University School of Medicine’s Medical Scientist Training Program. He then completed a pediatrics residency (Boston Children’s Hospital, Harvard University) and pediatric emergency medicine (PEM) fellowship (University of Tennessee, Le Bonheur Children’s Hospital). He worked briefly at Norton Children’s Hospital and then joined the faculty at the University of Tennessee as the Co-Medical Director of Emergency Services. Dr. Godambe obtained his MBA degree with a focus on quality (University of Tennessee) and then became the inaugural Medical Director of Medical Staff Quality. He moved to Children’s Hospital of The King’s Daughters (Norfolk, VA) where he became the inaugural Vice President of Clinical Integration and Quality and the Chief Quality and Safety Officer. Dr. Godambe has led CHKD to numerous quality and safety awards on their journey to becoming a high-reliability organization. As a Professor of Pediatrics, Vice Chair of Pediatrics – Quality and Safety, and Co-Program Director of the Improvement Science Fellowship with Eastern Virginia Medical School, he leads many educa-tional venues for students and trainees with regard to quality and safety. He is recognized as a leader in patient safety and quality improvement and has led or served on a myriad of state and national committees related to healthcare quality, safety, and emergency medicine. He is the regional co-leader for the Atlantic subsection of Children’s Hospital Solutions for Patient Safety (CH-SPS) and a clinical steering committee member for the National CH-SPS and Child Health PSO. He has served as a Senior Examiner for the Baldrige Performance Excellence Program. He is well versed in Improvement Science through his work experience and training in Lean, Six Sigma, Institute of Healthcare Improvement (IHI) Model for Improvement, and the Toyota Production System. He is currently an IHI Improvement Advisor. He has authored over 100 publications, chapters, and abstracts in emergency medicine, quality, and immunology. He is the co-editor of About the Editors

xixfive books: multiple editions of the 5-Minute Fleisher and Ludwig’s Pediatric Emergency Medicine Consult, PEM Question Book, and this textbook. He currently serves on the editorial boards of two journals and is a reviewer for mul-tiple clinical, safety, and quality journals. He has given over 200 national and international presentations.About the Editors

1© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 R. K. Shah, S. A. Godambe (eds.), Patient Safety and Quality Improvement in Healthcare, https://doi.org/10.1007/978-3-030-55829-1_1Introduction: A Case-Based Approach to Quality ImprovementSandip A. Godambe and Rahul K. ShahS. A. Godambe (*) Children’s Hospital of The King’s Daughters Health System, Norfolk, VA, USAe-mail: [email protected] R. K. Shah Children’s National Hospital, Washington, DC, USAe-mail: [email protected] 1.1A tertiary care free-standing hospital has a problem with catheter-associated urinary tract infections (CAUTIs). This problem is not new. The organization tackled CAUTIs 4 years prior with the creation of an over-arching structure which resulted in new processes and better outcomes. As the com-pliance with these refined processes improved, the absolute number of CAUTIs went down. However, in the past 18 months, the number of CAUTIs has slowly crept back up. This issue is further compounded by the fact that the rate has significantly worsened even as the organization has reduced their Foley catheter days dramati-cally. The clinicians only place catheters when they are most needed; hence the numerator has increased, while the denom-inator has decreased in the CAUTI rate equation. The executive leadership and Hospital Board demand an improvement from the quality and safety team. This can be the self-defeating prophecy for many teams trying to reduce the CAUTI rate – the absolute number of events is decreasing but the rate (which is used for benchmark-ing) continues to increase.Chapter Objectives• To demonstrate the burning platform of patient safety and quality improvement in the current healthcare era as it relates to the achievement of zero harm• To explain how varying improvement methodologies can co-exist to drive improvement in an organization with the use of an adapted simple, common language that fosters improvement across all layers of the enterprise• To connect the work of patient safety and quality improvement to the mission, vision, and values of an organization• To understand the value of learning best practices and methods from non- healthcare industries

2 Opening Question/ProblemThis chapter is not about CAUTIs or specific tac-tics to reduce these infections – that will be dis-cussed elsewhere in this text. Rather, this chapter discusses the improvement framework and approach toward patient safety and quality improvement that transcends individual hospital acquired conditions and can be broadly applied to quality improvement initiatives in the organization. Introduction/OverviewThere have been significant strides made to advance patient safety and quality improvement in the past two decades. Hospitals, and other organizations, reacted to the clarion call from the Institute of Medicine’s seminal work, To Err is Human [1]. Since this publication, hospitals and healthcare systems have made tremendous invest-ments in people, processes, and technology – all with an aim to improve the quality and safety of care delivery. We have seen improvement; how-ever, there are issues that still persist and have not improved at the same rate as other measures. Many organizations are struggling with their progress toward zero harm; they have seen a pla-teau in their improvement and are looking for novel approaches and strategies.Early in the journey, there was an educational component which was missing in this work. As such, initial efforts were appropriately targeted toward increasing capability (the ability, from a skills perspective, of healthcare workers to embark upon quality improvement initiatives) (Key Point Box 1.1).Much of the efforts immediately after To Err is Human focused on extrapolating the theoreti-cal underpinnings from systems science, reli-ability, and quality improvement from other industries to educate those of us in healthcare. This was initially quite successful, as there was a whole new lexicon introduced into healthcare. Previously fertile ground was now inundated with theoretical quality improvement applica-tions. As expected, improvement followed as the proverbial low-hanging fruit (Fig. 1.1) was harvested. Some of the success in the early 2000s was a result of the Hawthorne effect (which states that improvement will occur when those performing the work know they are being observed); however, not surprisingly, in many instances, these results were not sustained (Key Point Box 1.2).Nevertheless, healthcare was quick to embrace this renewed interest in the safety of their patients Sweet fruitGround fruitBulk of fruit5,6 σ : Address designs 1,2 σ : Logic and intuition4 σ : Improve internally3 σ : Demand improvementLow hanging fruitConcept and design : Rahul K. ShahFig. 1.1 Climbing the quality tree. (Image courtesy of Rahul K. Shah)Key Point Box 1.2 SustainA common problem in quality improvement is the ability to sustain projects for prolonged periods of time. Smart aim statements usu-ally include verbiage to indicate the degree of improvement over a prescribed period of time (6, 9 months, etc.). It is the leader’s role to ensure that the project “sticks” and that true improvement is achieved.Key Point Box 1.1 Capability Vs. CapacityCapability – the intellectual understanding, knowledge and practical application of improvement scienceCapacity – the ability to take on quality improvement projects

3and the quality of care delivery; furthermore, the public, government, and payers were expecting such improved care to be delivered quickly.During the past decade, it has become clearer that the low-hanging opportunities have been addressed. A clear understanding of the journey of healthcare improvement, via the continuous quality improvement framework, resulted in organizations realizing several disadvantages. They were in for the long-haul and real improve-ment would take years, not months. Improvement would be elusive, rather than straightforward. It would yield further disappointments, not all success.To increase, or at least continue, their trajec-tory of improvement, health systems need to change their level of sophistication. Figure 1.1 demonstrates a rubric, and guiding principle, used and presented by one of the editors (RKS) in explaining the complexity necessary to continu-ally improve outcomes for our patients. To under-stand where healthcare is at present in the quality improvement journey, one can overlay the improvements in healthcare, since 2000 to pres-ent, with the level of sophistication necessary to achieve sustained outcomes (Fig. 1.1).In the early 2000s, much of the improvements were a result of targeting low-hanging fruit and using basic resources to drive improvements. We would train teams on whatever improvement methodology aligned with our organizational quality improvement teams (Institute for Healthcare Improvement (IHI) Model for Improvement [2], Lean, Six Sigma, etc.). Usually, that basic theoretical education sufficed to collect the “easy to reach” improvement opportunities. This was essentially the era of demanding improvement.As we evolved our understanding and tech-niques, the issues became more complex and mandated differing strategies. Organizations started collectively focusing on improvement. Improvement science transcended the quality improvement department, such that it was con-sidered to be the job of hundreds of individuals in an organization. When leadership held teams (and themselves) accountable for outcomes and demanding improvement, said improvements were made to a higher degree of reliability. The next evolution in outcomes will require structures and processes that have specific and unique inter-nal improvements and address systems design.Healthcare is emerging from its, at times, insular history and is now turning to other indus-tries such as our airline counterparts, Toyota, the US Navy, Alcoa, and others, for models of opera-tional excellence that support a culture of safety and continuous process improvement. Dr. W. Edwards Deming [3] spoke of the importance of systems thinking as a key ingredient for improvement. His System of Profound Knowledge consists of four key points: Vignette 1.2Four years prior, the organization made the reduction of hospital acquired conditions, especially infections, a priority. A new structure was put in place. A physician and nurse co-led the CAUTI team which also included stakeholders from the inpatient floors, the operating room, and the emer-gency department. The team chartered this work and put in place processes to address the key drivers from their CAUTI road map, using the IHI Model for Improvement. The initial results were impressive – an 80% reduction in CAUTIs in just a few years. However, over the past 18 months, outcomes have slipped, and there has been an increase in CAUTIs. Much has changed in the past 5 years in hospitals with regard to quality improvement. The CAUTI team believes that they need to refresh their qual-ity improvement approach. They are strug-gling with how to do this with competing organizational priorities. This is further complicated by the ever-changing national perspective, and potentially competing improvement methodologies, which may be frustrating staff.

4appreciation of system, theory of knowledge, psychology of change, and understanding varia-tion. It has had significant impact on some of the aforementioned models of excellence [4]. The Theory of Knowledge incorporated the Plan-Do-Study-Act (PDSA) cycle which is the most com-monly discussed unit of improvement science-directed change.Dr. Donabedian emphasized the importance of systems awareness and design [5]. His widely used theoretical framework (commonly referred to as the Donabedian triad) is composed of three crucial points: structure, process, and outcome. In our organizations, we employ the Donabedian quality triad when embarking on projects or when delving deeper to understand why a sys-tem is not performing as expected (Key Point Box 1.3). By having the improvement team take a step backward and move “upstream” from the outcome, the role of structure and process becomes clear. The improvement team needs to look beyond outcomes and ask the provocative questions of what structures are in place and if we are holding teams accountable for the pro-cesses that we deem necessary to drive improvement.With a trend in CAUTIs that was contrary to our global aim, and continuing to affect patients, this organization took a pause. They evaluated not only the structure and processes but took a higher level approach to ask if they were using the correct methodologies. The initial key driver diagram from 4 years ago was reviewed and refreshed. Many members of the prior team had moved on from the organization or were not actively involved in the present work. A revised key driver diagram was created and shared throughout the organization.There exist several quality improvement methodologies and myriad permutations of the foundational methods. Many healthcare organi-zations are steeped in the understanding of the IHI Model for Improvement and Lean [2, 4]. The IHI Model for Improvement uses a conceptual framework to understand variation, clarify pro-cesses, plan tests of change, and measure and accelerate improvement and includes aims, key drivers, and measurement. Lean is an improve-ment methodology based on the tenets of reduc-ing waste and driving efficiency. It was derived Key Point Box 1.3 Donabedian Quality TriadStructure, Process, OutcomeVignette 1.3The initial work in CAUTIs for the organi-zation started approximately 4 years prior. The organization was admittedly and knowingly behind other organizations as they had lost focus and sustainment. To address this, a structure was put in place. Not only was thought given to the constitu-ency of the team (size, representation, need for contrarians, etc.) but also to its report-ing structure. The improvement team was explicit in its desire to recruit an executive sponsor to champion the work and provide organizational alignment – ultimately between executive management and the Board. The committee was chartered and reported to progressively more influential hospital level quality committees. The absolute number of CAUTIs were tabu-lated monthly and presented in a collated format, along with the other hospital acquired conditions, to management, lead-ership, and the Board in a consistent fash-ion. Once the improvement team’s membership and reporting structure had been clearly delineated, attention was turned to processes. For the CAUTI work, best practices were gleaned from literature, national collaboratives [6], hospital associ-ations, and infectious disease experts. In turn, a decision was made to adopt a bundle from a national collaborative. The bundle, consisting of five items, was adopted and adherence to it was measured.

5from the Toyota Production System (TPS) [7, 8] and focuses primarily on its technical tools.A case-based approach to quality improve-ment cannot be wedded to a particular quality improvement methodology. Organizations should have some latitude and resist being vehemently dogmatic, on which improvement methodology is employed. Of course, it is strongly suggested that an organization have a predominant methodology for quality improvement that is understood by the entire organization. However, to climb the quality tree, it must be conceded that, at times, additional methodologies may need to be incorporated into the strategy. Furthermore, we would caution that being resistant to ideas from other staff about their preferred improvement methodology may harm improvement culture in the long run.Simplicity is crucial to the message for our frontline team members, who may not under-stand the complexities of improvement and safety science, as they are the agents driving change. Recall the aforementioned discussion about the apple tree (Fig. 1.1). Now realize that these apples need to reach their customers or our front-line team members. The more complex the bureaucracy or the language, the more likely that the apples will spoil and not reach the mouths of our frontline team members. This would be crip-pling, as they are hungry for the skills that will make them better problem-solvers (Fig. 1.2). We need to realize that improvement science, while having multiple theoretical models, can be sim-plified to a common local language that is inclu-sive and respectful of all methods while still facilitating change across the health system continuum.1007550250ComplexModerateSimple languageLess engagementMore engagementSimplicity of message to frontline team membersFrontline engagement and likelihood for successBetterFig. 1.2 Simplicity of the message and delivery of the fruit of the quality tree. (Image created by Eric Cardenas)Vignette 1.4After the appropriate structure had been put in place with the necessary multi- disciplinary stakeholders, a clear reporting structure to executive leadership and the

6It is clear that operational success requires systems thinking and realignment which, in turn, requires a structured framework. Some frame-works are inherently complex, such as the Baldrige Framework for Performance Excellence [9], and require considerable organizational and individual commitment and planning. Others appear to be simple like the TPS (Fig. 1.3) which emphasizes the development of individuals, with a focus on the frontline and customers, and the creation of teams of problem-solvers that readily bring problems to the surface. The authors are not advocating for one over another – they each have a role. It is important for the reader to understand the basic tenets of these frameworks. The reality is that the ability to do the latter (TPS) well takes considerable organizational commitment and alignment and probably has not been mastered perfectly by any health system at the time of writ-ing of this text. Jamie Bonini, Vice President of Toyota Production System Support Center (TSSC), best described TPS as “an organizational culture of highly engaged people solving prob-lems (or innovating) to drive performance” (per-sonal communication). Implied in this statement is the importance of transparency, accountability, a focus on developing our frontline team mem-bers, and supporting a problem-solving culture. Toyota Production System TriangleMotivate % developpeople to sloveproblems“sea of problemsolvers”Use TPS toolscorrectly to exposeproblems“learn by doing”ManagerialTechnicalPhilosophyCustomer firstPeople are our mostvaluable resourceContinuousImprovementShop floor focus1.2.3.4.Modified with permission from Toyota Motor Co.Quality Improvement& SafetyHUMANDEVELOPMENTFig. 1.3 Toyota production system triangle. (Modified from an original figure from Toyota. Used with the permission of Toyota)Board, and an involved executive sponsor, attention was turned toward traditional quality improvement methodologies. A key driver diagram with a global aim, SMART (specific, measurable, applicable, realistic, timely) aim, appropriate drivers, and inter-ventions was created and then shared broadly throughout the organization. The key driver diagram and review of the CAUTI processes and outcomes were eval-uated by the Chief Quality Officer on a monthly basis. Resources (educational, personnel, financial, etc.) were deployed to the micro-units in need to properly rein-vigorate their teams. Small groups of front-line individuals were pushing back that they had competing priorities and were unable to do their core work. The CAUTI steering committee was appropriately wor-ried that this would, once again, set back the improvement project.

7This cannot be achieved overnight. Frankel et al. [10] proposed a Framework for Safe, Reliable, and Effective Care (Fig. 1.4) which describes the culture and learning system domains as being foundational and crucial to the success of safety and quality systems.Quality improvement efforts in a healthcare organization need to be cognizant of the organi-zational Culture (intentionally with a capital “C”). Culture is the shared norms of a system. There are hundreds of definitions of Culture. Indeed, each organization most likely uses some permutation of the aforementioned definition. The CAUTI vignette, which has been carried through this introductory chapter, has Culture as a key component. The authors and editors of this text have shared many examples of how quality improvement initiatives fail, or are not sustained, primarily due to the lack of appreciation of the importance of Culture. There is no quick fix or methodology to improve Culture. It is beyond the scope of our introductory chapter, in this case- based approach to quality improvement textbook, to expound upon Culture. However, it must be appreciated in these case vignettes that efforts to drive quality improvement, without an under-standing and appreciation of Culture, will not be successful.Creating an environmentwhere people feel comfortableand have opportunities to raiseconcerns or ask questions.Facilitating and mentoringteamwork, improvement,respect, and psychologicalsafety.Openly sharing data and otherinformation concerning safe,resepectful, and reliable care withstaff and partners and families.Applying best evidence andminimizin g non-patient-specific variation, with thegoal of failure-free operationover time.Improving work processes and patientoutcomes using standard improvementtools, including measurements over time.Regularly collecting andlearning from defectsand successes.Gaining genuineagreement on mattersof importance to teammembers, patients, andfamilies.Developing a sharedunderstanding, anticipationof needs and problems, andagreed-upon methods tomanage these as well asconflict situations.Being held to act in a safe andrespectful manner, given thetraining and support to do so.Learning systemCulture•••••••••PsychologicalsafetyAccountabilityLeadershipTransparencyReliabilityImprovement&measurementContinuouslearningNegotiationTeamwork & communicationEnagement ofPatients & FamilyFig. 1.4 IHI framework for safe, reliable, and effective care [10]. (Reprinted from www.IHI.org with permission of the Institute for Healthcare Improvement, ©2019)Vignette 1.5It was found, when digging deeper into the CAUTI outcomes, that the operating rooms and emergency department did not espouse the same values and Culture with regard to CAUTI as that held by the inpatient units. One can immediately see the problem and how it can spiral into a bigger issue. If two of the three stakeholders had a different cul-tural approach to CAUTIs, then there would be no shared mental model. The emergency department and operating rooms did not feel ownership of the issue, as they believed that their care was transient and the patient was ultimately admitted to the inpatient unit. To break this cultural logjam, the Chief

8A successful approach to those stakeholders that are recalcitrant, or do not see an issue as a “problem” to be owned, is to move the issue to a higher level and focus on the mission, vision, and values of the organization. This is not a quick solution, and the recalcitrant leader may need to be reminded frequently, perhaps at the start of each meeting on the topic, of their role in the organization and how that ties into the mission, vision, and values (Key Point Box 1.4).It would be disingenuous to state that, imme-diately after this meeting, these groups were engaged. Culture change takes time – often years. Once the Chief Medical Officer had the small group meet, she further charged them to report back to her monthly with their CAUTI data. At subsequent meetings, the Chief Medical Officer made it clear that the three leaders were account-able for the CAUTI outcomes in the organization.Rates are often used in quality improvement and take various forms in their presentations. The most common is the number of events divided by a frequency. For CAUTI, the rate is usually expressed as the number of catheter-associated urinary tract infections divided by the number of catheter days. Some individuals (board members, executive leadership, or non-clinical leaders) may not be able to immedi-ately grasp the significance of small changes in rates as having an impact on patients, especially as we near zero. Dr. Richard Brilli, Chief Medical Officer at Nationwide Children’s Hospital, has been a proponent on using actual Medical Officer brought the leadership of these three areas together in a small group meeting. The objective of the meeting was to discuss, in an open forum, why two of the stakeholders were not appreciating their team’s role in CAUTIs. Contrary to one’s impression, the 1-hour meeting did not per-severate on the pathophysiology of CAUTIs nor on specific tactics and strategies to reduce CAUTIs. Rather, a significant por-tion of the meeting addressed the mission, vision, and values of the organization. By elevating the meeting to a shared under-standing of the organization’s commitment to their patients, families, and community, the Chief Medical Officer was able to imbue the organization’s desired Culture to these teams. Of course, this broader realization did not happen overnight. The initial meet-ing with the Chief Medical Officer put in motion the goals of the three teams and laid out how their work on CAUTIs would be a microcosm of the bigger work and global aim.Key Point Box 1.4 Mission, Vision, ValuesMission – the role of the organizationVision – forward-looking statement of what the organization wants to achieve in the futureValues – principles and ideals that bring the organization togetherVignette 1.6The hospital’s Board had heard about the increase in CAUTIs and wanted this to be presented by the Chief Quality Officer at the next Board meeting. The Chief Quality Officer struggled with presenting the data as an absolute number of cases versus a rate (numerator/denominator). The Chief Quality Officer had also contemplated the best manner in which to show the executive leadership and Board other hospital acquired conditions. She believed that if the Board was engaged at present, and ask-ing for data regarding CAUTIs, she should seize this moment and put CAUTIs in con-text with other hospital acquired condi-tions. She struggled with how to best show the Board the entirety of the information in an understandable and meaningful way.

9event frequency data, as well as rates, to help organizations understand the scope of a prob-lem [11]. It is much more tangible for leader-ship, Boards, and frontline team members to know that there were, for example, 17 CAUTIs in the past year and 3 in the past quarter. To tersely state a rate for this audience would not be providing them the full context. As we con-tinue to climb the quality tree, outcomes are going to significantly improve, and the numera-tors (number of actual events) will continue to fall. Rates should also continue to drive down to zero. Dr. Brilli was among the first to stress the importance of zero as our goal for harm reduction. Tacit in this goal is that we may need to be agile in how we present our data – some-times as an absolute number of events and other times as a rate.When faced with an improvement project, it is crucial that the initiative is aligned with the orga-nizational mission, vision, and values (Fig. 1.5). We have seen that, in our institutions and when working with other organizations, both the front-line and executive leadership need to be able to see how their work connects to the trajectory of the organization.At Children’s National, under the leadership of our Executive Management and Board, we embarked upon a journey in which the account-able executives over Patient Care Services (Chief Operating Officer and Chief People Officer) crafted contemporary organizational values. These values are Compassion, Commitment, and Connection (Fig. 1.6).The importance of explicitly stating the orga-nizational core values, and using them as levers to drive engagement and improvement projects forward, cannot be understated (Fig. 1.5).Path to success1. Strategic objective2. Strategic objective3. Strategic objective4. Strategic objective5. Strategic objective6. Strategic objectiveEvaluation and improvementSystem strategic planSystem balanced scorecardLeadership scorecardDepartmentbalanced scoreboardDepartmentbalanced scoreboardDepartmentbalanced scoreboardDepartmentbalanced scoreboardDepartmentbalanced scoreboardDepartmentbalanced scoreboardVisionMissionValuesTo provide world-classhealth careTo be an independent,non-profitorganization and toprovide innovative,comprehensive care ofthe highest quality,always exceedingcustomerexpectations.QualityCompassionConfidentialityDignity/respectEqualityIntegrityIndividual performance reviewFig. 1.5 Organizational structure needed for successVignette 1.7Despite aligning organizational awareness around CAUTIs by using the Harm Index to demonstrate to the employees and Board that this issue was still pervasive, some employees were not making the connection to bundle compliance and the goals of the organization. Frontline employees were completing the CAUTI bundles approxi-mately 50% of the time on average, and, when looking at various microsystems, the bundle compliance ranged from 30% to 70%. Therefore, the quality improvement team was not surprised that the organiza-tion was still having a CAUTI every

10Starting each pertinent meeting with a safety story is hugely impactful. A safety story is a brief vignette of an event that occurred in the organiza-tion, region, or otherwise, told by a member of the committee. The story should be brief (90 sec-onds or less), and the chair of the committee should provide just a couple of minutes of discus-sion to connect the story, address open items, and move the meeting to the agenda items. An exam-ple of a patient safety story presentation would be: “I would like to start this meeting off with a patient story. The patient was on the hospitalist service on hospital day #3 when she spiked a fever. The child had multiple lines and was admit-ted for an aggressive respiratory infection. The child was pan-cultured and found to have a uri-nary tract infection with a Foley catheter, so this was deemed to be a CAUTI. The child was trans-ferred to the ICU for urosepsis, and required aggressive antibiotic therapy for 3 days. She was then discharged home after a total hospital stay of We serve all with warmth and kindness•Care for ALL children, and respect and value ALL colleagues.•Honor the diversity of all patients, families, staff, and the communities we serve.•Seek out diverse thoughts and perspectives.•Support and encourage those around us. We lead with dedication and discovery•Challenge ourselves, innovate, and push each other to always excel.•Demonstrate integrity and focus on what’s right, not just what is required.•Be a leader in education, learning, and improvement.•Be honest and speak up when we see a problem. We team up for success•Always put patients and families at the center of what we do.•Be present and focused in all our interactions.•Empower families and each other with the guidance to make decisions.•Reach out and create partnerships across teams. Fig. 1.6 Children’s National core values and behaviors with alignment to the organizational mission. (Courtesy of Children’s National Hospital, Washington, DC)45 days. The CAUTI steering committee heard from frontline staff that they believed there was no connection of their work to the goals of the organization. The CAUTI steering committee began to change their messaging. The team began to include the “why.” The leaders of this work started each of their CAUTI meetings with a patient story that related back to the organi-zational mission, vision, and values. Often times, a non-CAUTI story was utilized. This tactic spread organically through the organization and, before long, patient safety stories were shared at the top of each pertinent meeting. The patient stories gen-erally lasted about 2–3 minutes and were strategically used to connect the meeting, and work of the team, to the mission of the organization.

119 days. When reviewing the risk factors for the CAUTI, it was noted that the unit’s bundle com-pliance for CAUTI is only 60%. For this child, the CAUTI bundle was not performed each time for all elements. As a side note, the hospital cen-sus is high and the ICU is at full capacity” (Key Point Box 1.5).The specifics of how to tell a patient story are important to share as the authors have often seen patient stories taking 10–15% of an allot-ted meeting or note stories that are not con-nected back to the meeting agenda. Other times, the stories are so profoundly impactful (e.g., patient death or egregious deviation from care) that a portion of the meeting must be used to immediately address some area(s) of concern identified in the patient story. Such a story is not effective if it did not achieve its goal of con-necting the dots for the committee members and grounding the team in their work, but instead “hijacked” the agenda from the meet-ing’s intended purpose. A safety story should be a succinct vignette, preferably related to the organization in some manner and presented in 90 seconds or less, that is used to demonstrate organizational alignment and the work of those in the meeting. Additionally, the importance of patient and family participation on improve-ment committees cannot be overstated. We need to remember that, at some point, all healthcare providers will also be consumers of healthcare. We would want to be given the same respect and ability to be involved in our care or the care of our loved ones.We are confident that this introduction has provided the reader an idea as to what to expect in the ensuing chapters. Quality and safety is the paramount priority of most organizations glob-ally and unequivocally for healthcare organiza-tions. Naturally, there is much information as to how to proceed, but the journey to zero harm requires careful planning and time. Success takes a shared vision, simple and measurable strategic objectives, leadership and frontline engagement, common operational language, perseverance, and the desire to succeed.Thomas Nolan’s Framework for Execution [12] (Fig. 1.7) and the Toyota Production System Achievestrategic goalsManage localimprovementDevelop humanresourcesSpread and sustainProvitde leadersfor large system projectsProvide day-to-dayleaders for microsystemsFig. 1.7 Framework for execution. (Reprinted from www.IHI.org [12], with permission of the Institute for Healthcare Improvement, ©2019)Key Point Box 1.5 What Is a Bundle?A bundle is a group of process interven-tions (almost always evidence-based) put into place for a specific metric, which has been demonstrated to improve outcomes.

12Triangle (Fig. 1.3) are some of the simplest rep-resentations by which to drive improvement. Both will be discussed in the ensuing chapters. Nolan discussed the criteria necessary for break-through performance: (1) to define breakthrough performance goals; (2) to create a portfolio of projects that support these goals; (3) to deploy appropriate resources to ensure the success of these goals; and (4) to create the oversight and learning system to monitor and ensure success. High reliability, as discussed by Weick and Sutcliffe [13], is the goal for healthcare enter-prises and their combined membership. It is not for the faint of heart, but it remains elusive until the many aforementioned criteria are attained. We will be discussing their various components in depth in this text. Chapter Review Questions 1. Describe how quality improvement strategies have evolved over past decades.Answer: Initial quality improvement strate-gies focused on the low-hanging fruit, and, as improvements occurred, it became necessary to move to higher levels of sophistication and reliability. At present, organizations are on dif-ferent parts of the quality journey, and, as such, their improvement strategies have differing levels of sophistication (Fig. 1.1 and 1.2). 2. What is the difference between capacity build-ing and capability building?Answer: Capability building is the “abil-ity,” or skill set, for improvement science. Capacity is the “time,” resources, or organiza-tional ability, to improve. An individual may have capacity to lead improvement, but an improvement initiative will be stymied with-out capability.Editors’ CommentsEach chapter will be followed by a synop-tic chapter summary by the editors to put the article into the broader context of the textbook and healthcare quality improve-ment overall. To simply reiterate the abstract would not be of value. Rather, this concluding section for each chapter will attempt to pull the chapter and textbook together and be forward-looking in nature for the reader.This introductory chapter attempts to rekindle the burning platform in healthcare by pushing us to strive for zero harm. To do this, we implore readers to strive for zero harm. To increase the level of sophistica-tion in quality improvement, the authors stress the importance of the Donabedian quality triad of structure, process, and out-comes. In beginning quality improvement projects and when evaluating those that are in sustain mode, it is crucial to ensure the project has the right structure and that pro-cess measures are being completed and sustained as expected with controls in place for accountability.A key goal of this chapter is to also drive home the concept of absolute numbers of harm compared to a rate and how to engage an organization’s Board to understand and be able to participate in discussions regard-ing hospital-acquired conditions. Additionally, engagement of our frontline team members, our patients, and their fam-ilies is needed for success. We need a com-mon and simple operational language which everyone can understand and rally around.Finally, Culture is important when eval-uating why a quality improvement project has stalled or is not achieving the desired outcomes. Understanding your organiza-tional Culture and ensuring its alignment with quality improvement efforts is com-pulsory, especially with stalled initiatives. Many times, Culture is not explicitly addressed and is evaded to avoid poten-tially difficult conversations. One must use the levers necessary to prioritize and high-light the role of Culture in quality improve-ment initiatives.

13 3. How does Culture influence quality improve-ment initiatives?Answer: Culture drives improvement. Without attention to Culture, much improve-ment will be the result of the Hawthorne effect and will not be able to be sustained. The value of Culture development cannot be underestimated. 4. What are the elements of the Donabedian quality triad?Answer: Structure, Process, Outcome. 5. How can an organization’s mission, vision, and values be used as levers for quality improvement?Answer: It is crucial that staff appreciate and understand their role in quality and safety and how it aligns with the organization’s role. The mission, vision, and values help the front-line staff, manager, leader, and Board member connect their safety and quality work with organizational improvement efforts. 6. How can patients and their families be incor-porated into organizational quality improve-ment initiatives?Answer: It is imperative to include the voice of the family and patient in organiza-tional quality improvement. If we fail to include these stakeholders, then our work is not complete. It is quite easy to include patients and families by working with your Patient/Family Advisory Council, Volunteer Services, or other such liaisons in your organization. 7. Describe the characteristics of the ideal sys-tem for continuous process improvement.Answer: The purpose of this question is to get our readers to start thinking about the ideal system for continuous process improvement. The remaining chapters of this text provide further insights, and we will return to this very question throughout the text in the editor’s comments. For now, we will state that the ideal system for continuous process improve-ment understands this is difficult work that takes considerable organizational planning and foresight. Capability and capacity need to be built at the frontline level with significant senior leadership, and Board, commitment and visibility. The goal of this system is to develop processes and procedures that are clear, simple, and understandable and that occur reliably. The organizational culture needs to encourage bringing problems to the surface and, for the most part, local ownership of problem-solving. 8. True or False: Healthcare systems are unique and complex, so few concepts from other industries are applicable to healthcare.Answer: False. Healthcare has learned, and continues to learn, much from other indus-tries. Specific examples are included through-out the chapter. 9. Based on the discussions in this chapter, which of the following is important to carrying out a successful quality improvement project? A. Alignment with organizational goals and priorities. B. Inclusion of patients and/or their families. C. Assigned accountability and visible sup-port of senior leadership. D. Supportive culture that permits transparency. E. All of the above.Answer: E.References 1. Institute of Medicine. To err is human: build-ing a safer health system. Washington, DC: The National Academies Press; 2000. https://doi.org/10.17226/9728. 2. Institute for Healthcare Improvement. Boston, MA, 2019. http://www.ihi.org/. Accessed 30 Dec 2019. 3. Deming WE. The new economics for industry, government, education. 1st ed. Cambridge, MA: Massachusetts Institute of Technology, Center for Advanced Educational Services; 1994. 4. Langley GJ, Moen RD, Nolan KM, Nolan TW, Norman CL, Provost LP. The improvement guide. 2nd ed. San Francisco: Jossey-Bass Publishers; 2009. 5. Donabedian A. The quality of care: how can it be assessed? JAMA. 1988;260(23):1743–8. 6. Children’s Hospital Solutions for Patient Safety (CH-SPS). Cincinnati, OH. 2019. https://www.solu-tionsforpatientsafety.org/. Accessed 30 Dec 2019. 7. Ohno T, Bodek N. Toyota production system: beyond large scale. 1st ed. Portland: Productivity; 1988.

14 8. Toyota Production System Support Center (TSSC). 2019. http://www.tssc.com. Accessed 30 Dec 2019. 9. Baldrige Website. 2018. https://www.nist.gov/bal-drige/publications/baldrige-excellence-framework/businessnonprofit. Accessed 30 Dec 2019. 10. Frankel A, Haraden C, Federico F, Lenoci-Edwards JA. Framework for safety, reliable, and effective care. White paper. Boston: Institute for Healthcare Improvement and Safe and Reliable Healthcare; 2017. p. 1–31. Available on IHI.org 11. Brilli RJ, McClead RE, Davis T, Stoverock L, Rayburn A, Berry JC. The preventable harm index: an effective motivator to facilitate the drive to zero. J Pediatr. 2010;157(4):681–3. 12. Nolan TW. Execution of strategic improvement ini-tiatives to produce system-level results. IHI innova-tion series white paper. Cambridge, MA: Institute for Healthcare Improvement; 2007. Available on IHI.org 13. Weick KE, Sutcliffe KM. Managing the unexpected: sustained performance in a complex world. 3rd ed. Hoboken: Wiley; 2015.

15© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 R. K. Shah, S. A. Godambe (eds.), Patient Safety and Quality Improvement in Healthcare, https://doi.org/10.1007/978-3-030-55829-1_2Organizational Safety Culture: The Foundation for Safety and Quality ImprovementMichael F. Gutzeit, Holly O’Brien, and Jackie E. ValentineAbbreviationsAHRQ Agency for Healthcare Research and QualityCOSS Culture of Safety SurveyDSB Daily Safety BriefingEPT Error Prevention ToolsHRO High Reliability OrganizationRTI Rounding to InfluenceSSE Serious Safety EventSSER Serious Safety Event RateM. F. Gutzeit (*) Children’s Hospital of Wisconsin, Milwaukee, WI, USAe-mail: [email protected] H. O’Brien Children’s Hospital of Wisconsin, Department of Quality and Safety, Milwaukee, WI, USAe-mail: [email protected] J. E. Valentine Seattle Children’s Hospital, Department of Patient Safety, Seattle, WA, USA2Vignette 2.1A pediatric healthcare organization had a recent change in several key executive roles.Financial performance was meeting tar-get, and much of the Board of Directors agenda was devoted to the topic of strategic efforts to maintain healthy financial perfor-mance. While there are many quality and performance improvement efforts under-way in the organization, these lack coordi-nation and have multiple consultants and teams working in silos. Despite best Chapter Objectives• To explain and define the role of patient safety culture in healthcare organiza-tional culture• To understand the essential role of lead-ership in shaping the culture of an organization• To share specific examples of safety tools, behaviors, and language used in creating a patient safety culture that connects across an entire organization• To appreciate the role safety plays as an important component of a quality improvement program• To offer methods of sustaining advances in a patient safety culture

16 Opening Question/ProblemThis chapter is about the foundational elements of building an organizational safety culture start-ing with leadership and its commitment to safety. It is intended to provide a variety of options as well as a case example that is meant to be illustrative. Recognizing that each organization has a unique set of circumstances and issues, the information should be helpful regardless of where an organization is on its own patient safety journey. Additionally, the key principles and examples could be applicable to industry outside of healthcare. IntroductionAn organization is defined by its culture. Culture influences and is influenced by the mission, vision, and values of organizations. It is the com-mon denominator that drives performance, engagement, and sustainability. It is hard to directly measure culture but it can often be per-ceived and is judged by others through the first experience with that organization as a customer or team member. Culture also reflects the value placed on the flow of information and engage-ment both up and down the organizational chain of command and input from patients/families/clients and others with perspective (“Voice of Customer”) as well as an approach to inevitable and necessary growth through change manage-ment. A specific component of organizational culture in healthcare is the safety culture, which is one component of the organizational culture. The safety culture is the sum of factors which demonstrate a resolve to health and safety man-agement by leadership to the organization [1]. Figure 2.1 depicts conceptualizing the compo-nents of a patient safety culture.Most importantly, organizational culture includes behavioral expectations that are applied consistently. New hires are made aware of this Patient safetySafetyOrganizational cultureBehavior accountabilityLeadershipFig. 2.1 The components of a patient safety cultureintentions for collaboration, there is intense competition among internal groups for lim-ited resources. There have been several attempts to initiate a defining set of univer-sal values for the organization which would help to align current and future projects and workflow changes. In the past, there had been adoption of some val-ues defined at the leadership level, but middle-management and frontline staff were confused about universal implemen-tation of these values due to a lack of a consistent educational platform and expec-tations about their relevance to everyday work at the frontline. Following a gap analysis of the organization’s current state of safety, experience, and clinical out-comes, the Board of Directors and Executive leadership agreed that building an organizational safety culture is crucial to meet the desired quality improvement vision for the enterprise to achieve staff and patient experience performance goals, as well as maintaining a healthy financial profile. The key decisions at the executive level involve how to begin the culture work, what type of existing framework to use, and how the potential impact on cul-ture will be measured. The executive team determined that starting with safety as a core value meant starting with clear and apparent leadership commitment and direction to this undertaking which would be evident to all in the organization.

17culture through the onboarding process and prior to that with a hiring process that takes into account candidate attributes consistent with the organizational culture.To illustrate this by example, at one large children’s healthcare organization (Seattle Children’s), every new hire is required to take a four hour interactive Error Prevention class on the second day of employment; those in clinical care settings take an additional four hours of Patient Safety Orientation that shares more spe-cifics on keeping patients and staff safe. The purpose is to share the priorities of safety that are universally expected throughout the organi-zation. The orientation includes methods and resources to support a patient safety goal of eliminating preventable harm to patients and staff. Topics such as Infection Prevention priori-ties, integration of simulation into learning, delivering effective and equitable patient-cen-tered care for a diverse population, and compre-hensive language and interpreter services for families are shared. Using real examples from past safety-related events helps raise situational awareness and emphasize the vulnerabilities that exist in caring for children in complex social environments.Within 90 days of hire, all leaders at supervi-sor level and above are required to attend a four hour integrated Leadership Methods course (see Appendix 2.1). This course defines the organiza-tional expectation of the leaders’ role in safety. The course uses the Institute of Medicine (now known as the National Academy of Medicine) six domains of quality as a foundation.Organizational culture supports and enables a safety culture. It is imperative to appreciate that over time, a culture of patient safety reflects the existing normative culture in any organization. If the foundation of culture is not well established, a culture of safety will be difficult to sustain. Ideally, every individual in a healthcare organiza-tion is part of the safety culture regardless of their role or proximity to patient care, because every role contributes to the health of the organization and, ultimately, the safety of all. Building a Safety Culture Begins at the TopLeadership commitment to a safety and patient safety culture is absolutely necessary because leaders shape and model culture in ways that are tangible and intangible, explicit and implicit. To change and build culture, top executives must demonstrate the behaviors they want to see. In fact, Sammer et al.’s findings from a meta- analysis [1] showed senior leadership account-ability is key to an organization-wide culture of safety and that it is the leaders that design and implement the strategy and structure that guide safety processes and outcomes and ultimately the safety culture. This point is also made in a publication by Yates et al. [2]. In an editorial on “Creating a Culture of Safety,” by Dickey from 2005 [3], it was noted that improving a culture of safety must begin with the chief executive officer. The executive leadership team must enable and build safety culture knowledge. Sammer et al.’s findings also [1] revealed that safety culture is a complex phenomenon that is sometimes not clearly understood by hospital leaders, thus making it difficult to operational-ize. To understand culture it needs to be defined. The Agency for Healthcare Research and Quality’s (AHRQ) definition is: “The safety cul-ture of an organization is the product of indi-vidual and group values, attitudes, perceptions, competencies and patterns of behaviors that determine the commitment to, and the style and proficiency of, an organization’s health and safety management” [4]. Moving from Leadership to the FrontlineIn building a sustainable safety culture, it is important for the frontline staff to understand the mission, vision, and values of the organization. This helps generate a common purpose, lan-guage, and focus. There are many contributing factors that must come together over time to con-tinue advancing the priority of a safety culture as shown in Fig. 2.2.

18Choices made and behaviors demonstrated at the executive and other leadership levels will subsequently influence those same types of choices and behaviors at all levels of the organi-zation. One important and practical demonstra-tion of a culture focused on safety (and especially patient safety) is to observe whether staff hold themselves and each other accountable by cross- checking one another and provide real-time feed-back when deviation from generally accepted performance standards is identified such as fol-lowing hand hygiene policy and best practice (i.e., 200% accountability; see Key Point Box 2.1). Very simply put, each person holds co- workers and themselves equally accountable for patient safety. A 200% accountability concept must be supported by the ability to provide open, honest, and transparent feedback without fear of retribution or retaliation following unexpected outcomes that cause harm. This includes full sup-port for families and staff involved in these inevi-table events. Moving to Improving CultureUnderstanding the current state of organizational culture is usually the most important first step in building a patient safety-focused culture. Most validated psychometric surveys are indicators of the workforce’s perceptions of safety culture and engagement for those integrated survey tools. The administration of an annual or biannual cul-ture of safety survey is most often cited as a lag-ging (trailing) indicator of cultural safety, but it could also be considered a leading metric. To elaborate on this concept, consider that if the cur-Fig. 2.2 Hospital culture of patient safety contributing factors. (Reprinted from Sammer et al. [1], with permission from John Wiley and Sons)Key Point Box 2.1: 200% AccountabilityAn organizational expectation that each person is 100% responsible for following behavioral and best practice norms as well as holding others 100% accountable for the same

19rent culture norms continue as is in an organiza-tion, it will be predictive of the future organization cultural direction unless there is a change. If there is desire or a restlessness that improvement is needed in the current organizational culture, the results of such safety surveys should give an idea of how high the bar needs to be set to affect cul-ture change when planning for improvements.Some of the validated psychometric culture of safety surveys (COSS) (see Key Point Box 2.2) organizations use today to monitor workforce per-ceptions of culture are the Agency for Healthcare research and Quality (AHRQ), the Safety Assessment Questionnaire (SAQ), the Safety, Communication, Organizational Reliability and Engagement (SCORE), the Advisory Board and the Press Ganey Integrated Engagement, Resilience and Safety Culture Survey. These sur-veys will identify workforce perception of the safety culture at a point in time. Achieving at least a 60% survey response rate from staff gives the most meaningful results which can be analyzed and potentially acted upon. Ensuring anonymity is also crucial for participation and candid responses.The Joint Commission (TJC) requires, and other regulators recommend an assessment of the safety culture at a minimum of every 2 years with a validated survey. This is also required to receive top recognition on the Leapfrog Hospital Survey [5]. With a focus on leadership and culture, regu-latory agencies are looking for survey results shared from the board to the frontline teams with clear action plans and a continuous history of improvement. A Joint Commission Sentinel Event Alert, published in December 2018, noted the importance of leadership accountability to advance a strong safety culture and frontline team member’s willingness to report both near misses and patient safety events that reach the patient [6].To emphasize the importance of these safety survey tools and their potential use, consider an organization that received a sub-optimal Leapfrog Hospital Survey score. Further analysis revealed one of the most heavily weighted questions impacting the score was related to the administra-tion and organizational action planning from the culture of safety survey (COSS) results. The COSS had not been administered for over two years, and previous surveys lacked a clearly demonstrated organizational dissemination of results, communication to staff, and leadership oversight and follow-up on actions with the teams involved in the actions. The important lesson is that any survey must be linked to follow up com-munication and sharing of results along with a clear plan and timeline with support for action. This is important in building the desired culture in an organization and was demonstrated in six large hospitals researched by Campione and Famolaro [7].An effective strategy to achieving desired input leading to action could be to debrief the survey results with a team through an indepen-dent facilitator which might occur without the leader of the unit present. This encourages candid and comprehensive feedback and engagement of staff. Closing the loop on such discussions with staff and leaders is important as is celebrating successes and measurable improvements. As much as possible, a supportive, non-punitive, and actionable organizational response to low perfor-mance score is imperative. Open-ended com-ments from surveys can also provide additional insight if they reflect a systemic issue. Actions must be prioritized and using data whenever pos-sible helps sustain the effort.Key Point Box 2.2: Culture of Safety Survey (COSS)A survey to gain insight about how staff and others involved in the delivery of care organizational operations view the current patient safety practicesVignette 2.2Leaders had reviewed the annual culture of safety survey results, but did not have a real understanding of meaningful actions to take on for improvement nor any owner-ship or accountability from the leadership team on expectations. Safety appeared to be a lower priority overall in the organization without clear expectations

20 Leaders Being Present and Leading by ExampleThe following are ways that leaders can opera-tionalize and visibly demonstrate a commitment to building and maintaining a commitment to safety and reliability [8]. 1. Daily Safety Briefings (DSB) are recognized as best practice to achieve an enterprise-wide daily operational surveillance and manage-ment system to enhance the awareness and priority of safety. The Daily Safety Briefing starts at the local level with team or unit hud-dles throughout the organization. This struc-ture allows reports of safety and operational concerns to be communicated and resolved as quickly as possible. A system-wide, daily 15-minute huddle facilitated by an executive with reports from key operational leaders for escalation of high-risk issues, deficiencies, distractions, cross-departmental issues, and abnormal conditions, allow the leadership teams to become more sensitive to operations, and the immediate needs of patients, staff, and facilities are addressed. The timing for the DSB should be consistent from day to day with minimization of scheduled meetings dur-ing this interval to allow maximal participa-tion. Utilizing a modified weekend and holiday structure shows continuity of leader-ship support. Key factors for a successful and sustainable Daily Safety Briefing include leadership presence at the huddle and on the call, preparation by reporting teams, defined follow-up on concerns raised to build trust in the process as well as clear expectation and accountability to participate in the Daily Safety Briefings. 2. Rounding with purpose on individual units. In addition to being visible for operational lead-ers at the Daily Safety Briefings, executives should set the expectation for leaders at all levels to participate in mandatory rounding on a regular basis. The purpose is to connect with the frontline leaders, teams, and patients/fam-ilies to observe firsthand the work being done. This will allow a determination of work as it is actually being done compared to how it is imagined being done: reality vs. perception. Rounding promotes an opportunity for leaders to provide a few key strategic and tactical sys-tem items to staff and solicit their feedback on goals, priorities, concerns, and barriers. This effort supports building relationships and to close the loop on issues raised from previous rounding interactions or the Daily Safety Briefings. It gives leaders the opportunity to provide positive feedback and to recognize and reward those individuals who demonstrate the safety culture behaviors and language. Many different types of rounding methods are evolving across healthcare systems and are beneficial for building staff engagement, patient/family, experience, and culture of safety scores across all domains. Examples of and guidance on relevant and sustained action plans for performance improve-ments. With a baseline cultural assessment complete, a strategy was set and operation-alized that created clear standards and expectations. A curriculum was initiated for every leader and frontline workforce member in the organization to build capa-bility and capacity in safety culture behav-iors, terminology, and habits to reduce the probability of error. An aspirational goal of zero harm, like many other healthcare insti-tutions, was set. The board and senior exec-utives recognized it would require a large upfront commitment of time and resource allocation. Borrowing from examples in non- healthcare industries provided aware-ness that crucial elements of developing consistent system reliability and culture would be essential to achieve similar results. The executive team agreed to keep a visible commitment to safety as a top ongoing priority.

21rounding practices for leaders include the following: (a) Round with every patient every day: Operational leaders round on every patient every day with a focus on one or two important questions that could be related to improvement ideas from the patient experience or culture of safety sur-veys such as teamwork within and across teams and feeling safe to speak up and escalate an issue [9]. (b) Round once a month with every staff member – Rather than waiting for the annual engagement or safety surveys, organizations can implement a continu-ous feedback model in which each staff member has an opportunity to speak with their leader to share ideas and concerns or show appreciation and receive interval updates on goals and developments. (c) Executive Walk-Rounds – A method to coach and focus on key organizational goals. This is a way to validate that front-line teams understand the importance of specific priorities such as hospital acquired conditions (HACs), hand hygiene and other work important to improving patient outcomes, such as care bundle reliability. [10, 11] (Key Point Box 2.3).Whichever rounding method(s) is imple-mented, it is important to start with intention and purpose and build confidence, capability, and capacity in all leaders to round and close the loop on issues raised whenever, and as soon as possi-ble. Rounding times could be used to emphasize a specific organizational value or for recognition where individuals and systems have performed well. (Key Point Box 2.4).Consistent messaging across leadership levels will demonstrate the cultural priorities of the organization at the system level down to the unit or department level. Organizational Case Example: Embedding Safety Culture Tools, Behaviors, and LanguageThe following is an actual case of how Seattle Children’s (formerly Seattle Children’s Hospital) used a structured process to embed safety tools and behaviors to drive their safety culture at the frontline with leadership support. A consultant in high reliability organization was utilized to col-laborate on this journey. At the outset a standard-ized Safety Event Classification (SEC) taxonomy and algorithm system was used to classify reported safety events from a previous 12-month period. This is a method of defining and investi-gating thoroughly near miss events (NME), pre-cursor safety events (PSE), and serious safety events (SSE) to determine a baseline Serious Safety Event Rate (SSER). A serious safety event is defined as an unintended incident that reaches the patient causing moderate to severe harm, including death. In a serious safety event, clear deviations from generally accepted practices or standards have occurred, such as unknowingly going against policy due to lack of training or dis-tractions. An event classified as an SSE is gener-ally considered preventable. The Serious Safety Event Rate is calculated monthly as the number of serious safety events for the previous Key Point Box 2.3: Care BundlesEvidence-based practices that when per-formed collectively with high reliability have been demonstrated to improve patient careKey Point Box 2.4: Communication of Shared Learnings• Thematic leadership rounding such as a specific organizational value, error pre-vention tool, or regulatory concept• “Close the loop” follow-up on issues• Intranet posting of recognition for spe-cific examples of excellence in safety• Periodically starting the Daily Safety Brief with an example of a “good catch”

2212 months per 10,000 adjusted patient days [12]. The ultimate goal is zero serious safety events which is commonly used within an organization as one metric to determine the improvement in patient safety culture, systems reliability, and overall performance improvement. The transpar-ency of sharing safety event stories and meaning-ful safety data as learning opportunities had a significant and positive influence in improving Seattle Children’s organizational culture and reli-ability. It helped reinforce that everyone in the organization, no matter the role, contributed to improving the SSER.Vignette 2.3Seattle Children’s chose to invest a sig-nificant amount of resources to train all leaders using a leadership curriculum with dedicated weekly effort called Rounding to Influence (RTI) which set expectations on reinforcing and coaching to the safety culture journey (see Appendix 2.1). All frontline clinical and non-clinical leaders and workforce members were trained in error prevention tools (EPT), behaviors, and a cultural language (see Appendix 2.2), which in theory should reduce safety events [13]. The tools focus on reducing the probability of errors by enhancing communication, such as using standard structured formats for handoffs, repeating, and reading back information to ensure the receiver has the correct information or task. Specific tools and a brief explanation are described in Appendix 2.2. Frontline teams and leaders learned and applied these error prevention tools in both clini-cal and non-clinical settings. The purpose is to create a unified set of safety behav-iors and common organizational language that can help eliminate defects and errors as seen in other high reliability organiza-tions and peer institutions by building habitual excellence in the use of this lan-guage and behaviors. Progress toward a Key Point Box 2.5 5 × 5 RoundingA monthly observational and coaching rounding tool where a leader asks five staff member from five different disciplines about the penetration of culture tactics to influence behaviors with the knowledge and application of error prevention tools and behaviors at the frontlinesafety culture was accomplished at Seattle Children’s for the training sessions with built-in sustainability structures to ensure all new hires and leaders are on- boarded to the culture training. The Daily Safety Briefing was operational 7 days per week and recognized as a best practice during the consultant quarterly assessment. The reporting of patient safety events, both near misses and events that reached the patients, had almost doubled. Most impor-tantly, the overall outcome metric, the Serious Safety Event Rate, was steadily decreasing. However, when the consultant came to do a quarterly assessment on the safety culture strategic initiatives, the patient safety and executive teams were disappointed to hear the results of round-ing observations on the units. The consul-tant used a technique called 5 × 5 Rounding (Key Point Box 2.5) where five individuals from different disciplines and different areas were asked about their cur-rent understanding and application of the error prevention tools and safety behaviors.The results demonstrated most staff members recall an error prevention tool only 20% of the time (i.e., could share one or two safety tools out of 7 (Appendix 2.2)). The consultant felt the behaviors and lan-guage, although taught in the classroom had not penetrated to the frontline culture, as would have been expected at this point in the journey.

23 5 × 5 Rounding for Observational Assessment, Coaching, and ImprovementFollowing this assessment, the patient safety team, with support from leadership, instituted an observational 5 × 5 rounding/coaching assess-ment methodology on a monthly basis to observe and coach both the leaders on their Rounding to Influence and the workforce on the importance and use of the error prevention tools and safety behaviors. A specific and time-bound goal was set and measured. The patient safety team devel-oped a 5 × 5 Rounding script (Table 2.1) and used it to ask workforce members which error preven-tion tools they were familiar with or which ones resonated most with them. They asked for an example of an error prevention tool along with a situation demonstrating its proper application. Coaching occurred in the moment to reinforce learning.While rounding, it was clear that the highest performing areas were those where the local leader influenced the outcomes due to their per-sonal investment with understanding and utiliza-tion of culture principles.These observations and results helped guide the Rounding to Influence monthly focus in order for leaders to coach their frontline teams on the application of the error prevention tool. This, in turn, demonstrated the organizational importance and priority of a strategic initiative to eliminate preventable harm with these tools and behaviors. There was a direct correlation to the improvements seen in the overall results each month with the current Rounding to Influence message (Fig. 2.3).Eventually all directors were trained in 5 × 5 rounding so they could do their own assessment and coach the leaders in their reporting structure to further enhance knowledge and application of the tools and language at the frontline. For inter- rater reliability, the patient safety department continued to round on a monthly basis to track process and outcome metrics at a global level. Every month, leaders reported out their teams error prevention tool recall rate to senior execu-tives which demonstrated accountability to the safety culture work and determination to elimi-nate preventable harm. The hypothesis is that by developing habitual use of safety tools and asso-ciated behaviors, a reduction in the serious safety event rate would result (Fig. 2.4).Table 2.1 5 × 5 Observational coaching form used by patient safety and directors to assess penetration of error prevention tools and behaviors across roles at the frontline5 × 5 Rounding response capture toolUnit/departmentShift□ Day (7 am–7 pm)□ Night (7 pm–7 am)Role(leader, provider, clinical/support, non-clinical)Title(attending, RN, respiratory therapist, unit coordinator)Completed EPT class□ >6 months ago□ 3–6 months ago□ < 3 months ago□ Not yet completedEPT tools recalled□ ARCC□ QVV□ Read/repeat back□ SBAR□ Standardize H/O’s□ STAR□ Stop and resolveCan you give an example of using tool?□ Yes, correct use□ No, incorrect use□ No, no exampleCan you give an example of someone else using tool?□ Yes, correct use□ No, incorrect use□ No, no exampleSafety stories shared at meetings and huddles?□ Yes□ NoDo you know who your safety coaches are in your area?□ Yes□ NoTable 2.1 (continued)5 × 5 Rounding response capture toolUnit/departmentReceptivity□ Very receptive□ Somewhat receptive□ Neither□ Slightly non-receptive□ Very non-receptiveCommentsAdapted with permission from Healthcare Performance Improvement (HPI). Copyright Seattle Children’s 2014

24With the assurance that the proper foundation had been set up with leaders now having safety culture knowledge, a strategy had been identified and communicated, and key tactics to improve culture and reliability had been operationalized, the organization decided to pilot the integrated Safety Communication Organizational Reliability and Engagement (SCORE) safety culture survey. One valuable result of conducting this pilot was demonstrating the impact of leader rounding with frontline teams and closing the loop on issues raised (see Fig. 2.5). The survey was adminis-tered in four key areas of the organization, with a 85% response rate (n = 849). For the teams that could answer “yes” to the question “my leader rounds regularly” and also “yes” to the question “they closed the loop on concerns raised,” there was a positive correlations to all culture domains. Fig. 2.3 5 × 5 observational coaching tool results built in a tableau report. Ability to drill down by area and role. (Adapted with permission from Healthcare Performance Improvement (HPI). Copyright Seattle Children’s 2014)

25Fig. 2.4 This juxtapose graph demonstrates as the percentage of Error Prevention Tool Recall Rate (EPTRR) increases (blue line), the Serious Safety Event Rate (SSER) decreases (orange line) over time, Jan 2015 to Oct 2015. (Adapted with permission from Healthcare Performance Improvement (HPI). Copyright Seattle Children’s 2015)WalkRound (WR) feedback effect on cultureLearning environment with WR feedbackLearning environment without WR feedbackWork/Life balance with WR feedbackWork/Life balance without WR feedbackSafety climate with WR feedbackSafety climate without WR feedbackTeamwork with WR feedbackTeamwork without WR feedbackPersonal burnout with WR feedbackPersonal burnout without WR feedbackBurnout climate with WR feedbackBurnout climate without WR feedbackLocal leadership with WR feedbackLocal leadership without WR feedbackSource Date: Jun 2017Institution: Seattle ChildrensWork Setting: All Pilot Work AreasPosition(s:) All PositionsResponse Rate: 85% (N=849)0204060801004172364652693556437963755162Fig. 2.5 SCORE pilot leadership rounding effects on culture domains. (Image courtesy of Safe & Reliable Healthcare, LLC Copyright 2017. Data shared with permissions of Seattle Children’s)In particular, the culture domains of Learning Environment, Safety Climate, Teamwork, and Local Leadership showed a >30% improvement in results. Each of these domains represents cru-cial elements of a positive organizational safety culture. These results mimic previous studies demonstrating the influence of leadership round-ing on quality outcomes, workforce well-being/

26resilience and ultimately the improvement in an overall organizational patient safety culture to eliminate preventable harm to patients [14, 15].The greatest result change was demon-strated with the Agency for Healthcare Research and Quality culture of safety ques-tion: “the grade I give my unit for safety.” This improved over a full point on the Likert Scale (1–5) going from a 3.01 to a 4.02 and was sus-tained for two consecutive years; a change of this order of magnitude is considered signifi-cant (Fig. 2.6).The organization also celebrated a full year without a serious safety event and the serious safety event rate dropped by 80% (Fig. 2.7). The reduction in the serious safety event rate has overall continued to decline since the strat-egies with standards and expectations were operationalized.“Please give your workarea/unit in this hospitalan overall grade onpatient safety”Grading Patient SafetySurvey resultsFY12 = 3.00FY13 = 3.07FY14 = 3.07FY16 = 4.05FY15 = 4.03*FY17 = N/A*AHRQ Culture of Safety Survey provider engagement not conducted is 2017 Fig. 2.6 Three years into Seattle Children Hospital’s safety culture journey a full point increase was seen and sustained in this AHRQ Culture of Safety Survey Question. Likert Scale 1–5 (5 is best). (Data used with permission Seattle Children’s. Copyright 2017)Serious safety events per 10,000 adjusted patient daysFig. 2.7 Serious Safety Event Rate over time (lower is better), August 2011 to January 2019. (Used with permission from Healthcare Performance Improvement (HPI). Copyright 2019. Data used with permission of Seattle Children’s)Vignette 2.4The executive team was pleased to see the results of the efforts paying off in the multi-ple initial tactics to improve safety and drive the serious safety event rate down. They could also see the positive effects the leader-ship influence had on culture. An increase in safety behaviors and use of safety tools being supported by leaders resulted in a significant reduction of the serious safety event rate.

27These vignettes and case study have shown the power of leadership and culture on safety and an unmeasurable return on investment in relation to the ultimate goal of eliminating preventable harm. Figure 2.8 shows the exponential reduction in risk of harm when there is a focus on culture and safety as a core value.In Fig. 2.8 the vertical (Y) axis shows the risk of harm with HIGHER meaning less chance of harm. The lowest line in the graph shows improv-ing systems and processes, reduces the probabil-ity of error from 1 in 10 to 1 in 100, and is an important first step. The middle line denotes the exponential improvement in reducing the proba-bility of error with introducing the culture work. The highest line is achieved through efforts incorporating human factors consideration. For example, anesthetic gasses each have universal and unique connection shapes and colors. This makes it physically impossible to inadvertently connect an incorrect gas on the anesthesia machine. The healthcare environment is complex and always evolving. When possible, reducing variation and minimizing human error improve safety as demonstrated by the airline and nuclear regulatory industries.All safety is quality but not all quality is safety. Utilizing a performance improvement framework that focuses on improvement science will strengthen safety efforts and safety culture by anchoring these efforts into a comprehensive plan. Any one of many models (Six Sigma, Lean, Model for Improvement, etc.) may be helpful in changing organizational culture. Many health-care systems use the six domains of quality as a foundation to frame their safety and quality pro-grams. The six domains were developed by the Institute of Medicine in 2001 [16]. It is important Journey to improving reliabilityOptimizedoutcomesIntuitive designImpossible to do the wrong thingObvious to do the right thingBehavoir expectations for allFair, Just and 200% accountabilityCore values & vertical intergrationHire for fitEvidence-based best practiceFocus & simplifyTactical improvements (e.g. bundles)© 2020 Healthcare Performance Improvement. LLC. ALL RIGHTS RESERVEDHuman factorsintegrationReliabilitycultureProcess designLevel 5quality10-110-210-310-410-510-610-710-8ReliabilityFig. 2.8 The exponential effects of culture on reducing the probability of harm over time. (Used with permission from Healthcare Performance Improvement (HPI). Copyright 2010 Healthcare Performance Improvement)Vignette 2.5The healthcare system’s senior executives have been observing the evolving culture of the organization. They are realizing that while there is enthusiasm to work on qual-ity improvement, there is a need to clarify what that actually means. Therefore, the CEO has identified a small group to iden-tify one model of quality improvement that can then be shared throughout the system.The leadership is in full agreement. A broader team of leaders throughout the sys-tem has been assigned the task of conduct-ing an in-depth analysis of safety risks with the goal of reducing harm through targeted efforts and projects. This team is reviewing several models to determine the best option for a quality system and is tasked with tak-ing on the bigger responsibility of overall quality and performance improvement.

28to diversify the organizations quality improve-ment work across all the domains. Balance of the domains is essential for inclusiveness in decision- making and to appropriately distribute the effort. For example, if safety is the only focus, then other domains (efficiency, timeliness) might be out of balance and productivity, or workflow might be significantly negatively affected. A clear and practical example of this involves the implementation of an enterprise electronic health record (EHR) at one of the organizations affiliated with authors of this chapter (MG, HO). Knowing that medication errors are a common source of patient harm, the organization invested signifi-cant effort into prevention by including a variety of best practice alerts and similar “pop- ups” for medication ordering and dispensing prior to implementation. This led to the unintended con-sequence after “go-live” of burdening the end user with significant workflow challenges and delays due to the volume and frequency of these alerts. The domain of safety was overemphasized at the expense of the domains of timeliness and efficiency.Spreading initiatives and resources across the domains insures having a broad portfolio of qual-ity improvement projects. The organization should have a continuous improvement approach to quality. Improvement never ends. Change is inevitable. The balancing process is in prioritiza-tion and calibration of projects to system and workforce capacity. Feedback such as “There are too many projects,” “People can’t absorb all the changes,” and “Things are moving too fast” is legitimate. The leaders of the organization must be aware, support the changes, and help the front-line understand the desired key outcomes in an ever-changing environment. To do so, it requires a consistently applicable and understandable change management process (described below).A typical organizational portfolio will neces-sarily include initiatives such as reduction of hos-pital acquired conditions, improving patient/family experience work, staffing model projects, and financial initiatives. When the quality improvement model is defined, the work of man-aging change needs to be addressed as well. A surveillance system should be in place to deter-mine when too much change may be overwhelm-ing people and the system. This could include feedback from rounds, employee engagement surveys, focus groups, or unexplained delays in project timelines.Change is inevitable. Growth is necessary. Project teams must have change management as part of the plan while implementing projects. The leaders of the organization must be aware, sup-port the changes, be visible during the change, and help the frontline understand the desired key outcomes in an ever-changing environment.Change disturbs the current state. Anyone involved in an organizational change seeks to understand how the change will affect them, their role, their workflow, and their team.There are several change management models that can assist organizations. All have processes to guide teams through change. By using a con-sistent framework for change, leadership can demonstrate a commitment to providing the appropriate resources and processes for it.Another approach to achieve safety improve-ment used by many healthcare systems in the past few years is the high reliability organization (HRO) model. Though not specific to healthcare, many health systems find it useful in improving their cul-Vignette 2.6The leaders of the healthcare system are starting to feel more comfortable discuss-ing quality improvement methodology, have a clear plan to prioritize projects, and have developed a consistent process for rounding with staff. While rounding, they receive frequent feedback that there are a lot of changes and that the leaders seem to be missing an opportunity to communicate and discuss the most important changes affecting staff. Staff are reporting they don’t always have time to check emails and are requesting other forms of communica-tion to help them keep up on relevant changes.

29ture. Dr. Karl E. Weick and Dr. Kathleen M. Sutcliffe have published texts about this model which are often cited as excellent reference points [17]. They studied industries such as nuclear power and flight operatins on aircraft carriers which were able to significantly decrease serious safety events. By focusing on the principles of anticipation and containment of accidents, an HRO can ultimately reduce serious safety and harm events. Building an HRO is predicated on improvement through learn-ing. In other words, an organization must be will-ing to explore risks, learn from mistakes, and have the underpinning of a culture that supports identi-fying, analyzing, and mitigating errors and poten-tial errors. The foundational elements of leadership methods, detecting and bringing problems to awareness, error prevention tools, and focus on a culture of patient safety across the organization lead to high reliability (Fig. 2.9).The language used in the five principles noted in Fig. 2.9 can sometimes be confusing. The fol-lowing explanations are to help clarify.The first three are principles of anticipation for unexpected events; the final two are principles of containment when the system is impacted by an inevitable, disruptive significant event. 1. Reluctance to simplify: This is an appreciation that work in healthcare is complex and dynamic; thus people in this environment seek underlying, deeper causes when thing succeed or fail rather than superficial explanations. Consider an administered drug error in a patient. While it might be straightforward to identify the individual at the terminal point in the process, it is highly likely there are under-lying systems issues that led to the failure mode (Key Point Box 2.6). 2. Preoccupation with failure: Healthcare relies on human delivery of processes. Human sys-tems are inherently imperfect. Therefore, everyone is aware of and thinking about the potential for failure. Near miss events are seen as opportunities to learn about systems issues and improvement. Lack of significant events should lead to heightened vigilance rather than complacency. Preoccupation with failure is practiced when the organization focuses on the small close calls and near misses as fail-ures, studies how those failures occurred, and implements change to prevent the errors from reoccurring. Ideally, the organizational cul-ture and leadership should be encouraging staff to report error, seek out the failures, and affirmatively recognize staff for identifying the errors. 3. Sensitivity to operations: This is vital to a strong culture. This sensitivity is sometimes referred to as situational awareness. It is a constant awareness of how the current state supports or might jeopardize safety. Leaders should be rounding with the frontline staff in a standardized fashion and, most importantly, with a process to close the loop on issues. The frontline staff very often know what the safety concerns are and can provide input about potential resolutions. Leaders must be aware of the differences between the work as they imagine it is being done in distinction and how the work as it is actually done at the frontline. Workarounds or alterations to the Preoccupationwith failureReluctance tosimplifySensitivity tooperationsDeference toexpertiseCommitmentto resiliencePrinciples ofhigh reliabilityFig. 2.9 The principles of high reliability in organiza-tions. (Adapted with permission from Weick and Sutcliffe [17])Key Point Box 2.6: Failure ModeA cause of failure or one possible way a system may fail

30designed process might represent a sign of potential failure points. 4. Deference to expertise: People in HROs appreciate that those closest to the work are the most knowledgeable about the work. In a critical situation or emergency, the person with greatest knowledge of the issue might not be the person with the most seniority. In an HRO, everyone is expected to share con-cerns with others, and the organizational cli-mate is such that all staff members are comfortable speaking out about potential safety problems. The people that do the work must also be included in improvement proj-ects because they are the true subject matter experts to identify what can be successful in the system that supports their work. This prin-ciple leads to optimal organizational learning. 5. Commitment to resilience: High reliability organizations assume there is an ever present risk for failure, and they simulate rapid assess-ments of, and responses to challenging situa-tions. This also supports trusting, supportive relationships. Teams will be comfortable speaking up in unsafe situations because it is recognized that return to normal operations in times of duress relies on collaboration. Through leadership’s continuous commitment for appropriate resources, the staff will feel supported, competent, and empowered to lead and create change.By using all five of these principles of high reliability, the organization’s culture of safety will improve. But becoming a high reliability organization is a never ending process with a tire-less commitment to understanding where the next improvement can reduce risk.The organizational values have been estab-lished, but clear rules around behavior expecta-tions may not have been explicit. Rules regarding the behaviors within a culture are just as impor-tant as the values themselves.How do behaviors affect culture? Demonstrating and following expected behaviors by all staff and leaders is the real test of how effective the culture has been. At some point there will be an organizational challenge which will test the resolve of leadership to follow the culture path. One important consideration when building a successful and sustainable safety cul-ture is through developing appropriate account-ability known as a just culture [18, 19] (Key Point Box 2.7). If staff have a sense the culture allows retribution, retaliation, and other punitive mea-sures, it will stifle event reporting.Vignette 2.7While there have been many gains, the leaders still have strong concerns about some of the behaviors they are seeing. They are hearing reports of bullying, incivility, and disruptive behaviors. They are finding out that teams may be working well in their individual microsystem but they are not working well with each other across the system.Key Point Box 2.7 Just CultureBalanced accountability for organizational systems and the behaviors of individuals within those systemsVignette 2.8As the organization became more compe-tent and knowledgeable about the princi-ples and science of safety performance improvement and culture building, inevita-ble requests for additional resources, staff, and learning opportunities grew signifi-cantly. Leaders as well as board members were encouraged by the direction of change but were also wary of commitments to such resources without some validation of an acceptable return on investment.

31There is evidence that reasonable safety investments are consistent with financial health for organizations. There are data emerging which quantifies the financial outcomes of investment in safety culture [20]. One network that has been tracking this concept is the Children’s Hospitals Solutions for Patient Safety network. The net-work consists of >140 children’s hospitals work-ing together to reduce harm to patients. Their website references the aggregate cost savings from 2012 to 2018 (>$182 M) to these member organizations gained by harm avoidance result-ing from this effort over the past 6 years as well as the number of patients spared harm (>11,000) [21].It should be recognized that while a culture of safety is a top priority for organizations, it cannot be the ONLY priority. The financial commitment to safety must be included in the overall process of budgeting, sequencing, and change manage-ment capacity which is unique to every organization. Sustainability of a Safety CultureRecognizing the dynamic elements in play in all organizations, perhaps the most important ques-tion is how healthcare organizations maintain and sustain a culture of safety.As has been mentioned throughout this chap-ter, healthcare is changing at an unprecedented rate with many real and potential disruptors such as non-traditional organizations engaging in the field consolidation and expansion of organiza-tions and technology [22, 23]. These changes coupled with a consumer-driven focus that threat-ens to make healthcare more transactional can all lead to distractions and drift in organizational focus.Despite this, healthcare organizations must remain committed to their first mission and core principle: safe and effective care delivery. It is the expectation of staff, providers, patients, and families.There are a number of properties which have been identified to build and sustain a culture of safety. Any such list is open to healthy debate. We would like to share three that we feel should top any list of an organization hoping to sustain the hard won efforts to build a safety culture.First is the commitment of leadership (and the governing board) to prioritizing the vision of safety and elevating it to the highest level of orga-nizational attention which is often occupied by finance and strategy. Safety does not replace either one of these but is complementary to the long-term health of the organization. Leadership through actions, visibility, and modeling the behavior expectations is constantly reinforcing the safety culture.Next is developing the necessary components to build a high reliability focus that is based on having a process to anticipate, detect, analyze, and act upon inputs and events. This requires building the infrastructure components to ade-quately support these processes.Finally there must be a culture with account-ability, behavior expectations, and a way to develop leadership in these areas to weather the inevitable reality of organization change and turnover. These characteristics connect the board, leadership, staff, providers, learners, and patients/families.Leadership and the culture of an organization must be focused always on learning and improv-ing. The culture drumbeat must be relentless – culture never “coasts” or takes a break.Editors’ CommentsIt is well-known saying from business guru, Peter Drucker, that “Culture eats strategy for breakfast.” There should be no question as to why this textbook using practical, case-based approaches to safety and quality improvement in healthcare should use begin with the foundation of all improvement work: culture. The authors use a vignette to trace the accountability of culture. It should go without saying that culture starts at the top with an explicit organizational commitment to the culture of the organization. The core values described at one our institutions in Chap. 1

32 Appendices Appendix 2.1: Examples of Leadership Methods Curriculum for High Reliability• Safety as a core value – Link all decisions explicitly to safety and begin every meeting with safety story.• Leaders Round to Influence – Helps sets the tone and reinforces the behavioral expecta-tions for safety and reliability (Fig. 2.10).• 5:1 Feedback – Intentional focus on observing Fig. 2.10 4 C’s Rounding to Influence rounding template to highlight the importance of a safety culture focus. (Used with permission from Healthcare Performance Improvement (HPI). Copyright 2008 Healthcare Performance Improvement)is an example of how culture starts from the top. Culture then needs to be implemented and operationalized at the interface between management and the frontline. The authors develop the vignette to show strategies on how to ensure culture perme-ates these levels.A key image in this chapter demon-strates the exponential reduction in risk of harm when employing high-reliability cul-ture techniques. As we all anxiously wait for human factors integration, we must rely on culture to take us to the next levels of reliability. A theme throughout this text-book will be high-reliability principles as described by Weick and Sutcliffe in manag-ing the unexpected. Throughout the text, we will see how these principles relate to the specific topic of the chapter. This may appear redundant, but it is intentional. As editors we want the reader to be left with a thorough understanding of how high- reliability principles must be weaved into operations, management, and leadership to drive culture and thus outcomes. The authors end the chapter discussing one of the most pressing issues in safety and qual-ity improvement in healthcare – sustaining the gains. Safety of culture surveys are dis-cussed as an objective way to measure and track culture; however the urgent need to ensure our safety and quality gains is solid-ified, and hardwire cannot be underesti-mated. Culture is core to safety and improvement as thoroughly described in this chapter.

33and calling out five positive workforce behav-iors for one corrective feedback.• Performance Accountability – A fair and just culture promotes an environment of trust, fos-ters self-reporting of error, and increases staff engagement. Recognizing honest mistakes from knowing violations.• Daily Safety Brief.• Unit Safety Huddles. Appendix 2.2: Examples of Error Prevention Tools I. STAR- Stop Think Act Review, a self-check on every task, “Am I getting the result I expected with my action?” (labeling a specimen, paying a vendor, sending an email). II. Read back/repeat back and asking clarify-ing questions. III. Stop and resolve – not proceeding in the face of uncertainty and taking the time to go and ask for help. IV. QVV – Qualify, Validate, and Verify with a trusted source when something does not feel right or information received is not what is expected (HPI © 2006). V. SBAR – Situation, Background, Assessment and Request/Recommendation. Used in both written and verbal communication to relay key information that requires an action from another individual. VI. ARCC – Used to escalate concerns in a non-judgmental way by Asking a question. If that does not change course, the next step is to Request a change. If the issue is still not addressed or changed, then then the next step is to raise a Concern and ulti-mately the last step is to go up the Chain of Command (HPI © 2006). VII. This is similar to AHRQ TeamSTEPPS and CUS (I have a Concern, I am Uncomfortable and this is a Safety issue). VIII. Error prevention tool(s) and behaviors – Come in a variety of frameworks to enhance communication, safety and pro-cess reliability (Figs. 2.11 and 2.12).Safety behaviorsError prevention toolsI commit to:I commit to:I commit to:ARCC:STAR:Stop and resolve when uncertainQVV: Quality, validate, verifyThink through the detailsSpeak up if I have a concernReport issues promptlyPromote a questioning environmentEncourage an open, fair and just cultureAcknowledge safe behaviors when I seethemCommunicate clearly and accuratelyQuestion and coach othersParticipate in resolving issuesAsk question; Request a change;voice a Concern; and Chain of CommandStop: Pause for a moment; Think: Focus on the act; Act: Perform the acr;Review: Check for desired resultsQualify: Do I trust the source;Validate: Dose it make sense to me?;Verify: Check with an expert sourceRepeat-backs and read-backs using:SBAR:Standardize handoffs (5Ps)One or two clarifying questionsPhonetic and numeric clariificationsSituation, Background, Assessment, RecommendationPatient (or Project); Plan; Purpose of plan;Problems; PrecautionsWe are allaccountable forpreventing errorsWe respect others’input and opinionsregardsing safetyWe work as oneteam to create asafe and reliableenvironment•••••••••••Fig. 2.11 Safety behaviors and error prevention tool integrated to customer service standards for safety and reliability. (Adapted with permission from Healthcare Performance Improvement (HPI). Copyright Seattle Children’s 2012)

34 Chapter Review Questions 1. Organizational culture is the same as a culture of patient safety. True or False?Answer: False. 2. Describe the building blocks of a patient safety culture.Answer: Leadership, Behavior and Accountability, Organizational Culture, Safety – all are foundational to a culture of patient safety. 3. Organizational culture is independent of lead-ership changes.Answer: False. Leadership sets the course for culture and significant changes in leader-ship have an effect in some way on organiza-tional culture. 4. Error prevention tools should be learned by a few key individuals in an organization to avoid confusion.Answer: False. It is essential that everyone in the organization is responsible for safety and all contribute in some way to patient safety. 5. Identify the six domains of qualityAnswer: The acronym STEEEP: Safe, Timely, Effective, Efficient, Equitable and Patient (Family) Centered. 6. High reliability organizations exist in many industries, including healthcare.Answer: True. Healthcare has learned much from other industries such as airlines about HROs. 7. What is a just culture?Answer: Balanced (appropriate) account-ability for organizational systems and the behaviors of individuals within those systems. It is a way to develop a non-punitive culture. 8. Can investing in patient safety lead to finan-cial benefits to an organization? Why or why not?Answer: Yes. There is growing evidence that focusing on patient safety can improve patient safety and have a beneficial effect on organizational finances.References 1. Sammer CE, Lykens K, Singh KP, Mains D, Lackan NA. What is patient safety culture? A review of the literature. J Nurs Scholarsh. 2010;42(2):156–65. 2. Yates GR, Bernd DL, Sayles SM, Stockmeier CA, Burke G, Merti GE. Building and sustaining a sys-tem wide culture of safety. J Qual Patient Saf. 2005;31(12):684–9. 3. Dickey NW. Creating a culture of safety. J Patient Saf. 2005;1(2):75.Fig. 2.12 High level overview of error prevention tools and behaviors summarizing objectives of use and purpose. (Adapted with permission from Healthcare Performance Improvement (HPI). Copyright Seattle Children’s 2012)Self-check and cross-check•Stop, think, act, review (STAR)•Ask a question, request a change, voice a concern, go up chain of command. (ARCC)Ask a question/ask for and offer help•Stop and resolve – Do not proceed in the face of uncertainty •Qualify, validate, verify (QVV) – Use an expert source when in doubt Communicate clearly•Repeat-back read-back ask clarifying questions•Situation, background, assessment, recommendation (SBAR)•Handoff standards for patients and projects

35 4. Agency for Healthcare Research and Quality (AHRQ). Hospital survey on patient safety culture; 2002. http://www.ahrq.gov/qual/patientsafetyculture/hospsurvindex.htm 5. Leapfrog Hospital Survey. http://www.leapfroggroup.org/data-users/leapfrog-hospital-survey 6. Joint Commission. Sentinel event alert 60: develop-ing a reporting culture: learning from close calls and hazardous conditions. 2018; Issue 60. https://www.jointcommission.org/sentinel_event_alert_60_developing_a_reporting_culture_learning_from_close_calls_and_hazardous_conditions/ 7. Campione J, Famolaro T. Promising practices for improving hospital patient safety culture. Jt Comm J Qual Patient Saf. 2018;44(1):23–32. 8. Clapper C, Merlino J, Stockmeier C. Zero harm: how to achieve patient and workforce safety in healthcare. 1st ed. New York: McGaw Hill Education; 2018. p. 73–98. 9. Hudson-Covolo JL, Rivers R, Irwin B. Daily inten-tional nurse leader rounding on patients. J Perianesth Nurs. 2018;33(1):90–5. 10. Reinertsen JL, Johnson KM. Rounding to influ-ence: leadership method helps executives answer the “hows” in patient safety initiatives. Healthc Exec. 2010;25(5):72–5. 11. Healthcare Performance Improvement White Paper. Leadership method for performance improvement: cascading roles and responsibilities. May 2012. 12. Throop C, Stockmeier C. Healthcare performance improvement white paper. The HPI SEC & SSER Patient Safety Measurement System for Healthcare (HPI 2009–001). 13. Brilli RJ, McClead RE, Wallace CV, Stoverock L, Berry JC, Wheeler TA, Davis JT. A comprehensive patient safety program can significantly reduce pre-ventable harm, associated costs, and hospital mortal-ity. J Pediatr. 2013;163(6):1638–45. 14. Sexton JB, Adair KC, Leonard MC, et al. Providing feedback following Leadership WalkRounds is associated with better patient safety culture, higher employee engagement and lower burnout. BMJ Qual Saf. 2018;27(4):261–70. 15. Sexton JB, Sherek PJ, Thomas EJ, et al. Exposure to leadership rounds in neonatal intensive care unit is associated with a better patient safety culture and less caregiver burnout. BMJ Qual Saf. 2014;23:814–22. 16. Institute of Medicine (IOM). Crossing the quality chasm: a new health system for the 21st century. Washington, DC: National Academy Press; 2001. 17. Weick KE, Sutcliffe KM. Managing the unexpected. 2nd ed. San Francisco: Jossey Bass; 2007. 18. American College of Healthcare Executives and IHI/NPSF Lucian Leape Institute. Leading a Culture of Safety: A Blueprint for Success: Boston, MA.:American College of Healthcare Executives and Institue for Healthcare Improvement; 2017. 19. Developing a reporting culture: Learning from close calls and hazardous conditions. Sentinel Event Alert. 2018;(60):1–8. PMID:30540371. 20. Slawomirski L, Auraaen A, Klazinga N. The eco-nomics of patient safety: strengthening a value- based approach to reducing patient harm at national level, OECD health working papers, vol. 96. Paris: OECD Publishing; 2017. https://doi.org/10.1787/5a9858cd-en. 21. Solutions for Patient Safety. https://www.solutions-forpatientsafety.org/. “Our Results”. Accessed 05 May 2019. 22. Egan, Y, Kimmell, J.What ‘Amazon health care’ could look like in 5 years. Advisory Board (News and Blogs);2018. 23. Cohen JK. Google turns 20: here are 7 ways it’s changing healthcare: Beckers health IT and CIO report, September 10th, 2018.

37© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 R. K. Shah, S. A. Godambe (eds.), Patient Safety and Quality Improvement in Healthcare, https://doi.org/10.1007/978-3-030-55829-1_3Creation of Quality Management Systems: Frameworks for Performance ExcellenceAdam M. Campbell, Donald E. Lighter, and Brigitta U. Mueller IntroductionIn the current healthcare environment, creating value for the patient while providing safe, high- quality care is paramount. Listening to patients and families is essential to ensuring a positive experience. As innovative payment methodolo-gies emerge, the quality of care is even more important at the system level. For hospitals, whether freestanding or system-based, to remain competitive and continue to provide excellent care, a mature structure for a Quality Management System (QMS) must be in place.A. M. Campbell (*) Children’s Hospital of The King’s Daughters, Department of Quality and Safety, Norfolk, VA, USAe-mail: [email protected] D. E. Lighter Institute for Healthcare Quality Research and Education (IHQRE), and Physician Executive MBA Program, University of Tennessee, Department of Business Administration, Knoxville, TN, USAe-mail: [email protected] B. U. Mueller ECRI, Plymouth Meeting, PA, USAe-mail: [email protected]• Learn how to use the Quality Management System to foster a culture of quality improvement and safety at all levels of the organization.Chapter Objectives• Understand the process of evaluating and improving your organization’s Quality Management System (QMS).• Define the holistic structure needed for a highly reliable Quality Management System.• Appreciate how a Quality Management System promotes quality improvement.• Assess how accreditation requirements and oversight ensure a proper Quality Management System is functional within your organization.Vignette 3.1Great Care Hospital (GCH) recently hired a new CEO, Dr. Maggie Improverson. Her

38After assessing the situation at GCH, Dr. Improverson took a multi-faceted approach to build the structure and processes necessary to ensure safe, high-quality care. First, it was impor-tant that the organization understood the history of Quality Management Systems. W. Edwards Deming helped set many of the standards and approaches to quality management, called Total Quality Management (TQM), that continue to thread through our practices in healthcare quality improvement. Deming’s landmark book, Out of the Crisis [1] published in 1982, quickly became a foundational guide to performance improve-ment. Principles such as those embodied in his 14 Points for Management were adapted for health-care by Lighter in the text Principles and Methods of Quality Management in Health Care [2] and are summarized below in Key Point Box 3.1. 1. Stay in business – Healthcare leaders must understand customers’ value proposition and respond accordingly if they want to remain in business. Interestingly, Deming included the provision “to provide jobs,” which perhaps can be translated in the healthcare industry as an admonition to ensure products and services are tailored to the marketplace to make sure that workers are practicing “at the top of their licenses.” 2. Adapt to the new economic age – In short, change is inevitable, and leaders will find it fruitless to resist the changes that are affect-ing healthcare today. Not only must leaders cope with the change, but they must, in turn, encourage staff and co-workers to find ways to innovate solutions, thus ensuring that an organization thrives in the new business environment. 3. Eliminate the need for inspection – Healthcare is probably one of the most heavily regulated and inspected industries except perhaps the nuclear power industry. Myriad organiza-tions like The Joint Commission (TJC), DNV GL, Centers for Medicare and Medicaid Services (CMS), the National Committee on Quality Assurance (NCQA), the Utilization Review Accreditation Commission (URAC), and others oversee experience at other hospitals included a distinct focus on quality and safety. The first order of business was to assess the state of the safety and quality of care within GCH. As leaders at Great Care Hospital (GCH) progressed through their quality journey, they often heard one common theme: “Our operations and work on qual-ity improvement are functioning in siloes. No one works together, and our objectives are never defined.” A large amount of staff time and effort was being put into quality improvement and safety projects, but the organization’s results were not changing for the better, leading to staff and leader-ship frustration. Leadership was concerned that engagement would fall, and care would become less and less safe. Given this situa-tion, the leaders at GCH decided to focus on how they currently managed overall quality and safety at their hospital and started the process of changing the struc-ture and culture of improvement.Key Point Box 3.1 Deming’s 14 Points for Management Adapted for Healthcare 1. Stay in business. 2. Adapt to the new economic age. 3. Eliminate the need for inspection. 4. Reward quality. 5. Improve constantly. 6. Institute on-the-job training. 7. Help people and machines do a better job. 8. Drive out fear. 9. Break down barriers. 10. Eliminate slogans, quotas, and man-agement by objective. 11. Restore pride in workmanship for hourly workers. 12. Restore pride in workmanship for managers. 13. Institute education and self-improvement. 14. Make quality everyone’s job.

39hospital, medical practice, and payer opera-tions to ensure compliance to standards and a baseline level of quality. In spite of all of this oversight, US healthcare continues to face challenges in performance compared with similar countries around the world [3]; simi-lar to Great Care Hospital, performance improvement has not been integrated into operations to ensure that quality outcomes are the norm. 4. Reward quality – Hospitals, physicians, and payers must learn to be both trusted vendors as well as find suppliers with whom to build trusting relationships. The healthcare indus-try is moving in this direction, with value-based purchasing pushing providers to ensure quality performance. Great Care Hospital spends substantial resources on quality of services, however, has been unable to “move the needle” to achieve the next level of performance and create the trust rela-tionship that ensures its customers are engaged with the institution. 5. Improve constantly – Deming was particu-larly prescient with this recommendation, and the principle behind the advice has been demonstrated in numerous industries besides healthcare. For example, automotive safety has benefited immensely from adopting tech-nology to reduce the chance for errors, and these trends have been observed throughout the world as a significant differentiator among automobile brands [4]. Similarly, healthcare organizations like the Henry Ford Health System in Detroit have used patient safety as a key way to distinguish themselves in the marketplace, providing safety data on their website and making patient safety their priority [5]. The philosophy of continuous quality improvement constantly reinforced by leaders can lead to superior performance in providing safe care to customers. 6. Institute on-the-job training – Healthcare workers are accustomed to the need for con-tinuing education requirements to maintain certifications and licensure, but on the job (OTJ) training goes beyond the occasional in-service or medical conference. Continuous improvement demands continuous learning, and that learning needs to be shared with everyone associated with processes that impact performance. Rather than waiting to convey new knowledge at the next depart-mental meeting, methods of distributing new ways of improving a process through regular daily communications, such as lean huddles or person-to-person communications, have to be created. 7. Help people and machines do a better job – Just as point 6 demands the institution of OTJ training, this point stresses that leaders must find ways of continually enhancing the interface between people and the machines used to deliver services. In today’s healthcare environment, human factors design is becom-ing more germane to the elimination of errors and increasing safe behavior. The goal of human factors design is to “mistake proof” equipment and processes, creating a system that supports the safety of the patient and staff. 8. Drive out fear – The use of fear as a motivat-ing factor in healthcare organizations has long been recognized as being ineffective. The days of the domineering surgeon who throws instruments and berates staff when problems arise are behind us, and the use of approaches such as Crew Resource Management and Just Culture have helped normalize behavior in healthcare institutions. 9. Break down barriers – This principle may be one of the major impediments to continuous improvement at Great Care Hospital. One of the crucial requirements of leadership is the ability to identify and then demolish barriers to effective communication and collaborative work. The senior staff at GCH likely will need to spend a great deal of time finding those processes that compete between departments and work to align the work of these departments to achieve synergy. 10. Eliminate slogans, quotas, and management by objective – The key message is that lead-ers should focus on the system as the arbiter of poor performance, rather than the work-force. In nearly every situation where perfor-mance lags, system and process design are

40flawed, and the workforce is trying to make the poor process work. Staff members fre-quently find “workarounds” to compensate for the defective process until leaders listen to workers and find ways of redesigning the process. Slogans, exhortations, quotas, and numeric targets can never counteract an inad-equate process. 11. Restore pride of workmanship for hourly workers – How often does one hear health-care workers, including physicians and other caregivers, complain that “all we do is move numbers through the system”? Giving front-line caregivers that opportunity to enjoy their work, realize the good they’re doing for the people for whom they provide care, and be appreciated for a successful care interven-tion, will reinforce the reasons that many of these professionals chose healthcare as a career and will lead to higher productivity and work satisfaction. 12. Restore pride of workmanship for manag-ers – Managers, too, need reinforcement for a job well done. Relieving the concentration on goals and targets as the sole motivating factors and finding approaches and measures that enhance customer engagement and satis-faction, then linking those customer param-eters to managers’ recognition, can help managers regain a sense of purpose that often is the motivator to growth in leadership positions. 13. Institute education and self-improvement – “Everyone in the organization should enjoy a sense of wellness, and programs that encour-age self-improvement through training and education programs have the potential of raising morale and worker engagement”. For organizations like GCH, lack of worker engagement can lead to poor performance, and lack of engagement will impact commit-ment to change and improvement. 14. Make quality everyone’s job – When the focus of a healthcare organization becomes excellent patient care, rather than just bud-gets and volume, customers will feel the dif-ference and become engaged with the organization. Workers will similarly feel that coming to work is something that is fulfill-ing, leading to improved performance and collective success.For organizations, like GCH that has stalled in its quality journey, Deming has some important ideas. First is the idea of a transformation. Dr. Improverson will be leading GCH on a journey to higher quality and greater safety through cultural and structural change. As the journey progresses, GCH will be able to set the foundation for future innovation and sustainable change. To promote this journey, it must be understood that incremen-tal changes are unlikely to motivate staff to engage in moving the hospital to the next level of perfor-mance. The transformation is driven from the top, i.e., leaders must support the change with plans and resources that identify performance factors and delineate approaches to evaluating, measur-ing, analyzing, improving, and sustaining new processes to take the organization to a higher level of customer satisfaction and economic achieve-ment. Leaders need to make the transformation part of everyone’s job, not just use catch phrases and slogans. Leaders need to ensure that workers and managers “own” the change and take credit for the improvements. GCH’s leaders will need to create a new work environment in which all these factors are addressed effectively. How can that happen? As we work through this chapter, we will see the path taken by Dr. Improverson. High Reliability OrganizationsVignette 3.2GCH has started examining the principles that will guide the hospital to higher perfor-mance. We have seen that Dr. Improverson has educated the staff on Deming’s under-lying principles of high performance. In addition, Dr. Improverson wants to instill the principles that will allow the organiza-tion to become aware of possible errors that could occur as well as ensure that the staff

41GCH wants to become a high reliability orga-nization (HRO). This term has become a buzz-word in the healthcare industry. It was first coined by Weick and Sutcliffe in 2007 in their book Managing the Unexpected [6]. The authors stud-ied diverse businesses that must maintain struc-ture and function in uncertain situations where there is a constant potential for error that can have disastrous consequences. They found that successful organizations used “mindful organiz-ing,” expressed in a set of five principles, three principles of anticipation, and two of contain-ment (Table 3.1). Organizations that observe these principles experience fewer accidents despite their complexity of operations because that complexity becomes more understandable and thus manageable. People in these organiza-tions focus both on performance-sustaining pro-cesses and increased efficiency, allowing them to not only catch errors early but also to use fewer resources to fix them [6]. Industries that are often mentioned as examples of HROs include avia-tion, nuclear power plants, and submarines but could certainly also include space travel or the Disney theme parks.There are many examples from other indus-tries that detail failures in safety systems that led to catastrophic events. While it may seem that these events are unrelated to healthcare, by exam-ining the underlying causes and failures, the sim-ilarities become clearer. A case in point was the January 28, 1986, explosion of the Space Shuttle Challenger. Given past launch pad explosions and other space-related events, the space pro-gram, in general, has been associated with high-risk/increased safety scenarios. Nevertheless, the Challenger broke apart 2 minutes into its tenth mission due to a failure of an O-ring in one of the rocket boosters. The failure was discovered after an extensive root cause analysis, and one of the major enabling factors was a culture of compla-cency and reluctance to speak up.How does this apply to healthcare? Healthcare experts have tried to use parallels to these indus-tries and apply them to the complex environment of caring for a vulnerable population, our patients. In healthcare, root cause analyses, hall-marks of high reliability organizations, are also performed to investigate potential serious safety events. One of the reasons people resist these comparisons is the oft-cited comment: “People are not widgets”, patients and their diseases have much more variability than airplanes, rockets, or submarines. This has made the implementation of HRO principles in healthcare a challenge, but nonetheless very important within the quality journey.When the principles of HRO are appropriately translated into the vocabulary of healthcare, it becomes clear that this framework, in fact, very much applies [7]. The goal, after all, is to identify Table 3.1 High reliability principlesAnticipationPreoccupation with failureRegarding small, inconsequential errors or deviations from the norm as a symptom that something is wrongRefusing to “normalize,” i.e., getting used to small deviationsAbsence of errors does not mean lower vigilance or complacencySensitivity to operationsPaying attention to what’s happening on the frontlineMake sure that people understand the impact of their work on the larger groupSituational awarenessReluctance to simplifyEncouraging diversity in experience, perspective, and opinionRespect and value the skepticsContainmentCommitment to resilienceDeveloping capabilities to detect, contain, and bounce-back from events that do occurLearn from mistakesDeference to expertisePushing decision making down and around to the person with the most related knowledge and expertiseEncourage people to ask for helpand Quality Management System are resil-ient if an error should occur. To accomplish this, she has introduced the concept of high reliability. She has put in place principles that ensure GCH does the right thing every time and that safety and quality principles are applied by all frontline staff.

42problems before they occur. Let us take the five tenets of the HRO approach and apply them to the healthcare sector (Table 3.1). Preoccupation with FailureWe know that errors, mishaps, or even disasters can happen at any time. Ideally, we prevent issues from happening by thinking through the “what if” scenarios ahead of time. For example, when we were preparing for potential Ebola patients, we simulated and repeatedly trained, always thinking of the “what ifs.” When a new unit or a new hospital is opened, hopefully, as Failure Mode, Effects, and Criticality Analysis (FMECA) has been conducted ahead of time (see Key Point Box 3.2) However, sometimes, we become aware of the risk through a report of a near-miss or “good catch” event that needs to be taken as seri-ously as an event that did reach the patient.Small signals may indicate future problems. When organizations analyze safety events, they classify them based on the severity of impact to the patient and the timeframe in which they were identified. For example, near-miss events are events that are identified and stopped before they reach the patient. In an HRO, near-miss events are of the utmost importance, as they identify ways in which errors were prevented from reach-ing and/or harming the patient. Please see Key Point Box 3.3 for more information on the event classification system many organizations use.Key Point Box 3.2 Failure Mode, Effects, and Criticality Analysis (FMECA)FMECA are methods designed to identify potential “failures” in a process before they occur. After mapping the process, a brain-storming team will assess the process and identify the steps in the process that may be high risk and be susceptible to failures. Each gap is rated using a scoring methodol-ogy that looks at occurrence rate and sever-ity of risk. Example: a hospital performed an FMECA to identify areas of risk during an Emergency Department lockdown procedure.Key Point Box 3.3Near-Miss Event: An event that does not reach the patient and causes no detectible harmPrecursor Safety Event: An event that reaches the patient but only causes mild or no detectable harm and that has the capacity to harm the next time it occursSerious Safety Event: An event that reaches the patient and causes moderate to severe harmVignette 3.4To explain this important topic, Dr. Improverson used a recent, real event that happened at another hospital:A patient received vecuronium, a para-lytic agent, instead of versed, an anxiolytic and sedative agent without paralytic prop-erties, and later died. The investigation found that the nurse overrode the Pyxis machine and pulled the wrong medication. But: (1) she was not the regular nurse for this patient; (2) to find versed in the Pyxis machine, she typed in “VE,” and the first drug that appeared was vecuronium; (3) double-check of medication was not per-formed; (4) patient was not monitored in radiology unit; (5) why would vecuronium even be stored in a Pyxis machine on a step-down unit? This nurse is now accused of murder, but there are so many system issues involved that just blaming one per-son is over-simplifying the events.

43 Sensitivity to OperationsThe earliest indicators of threats typically appear in small changes in organizational operations. These observations, most often by frontline workers, are important signals and, if acted upon, can help avoid the emergence of more wide-spread problems. Reluctance to SimplifyAnyone who has ever participated in a root cause analysis (RCA) has realized that the first (and most obvious) answer is never the full explana-tion for what happened. It is recommended to ask at least five times “why” to get deeper and deeper into the multitude of events that contributed to the failure or error. Remember the Swiss cheese model, where several holes need to line up for an error to make it through the whole “cheese” [8]. It takes many holes that just by chance line up for an error to make it through all the safeguards. Commitment to ResilienceDespite our best efforts and past successes, errors will occur, and safety will be threatened; HROs learn from mistakes instead of being paralyzed by them. Events like the one described above will shake an organization to its foundations but will hopefully also lead to many new improvements at that organization and throughout the healthcare industry.A promising movement is to learn not only from mistakes but also to adopt practices from areas where things go right. This is called the Safety II approach, compared to Safety I (learn-ing from past mistakes) [9]. See Chap. 12 Safety II for more details. Deference to ExpertiseHighly reliable organizations identify the person with the greatest expertise, instead of expecting the most senior person to come up with answers, when addressing issues. To take full advantage of the existing expertise, a hospital or other health-care environment needs to have a culture where everyone is able and willing to speak up, is feel-ing respected, and is commended for their input.Now that Dr. Improverson has instilled the foundational principles of high reliability, she must assess the organization’s current state as it relates to the implementation of a full Quality Management System. To accomplish this, senior leaders will have to provide structural support and resources to properly develop and maintain the QMS. A large part of the foundational sup-port for the QMS is via accreditation processes. As Dr. Improverson evaluates her organization, she must assess if the current accreditation agency and related processes are meeting GCH’s needs. Regulatory and Accreditation RequirementsHospital accreditation is a voluntary process. However, in order to be able to participate in fed-eral programs and bill Medicare and Medicaid for Vignette 3.3To help staff understand how important the recognition of safety events is and how even apparently unrelated or isolated inci-dences can lead to major problems, Dr. Improverson used a couple of recent issues that happened at GCH: 1. One patient with C. difficile infection? This can easily spread to a whole unit! 2. An infusion pump showed frequent occlusion alerts. Fortunately, several nurses reported this and a design flaw with these brands of pumps was found.She was able to show with these exam-ples that healthcare is not a static environ-ment and those new threats can occur at any time. All staff must think about poten-tial risks even in their daily routine work. She emphasized that complacency is a threat to safety and that highly reliable hos-pitals are always aware that they are operat-ing in a high-risk environment and that there is no “routine day.”

44services provided, hospitals and other healthcare entities must ultimately be accredited by the Centers of for Medicare and Medicaid Services (CMS) or one of the organizations that were given the authority to do so on behalf of CMS (called deemed authority) [10]. Accreditation provides an acknowledgment that the organization is commit-ted to patient safety and quality of care and strives for continuous quality improvement. There is evi-dence that the quality of care and patient satisfac-tion scores are higher in accredited hospitals [11]. Since federal payers cover so many patients, about 75% of all hospitals in the United States have decided to become accredited.Only accrediting organizations that adhere to the Conditions of Participation (CoP) and the Interpretive Guidelines (IG), the CMS manual, will be approved as having deeming authority through CMS. Accreditation can be obtained directly through CMS or its state agency, but very few organizations choose this pathway. There is ongoing controversy whether organizations with deeming authority are thorough and rigorous enough to satisfy CMS standards, and CMS regu-larly conducts validation surveys to verify the accuracy of the other organizations’ findings.The Joint Commission (TJC) is the largest accrediting body, focusing mostly on hospitals, including children’s and adult hospitals, acute and long-term care, as well as psychiatric hospi-tals, rehabilitation and specialty hospitals, sur-gery centers, and home health agencies. It received deeming authority in 1966 from CMS. It is constantly revising and updating its processes. In 2003 TJC started to include the National Patient Safety Goals, and in 2017 it introduced a new scoring grid that visually depicts the severity of the findings, the Survey Analysis for Evaluating Risk™ (SAFER™) matrix [12]. The SAFER™ matrix evaluates the likelihood for harm (low-moderate- high on the y-axis) against the preva-lence of the finding (limited-pattern-widespread on the x-axis).DNV GL received deeming authority for hos-pitals from CMS in 2008 and is accrediting a growing number of hospitals in the United States and internationally. Dr. Improverson, upon assessment of the current accrediting body for GCH, realized that merging the requirements of accreditation (through an organization such as DNV GL) with the structural benefits of QMS could greatly benefit GCH. From her experience, Dr. Improverson knows that structural criteria like those set forth in ISO 9001:2015 would addi-tionally benefit their journey. ISO 9001: 2015As GCH builds its QMS, there are certain structural criteria that are useful to follow. One such set of criteria are the ISO 9001: 2015 stan-dard. Within an organization like GCH, that is restructuring and improving its Quality Management System, these structures are essen-tial. Thus, the leaders of the hospital invested in training and resources to build this structure. Also, ISO 9001-2015 sets forth recommenda-tions for how GCH hospital can set up committee Vignette 3.5Upon her review of GCH’s current QMS structure, Dr. Improverson noticed that she was not being made aware of concerns and risks within the hospitals in a timely man-ner. Also, she was not given regular updates on progress. Dr. Improverson saw the need to restructure the sharing of this informa-tion, as she would need to give regular updates to the Board on major objectives and initiatives across the hospital. She planned to look at ways to set forth the nec-essary institutional structures to accom-plish better communication and strategic planning. She decided to start the integra-tion of ISO 9001:2015 principles in the organization. She also knew that this could be linked to the accreditation process in the future and would lead to better integration of strategy and outcomes. One such accred-iting body, DNV GL, links CMS require-ments to ISO 9001: 2015 standards. This seemed like a great opportunity for GCH.

45structures and senior leadership oversight to guide the strategic implementation of the QMS.The International Organization for Standardization (ISO), an independent, non- governmental international organization, was created over 70 years ago to ensure that products and services are safe, reliable, and of good qual-ity (https://www.iso.org/about-us.html) [13]. ISO standards provide a basic model for a Quality Management System for any industry and are updated regularly. The most current version is ISO 9001:2015. Although healthcare was late to adopt these standards, its use has become increas-ingly more common, and accreditation agencies such as DNV GL have made adherence to ISO 9001:2015 an integral part of their process.ISO 9001:2015 is not prescriptive and can eas-ily be combined with other quality management approaches, such as Lean and Six Sigma, the Toyota Production System, or the Malcolm Baldrige National Quality Award criteria. The Baldrige Award and ISO focus on leadership, strategy, customers and markets, as well as the workforce, process management, and results while assessing for continuous improvement, innovation, and agility. A commonly used tool in ISO is the Plan-Do-Study-Act (PDSA) approach.Having a QMS focuses the organization on what is important and helps make regulatory compliance more achievable. Furthermore, regu-latory standards often have not addressed basic management needs such as continual improve-ment, control of documented information, cali-bration of medical equipment, process-based internal audits/surveys, corrective action, and risk assessment. Coupling regulatory require-ments with ISO 9001: 2015 addresses these needs.ISO 9001:2015 has seven key tenets: customer focus, leadership, engagement of people, process approach, improvement, evidence-based decision making, and relationship management [14]. As can be seen in Fig. 3.1, the ISO principles help guide the QMS. In addition, we can see that the ISO principles also embody several of Deming’s principles.ISO 9001: 2015 presents criteria, organized into “clauses,” very similar to the criteria used in the Baldrige framework (see below). The most germane to QMS are:Understand current and future patient needsCreate valueAttract and retain patient confidenceCreate unity of purpose and direction (Quality policy; mission, vision and values)Strategic alignment (policies, processes and resources)Create conditions to achieve objectivesInvolve people at all levelsRespect people as individualsRecognize, empower and enhance peopleQMS comprised of interrelated processesWork systematically, not in silosUnderstand how results are achieved by systemOngoing focus on improvementRisk-based approachShow agility, react to changesData-centric (dashboards/scorecards/reports/governance)Covert data to information (analyze)Continual improvement (decide)Manage relationships with interrelated partiesLong-term trust with suppliersCreate value for hospital and suppliers/contractors (vendor management)•••••••••••••••••••••Customer focusLeadershipEngagement ofpeopleProcess approachImprovementEvidence-baseddecision makingRelationshipmanagementFig. 3.1 ISO 9001: 2015 principles

46• Clause 5: Leadership (organizational commit-ment and oversight)• Clause 6: Planning (addressing risk, risk-based thinking)• Clause 7: Support (resources, competency training, document control)• Clause 8: Operation (products and services, supply chain/management)• Clause 9: Performance Evaluation (how are we doing, problem identification)• Clause 10: Improvement (corrective action, continual improvement)To be compliant with ISO 9001:2015, the hos-pital must demonstrate its ability to provide prod-ucts and services that meet customer and regulatory requirements [15]. This starts with understanding what the strengths (and weak-nesses) of the organization and the requirements of the stakeholders are. However, it also sets lim-its: QMS cannot over-reach and thus must have boundaries to ensure proper scope [14, 15].ISO 9001:2015 (like other quality manage-ment systems) puts a heavy emphasis on leader-ship. Top management, including the Board, is not only ultimately responsible for the quality of care, but they are instrumental in assuring the success of the QMS. They must set directions and develop strategies to achieve the goals and objec-tives of the organization. Healthcare operations are complex, and many processes are dependent on each other. Standardization or at least harmo-nization among different areas is key to an effi-cient and smooth process. ISO ensures that the organization embodies a process orientation, focusing on inputs, process steps, and outputs of the process. Key elements include items such as resources, physical environment/facilities, and core competency (via job description and train-ing processes) and policy requirements.ISO 9001:2015 requires the organization to define and manage its risks associated with clini-cal service provision, including resources, equip-ment, and infrastructure. This includes both pro-active and retroactive evaluations, some of them very familiar in the healthcare environment including root cause analyses (RCA), Failure Mode, Effects and Criticality Analysis (FMECA), emergency preparedness, and others. Action plans and improvement process prioritization within the organization is based on the risk orien-tation of the process. Baldrige Performance Excellence AwardThe Malcolm Baldrige National Quality Award program was founded in 1987 when Secretary of Commerce Malcolm Baldrige observed that US companies were failing in their efforts to compete internationally. Baldrige focused the Department of Commerce on stimulating US industry to apply “quality control” to their enterprises to lower costs and improve competitiveness. Section 2(a)8(A) of the law states, “[the act helps quality and productivity by] helping to stimulate American companies to improve quality and pro-ductivity for the pride of recognition while obtain-ing a competitive edge through increased profits,” and subsection (B) goes on to say, “recognizing the achievements of those companies which improve the quality of their goods and services and providing an example to others” [16].The Baldrige award was codified by law not just to enhance business productivity and profit-ability but also to recognize those companies through the award process that provide an exam-ple for others to follow their lead. This dual pur-pose has guided the program since 1987 with demonstrated success at changing several busi-ness sectors in the United States. Criteria were developed initially for manufacturing companies, but over the years, new sectors were added, Vignette 3.6As GCH has matured in their QMS jour-ney, they have set themselves up to begin the “Baldrige Journey.” Baldrige is the pre-eminent award for quality and safety in the United States. The tenets of QMS and ISO 9001: 2015 lend themselves nicely to the criteria for the Baldrige Award (discussed below).

47including healthcare in 2000, and the healthcare sector has become one of the most active in adopting the framework and competing for the award.The framework consists of seven categories, each of which has several levels of criteria that do not serve as standards but rather ask “how” ques-tions about an organization’s structure, functions, and results. We will examine these areas in more detail, but first let us understand the foundation of the criteria, i.e., the Baldrige Core Values (Table 3.2). The Baldrige Framework is con-tained in a comprehensive booklet with updates every 2 years and is available for purchase at https://www.nist.gov/baldrige/products-services/baldrige-excellence-framework [17].The Framework starts by requiring the cre-ation of an organizational profile (OP) that delves deep into the enterprise structure and relation-ships (Table 3.2). In some cases, creating the OP provides leaders and managers with an under-standing of their organization that has eluded them in the past. The Baldrige Core Value of “Systems Perspective” requires everyone in the organization, but particularly leaders and manag-ers, to have an understanding of how work sys-tems are created and interact so that they understand the overall system, rather than the little piece of the system with which they are engaged. The OP provides that overview that is hard to achieve in any other way. The OP serves as the organizing resource for all of the rest of the Baldrige Framework and Criteria. Each Category of the framework must relate to one of the com-ponents of the OP, or the systems concept cannot be achieved. Table 3.2 lists the elements of the OP, which provide that comprehensive view of the organization and help connect processes and work systems for improvement.As the starting point of the Baldrige Journey, the OP forms the foundation of responses to the framework criteria. Criteria are written at three levels: 1. Basic items – the titles for each item. 2. Overall items – questions in boldface in the criteria booklet; these questions are the sub-ject headings for the multiple items that sum-marize the multiple questions. 3. Multiple items – the specific questions to address that get into the detail of the item.Most organizations will focus on multiple items, but some will find it difficult to respond to these very detailed questions. Usually, less mature organizations find it difficult to respond to questions at multiple levels, which is one way to identify opportunities for improvement (OFIs). If a question in the multiple items appears to be relevant, but there is no apparent approach to address the question, then the orga-nization has an OFI that requires an intervention.Additionally, the Baldrige Framework uses a mnemonic to gauge the effectiveness of a work process or work system – ADLI:• Approach – methods the organization uses to address a process, e.g., a process outline or description• Deployment – the extent and effectiveness that the approach is applied throughout the organization• Learning – collection and analysis of data and experience from the day-to-day operation of the process to improve the process and other similar processes throughout the enterpriseTable 3.2 Baldrige organizational profileP.1: Organizational descriptionOrganizational environment Healthcare service offerings Mission, vision, values, culture Workforce profile Assets (facilities, equipment, intellectual property) Regulatory environmentOrganizational relationships Organizational structure, including governance Patients, other customers, stakeholders Suppliers, partners, collaboratorsP.2 Organizational situationCompetitive environment Competitive position Competitiveness changes Comparative dataStrategic context Strategic challenges and advantagesPerformance improvement systems

48• Integration – synchronization of all the ele-ments and measures supporting process to achieve overall organizational goalsADLI is a method of evaluating organizational effectiveness and maturity. Almost every organi-zation has approaches for key processes, so the next level of maturity involves the extent of deployment of the approaches throughout the organization. Next, the question arises about whether the organization collects data about the operation of the process, i.e., how is the approach working? Finally, the highest level of maturity of application of the Baldrige Framework depends on how well the organization extends these approaches, deployment, and learning to all orga-nizational processes. Integration indicates that the organization’s processes are all working together to achieve strategic objectives.Does that sound like GCH? Analysis of the work systems at GCH will likely reveal that there are several approaches, but deployment, learning, and integration are lacking – all leading to sig-nificant opportunities for improvements for lead-ers, managers, and the workforce. So, what can the hospital do?Many healthcare enterprises have adopted the Baldrige Framework as the organizing approach for achieving the transformation that Deming recommends. The Framework promotes analysis of organizational processes using the Multiple Criteria and ADLI to assess the efficiency and effectiveness of the organization’s work. The Framework is briefly outlined in Table 3.3, and we’ll discuss some of the key elements that apply to healthcare entities like GCH.Category 1, Leadership, is probably the cru-cial opportunity for improvement for Great Care Table 3.3 Baldrige categoriesCategory 1LeadershipSetting vision and valuesPromoting legal and ethical behaviorCommunication and engagement of the workforce, key partners and customers, patientsCreating an environment for successCreating a focus on actionGovernanceResponsible governance systemPerformance evaluation of leaders and governanceLegal/regulatory complianceManagement of ethical behaviorSocietal contributions – societal well-being and community supportCategory 2Strategy developmentStrategy development processInnovationData analysis and decision supportWork systems and core competenciesStrategic objectives – balancing objectives among stakeholdersStrategy implementationAction plan creation, implementation, modificationResource allocationWorkforce plansPerformance measuresPerformance projectionsCategory 3Customer expectationsListening and learning from current and potential customersMarket segmentationHealthcare service offeringsCustomer engagementRelationship managementCustomer support and accessComplaint managementSatisfaction, dissatisfaction, engagementUse of voice of the customer data and market data

49Category 4Measurement, analysis, improvementPerformance measure data trackingComparative dataMeasurement agilityOrganizational performance reviewProjection of future performanceContinuous improvement and innovation using dataInformation and knowledge managementEvaluating data quality and availabilityOrganizational knowledge managementSharing best practicesOrganizational learning managementCategory 5Workforce environmentWorkforce capability and capacityRecruit, hire, onboard new workersWorkforce change managementWork accomplishment leveraging core competencies to reinforce customer serviceWorkplace safety, health, accessibilityWorkforce benefits and policiesWorkforce engagementDrivers of worker engagementAssessment of engagementOrganizational culture – communication, performance management, safety, engagementManagement of workforce performanceDeveloping the workforce (personal improvement)Effectiveness and efficiency of learning and development systemsCareer developmentCategory 6Work processesService and process design requirements and conceptsProcess implementation to address patient expectations and preferencesSupport processesService and process improvementSupply network managementInnovation managementOperational effectivenessManaging operation cost, efficiency, and effectivenessSecurity and cybersecuritySafety and emergency preparednessCategory 7Healthcare and process resultsResults for patient and customer service processesWork process effectiveness and efficiency resultsSafety and emergency preparedness resultsSupply network management resultsCustomer resultsPatient and customer satisfactionPatient and customer engagementWorkforce resultsWorkforce capability and capacityWorkforce climateWorkforce engagementWorkforce developmentLeadership and governance resultsLeadership communication and engagement with workforce, partners, patients, customersGovernance accountability resultsLaw, regulation, and accreditation resultsEthical behavior resultsSocietal well-being and key community support resultsFinancial, market, strategy resultsFinancial performanceMarket performanceStrategy implementation resultsTable 3.3 (continued)

50Hospital. Leaders have agonized over the perfor-mance of the hospital for some time, but no clear direction has emerged from their angst, and there aren’t any clear pathways to the performance excellence goals that they want to set. Baldrige organizations have developed Leadership Systems that employ behaviors that encourage employees to achieve stretch goals by clarifying vision and values through more advanced com-munication with all stakeholders, particularly the workforce. The work environment likely needs a redesign to create a focus on action, as well as inspiring and rewarding success. How might leaders achieve these goals? Using the experi-ence from nearly two decades of Baldrige health-care recipients is a good start [18]. Every Baldrige Award recipient provides a summary of its appli-cation to share with the public as a way of ensur-ing that the bright ideas and innovations that their teams have implemented are shared with others which can adapt these ideas to their organiza-tions. For example, Memorial Hospital and Health Care Center (2018 recipient) has shared information on its leadership practices via the Baldrige website.Additionally, each award recipient provides a contact person if someone wishes to get more detailed information about the organization’s approaches. So, GCH’s leaders need only click on the contact link on the website to send an email to the contact person and arrange a phone call to learn more. They may learn, perhaps, that Memorial’s leaders make daily administrative rounds and participate in regular “town hall” meet-ings, send hand-written “thank you” cards for exemplary employee actions to improve patient care (“Really Impressive Moments”), or send the “Friday Facts” email every week. Most Baldrige recipients are eager to share these approaches with others and often present their best practices at con-ferences and online meetings.Once a leadership system is in place, the team should turn to the other categories, and most organizations that commit to the Baldrige Journey appoint “Category Champions” for each of the first six categories. Often these champions are leaders from the C-suite; for example, the CEO might lead Category 1, Leadership, and if the organization has a planning department, the head of that group might lead Category 2, strategic planning. Each category is assigned to the expert in that area to ensure that the information needed to respond to the multiple criteria can be expertly addressed. Note also that each Approach-Deployment (AD) category has one or more associated results items to ensure that results are linked with approaches and deployment.As the Category Champions organize teams to respond to each of the Baldrige categories and the detailed questions in the framework, they will select people from around the organization who have intimate knowledge of how each approach, or process, is deployed within their divisions or departments. As information is gathered, each of the items in ADLI needs to be addressed so team members will be tasked to answer questions like:• Approach –What part of the overall organizational work system does our department perform? –What process or processes do we use in our department to implement our piece of the overall work system?• Deployment –How is the process implemented within our department? –Who is involved in ensuring the process is done properly? –How well is the process running, e.g., does everyone follow the process in the same way?• Learning –How do we measure process performance, i.e., what metrics do we use to determine efficiency and effectiveness? –How do we collect internal and external customer experience data with the process? –How do we integrate the information (quantitative and qualitative) to inform improvement plans? –How do we incorporate this integrated learning from the measures into improving the process? –How and when do we re-measure to ensure that improvement plans are effective?

51• Integration –How is our performance improvement activity used by other departments to enhance this process or other similar or related processes? –How do we access and use performance improvement results from other depart-ments to augment our efforts?It is interesting to see the effect of this effort on organizational learning. Many times, as Category Champions are doing their analyses, they immediately find opportunities for improve-ment in their approach or deployment that can be the subject of improvement efforts, but even if A-D issues do not arise, there will inevitably be issues in measurement of performance or in the ability of each department to share and integrate their experience with others. In any event, just the process of conducting a Baldrige review virtually always spurs the Category Champions to identify issues that they can address to meet their own strategic objectives better.GCH is poised to make significant gains using the Baldrige Framework. Not only will the frame-work provide the structure for organizing the hos-pital to make more cogent goals, but the use of ADLI will also help create a focus on the action ensuring that appropriate efforts will be made to achieve those goals. Moving from the broad agenda set by Deming’s 14 Principles to action plans using the Baldrige Framework is achievable, regardless of organizational size. Managers now have tools to attain performance goals, and GCH will soon embark on the Baldrige Journey (Table 3.3). ConclusionsOver months and years, Dr. Improverson trans-formed the structure, performance, and most importantly, the culture of GCH. From under-standing the history of quality systems and high reliability principles to the importance of founda-tional elements like information technology and the workforce and customer focus, a system was put in place to ensure that problems were sur-faced and addressed. Through this system, the quality and safety of care increased to the level of a top performing hospital. Dr. Improverson embodied the appropriate role of leadership to guide the hospital through the transformation and put in place a system that was built on valid data, a satisfied workforce, and most importantly, a satisfied patient and family.Editors’ CommentsThis chapter is at the core of quality improvement and patient safety in health-care – how does one (e.g., a hospital, a department, a quality leader) utilize a Quality Management System to drive toward higher levels of reliability? The authors answer this query by showing the readers in a simple manner the complexities of Quality Management Systems and the predominant systems that exist currently in American healthcare. The authors expound upon the 14 Points of Management, one of Deming’s major early contributions. The editors would be remiss to not recommend Deming’s book titled, The New Economics for Industry, Government, Education – 2nd Edition [19]. We have our own hospital-based quality improvement and safety teams reading this book which serves as a way to have the learner understand the beauty, simplicity, and provide confidence in quality improvement. The authors, as seen multiple times in this textbook, make the important connection between Quality Management Systems and the journey toward high reliability. It is important to show explicitly how these two major con-cepts intertwine; the authors do this nicely in the middle part of the chapter. The authors, throughout the chapter, demon-strate several types of Quality Management Systems and how they can drive improve-ments; the chapter ends with a thorough discussion of the Malcolm Baldrige National Quality Award program. One may argue that their organization “will never get there,” “is not ready,” “doesn’t know where to start” on the Baldrige Journey; however,

52 Chapter Review Questions 1. Why do health systems need to become learning organizations?Answer: A learning organization exhibits the willingness to change and embraces continu-ous quality improvement. In the current healthcare environment, one based on value, quality, and safety, continuous improvement is essential. Also, learning organizations fol-low high reliability principles, making them agile in response and resilient to error. 2. What is the benefit of training an organiza-tion to be problem solvers?Answer: As issues, or errors, arise, organiza-tions must be able to solve problems to ensure that mistakes to not repeat. A system must be able to identify high-risk problem areas and have standardized processes for solving them. Only then, can an organization improve its safety and quality performance and underlying culture. 3. Within the Baldrige Framework, what does ADLI stand for? A. Approach-Deployment-Learning- Integration. B. Alignment-Deployment-Learning- Integration. C. Approach-Deployment-Learning- Information. D. None of the Above.Answer: A 4. When you are paying attention to what is happening at the frontline, which high reli-ability principle is being followed? A. Reluctance to Simplify. B. Deference to Expertise. C. Preoccupation with Failure. D. Sensitivity to Operations.Answer: D 5. How can organization’s benefit from going through the accreditation process (e.g., DNV GL accreditation)?Answer: Accreditation is an opportunity for systems to identify high-risk processes and develop sustainable solutions to issues. Also, using criteria like ISO 9001: 2015, organiza-tions can design their quality management system to ensure appropriate structures are in place to support the QMS and that effective solutions to high-risk problems are overseen. 6. What does “having a culture of safety” mean? A. Making sure that patient feel safe. B. Making sure that employees feel safe. C. Having metal detectors at the hospital entry points. D. Always be prepared that something could go wrong. E. Having job security.Answer: D 7. What is the importance of “near-miss events” (select all that apply)? A. They provide a learning opportunity. B. They can be warning signals. C. They can indicate sloppy work. D. They can be used to determine who needs to be disciplined. E. Focusing on near-miss events will help prevent real events.Answers: A, B, E 8. What are important ISO 9001: 2015 princi-ples (select all that apply)? A. Involve people at all levels. B. Work systematically, not in silos. C. Use a risk-based approach. D. Ongoing focus on improvement.Answers: all of the above 9. What is the role of senior management/lead-ership in a quality management system?the authors answer all these questions and discuss the value in being on the quality journey via the Baldrige Criteria, etc. The key take away from this chapter is that a Quality Management System is a keystone in a successful safety and quality improve-ment program. Without deliberately build-ing a Quality Management System, a healthcare organization will become stag-nant and ultimately suffer significantly. This chapter provides an excellent roadmap to embark upon the quality journey – with a roadmap.

53Answer: Senior management (both in ISO 9001:2015 criteria and Baldrige Performance Excellence criteria) oversee strategic inputs into the QMS and ensure that it is properly resourced. Senior management also tracks progress on key initiatives and maintains the oversight of high-risk areas throughout the system. 10. Dr. Improverson has begun an initiative to train the entire organization on QI methods and “Safety First” Culture. Which of the high reliability principles will Dr. Improverson be addressing in this initiative?Answer: Dr. Improverson will be addressing all HRO principles. (1) Preoccupation with Failure: with a “Safety First” mindset, staff will be able to proactively address safety concerns before they happen. This will mani-fest itself as “near- miss” events in the organi-zation’s safety event reporting system. (2) Sensitivity to Operations: training all staff will give QI and safety culture capability to the frontline. (3) Reluctance to Simplify: staff will be able to break down complex problems and develop appropriate solutions. (4) Commitment to Resilience: staff will be agile and empowered as problem solvers, and leaders will be able to support staff when an error does occur. (5) Deference to Expertise: staff will be able to reference experts within the system as well as grow as experts themselves.References 1. Deming WE. Out of the crisis. Cambridge: Massachusetts Institute of Technology; 1982, 1986, 2018. 2. Lighter D, Fair D. Principles and methods of qual-ity management in health care. Sudberry: Jones & Bartlett; 2004. 3. Raleigh VS. Trends in life expectancy in EU and other OECD countries. OECD Health Working Paper No. 108. Organisation for Economic Co-operation and Development (OECD) Directorate for Employment, Labour and Social Affairs Health Committee. 2019. 4. Trends in automotive safety: eliminate the blind spot. https://www.renesas.com/us/en/products/automo-tive/auto-safety-eliminate-blind-spot.html. Accessed March 2019. 5. Henry Ford’s commitment to quality and safety. https://www.henryford.com/about/quality. Accessed March 2019. 6. Weick KE, Sutcliffe KM. Managing the unexpected: sustained performance in a complex world. 3rd ed. Hoboken: Wiley; 2015. 7. Hines S, Luna K, Lofthus J, et al. Becoming a high reliability organization: operational advice for hos-pital leaders. Rockville: Agency for Healthcare Research and Quality; 2008. 8. Reason JT. Human error. Cambridge, UK/New York: Cambridge University Press; 1990. 9. Braithwaite J, Wears RL, Hollnagel E. Resilient health care: turning patient safety on its head. Int J Qual Health Care. 2015;27:418–20. 10. Viswanathan HN, Salmon JW. Accrediting organiza-tions and quality improvement. Am J Manag Care. 2000;6:1117–30. 11. Lam MB, Figueroa JF, Feyman Y, Reimold KE, Orav EJ, Jha AK. Association between patient outcomes and accreditation in US hospitals: observational study. BMJ. 2018;363:k4011. 12. Facts about the SAFER™ matrix scoring process. 2018. https://www.jointcommission.org/facts_about_the_safer_matrix_scoring/. Accessed 13 Mar 2019. 13. International Organization for Standardization. https://www.iso.org/about-us.html. Accessed 13 Mar 2019. 14. Cochran C. ISO 9001:2015 in plain English. Chicago: Paton Professional Books; 2015. 15. Levett JM, Burney RG. American society for qual-ity. Using ISO 9001 in healthcare: applications for quality systems, performance improvement, clinical integration, accreditation, and patient safety. 2nd ed. Milwaukee: ASQ Quality Press; 2014. 16. Public Law 100-107. 1987. at https://www.govinfo.gov/content/pkg/STATUTE-101/pdf/STATUTE-101-Pg724.pdf. 17. The Baldrige framework booklet. Available at https://www.nist.gov/baldrige/products-services/baldrige-excellence-framework 18. Baldrige award recipients listing. https://www.nist.gov/baldrige/award-recipients. Accessed March, 2019. 19. Deming E. The new economics for industry, govern-ment, education. 2nd ed. Cambridge, MA: MIT Press; 2000.

55© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 R. K. Shah, S. A. Godambe (eds.), Patient Safety and Quality Improvement in Healthcare, https://doi.org/10.1007/978-3-030-55829-1_4Reliability, Resilience, and Developing a Problem-Solving CultureDavid P. Johnson and Heather S. McLeanAbbreviationsHRO High reliability organizationLOR Level of reliabilityMFI Model for ImprovementmFMEA Modified failure mode and effects analysisPDSA Plan-Do-Study-ActSMART Aim Specific, measurable, achiev-able, relevant, timely aimD. P. Johnson Monroe Carell Jr. Children’s Hospital at Vanderbilt, Department of Pediatrics, Vanderbilt University School of Medicine, Nashville, TN, USAe-mail: [email protected] H. S. McLean (*) Duke Children’s Hospital, Department of Pediatrics, Duke University School of Medicine, Durham, NC, USAe-mail: [email protected] Objectives• Understand how to use the Model for Improvement to drive change in an organization.• Understand how to incorporate princi-ples of reliability science within the Model for Improvement framework to achieve an outcome by coupling safety culture and process design.• Learn how to incorporate the power of problem-solving and the expertise of frontline staff to achieve an outcome.• Recognize the interplay of problem- solving techniques like the Model for Improvement with reliable process design and resilient safety culture.Vignette 4.1A tertiary care, academic hospital has a problem with hand hygiene compliance rates. This is not a new problem, but with the added focus of a new “Zero Harm” campaign, the executive team demands that all hospital units achieve hand hygiene per-formance rates of 100%. Despite efforts to educate and remind employees to clean their hands, hand hygiene performance rates remain at 85–92% with wide variation across areas. The Chief Medical Officer seems visibly frustrated that these efforts have failed, and she charges the quality improvement team with fixing the problem.

56 Opening Question/ProblemThe intent of this chapter is not about how to per-form effective hand hygiene but rather how to design a quality improvement project using high- reliability concepts operating within the Model for Improvement framework [1]. This case-based example of a quality improvement project about hand hygiene will illustrate these concepts throughout the chapter [2]. Understanding the principles of high-reliability science that couples reliable process design with the values of resil-iency in the safety culture are key to achieving and sustaining higher levels of performance. While the results achieved in this case are not at the level of a highly reliable process (>1 error in 10,000–100,000 events), it illustrates how a proj-ect team can incorporate specific change con-cepts with known levels of reliability to achieve their desired level of reliability. Model for ImprovementWhen facing a difficult problem or task, one needs a structured problem-solving technique to provide a framework for an effective, focused, and disciplined approach. The Model for Improvement (MFI), the subject of The Improvement Guide, provides such a framework for any type of improvement task – from personal life, to industry, and, of course, to healthcare [1]. Vignette 4.2A multidisciplinary improvement team, representing the medical-surgical care units, has been assembled including two nurse managers, an infection prevention nurse, medical director, nursing care assis-tant, and hand hygiene auditor. The team evaluates the system that has been in place for years and notes the presence of a hospital- wide hand hygiene auditor pro-gram that directly observes and records encounters, monthly dashboards display-ing unit performance, posted signs, and intermittent educational programs target-ing units with poor compliance. Hand hygiene performance data with targets are included on the unit, and aggregated bal-anced scorecards are updated monthly. Managers are held accountable for meeting these targets annually and share data with nursing staff during monthly meetings. Hand hygiene results have been plotted on a statistical process control chart that dis-plays the combined units’ monthly average hand hygiene compliance percentage. The results show a baseline median of 87% and wide variation with a range of 64–94%. The team decides to use the Model for Improvement to design the project. They map the process, conduct a modified failure mode and effects analysis (systematic method of identifying and addressing potential failures) (Key Point Box 4.1), develop a SMART aim statement, examine key drivers, and prioritize interventions in a key driver diagram (Fig. 4.1) [1].Key Point Box 4.1 Modified Failure Mode and Effects Analysis (mFMEA)A simplified version of the method used by process and product designers to identify and address potential failures before imple-mentation of change. This method is used in a proactive manner rather than tools that evaluate a problem that has already occurred such as root cause analysis or cause and effect (fishbone) diagram. The mFMEA is used with the project team as a group exercise with the goal of defining a high-level process map, then identifying failures in each step of the process, and finally proposing solutions to address the failures. The solutions proposed in this exercise can be used to populate the “inter-ventions” section of the project key driver diagram.

57This model, comprised of three questions fol-lowed by the Plan-Do-Study-Act (PDSA) cycle (Fig. 4.2), has proven to be a powerful tool for driving improvement. The three questions include (1) What are we trying to accomplish? (2) How will we know that a change is an improvement? (3) What change can we make that will result in improvement?Many healthcare professionals believe improvement occurs by holding meetings which often don’t result in actionable plans. The topics in these meetings may drift, and often a consen-sus plan is never reached. The attraction of the MFI is twofold. First, a small amount of initial planning to sequentially answer the three ques-tions, followed by focused testing (PDSA cycles) based on those theories, can result in more effi-cient improvement. Second, as the MFI becomes more widely used in healthcare, it can serve as a universal language between hospital units, hospi-tals, and health systems to help spread quality improvement successes. Question 4.1: What Are We Trying to Accomplish?While this question seems easy enough to answer, many teams experience difficulty articulating exactly what they are trying to accomplish unless they specifically set out to answer this question. In healthcare, team members often come to the table to discuss a problem, and lengthy conversa-tions can ensue with multiple ideas put forth. If each member were asked exactly what they are trying to accomplish, few would be able to articu-late a goal, and many might articulate contrasting or conflicting goals. This poses numerous prob-Smart aimKey driversInterventions (level of reliability)Increase health careworker (HCW)compliance withhand hygieneprotocols from 87%(LOR 1) to 95% (LOR2) by June1, 2014on two inpatientunitsHand hygiene suppliesconsistently available at thepoint of careReal time identification offailuresCompliance with hand hygieneisthe normProvide compliance information-Send weekly control chart byemail & post on each unit-Send weekly identification offailures (LOR 1)Develop restocking & repairprocess for supplies byEnvironmental services & nursingassistants (LOR 2)HCWs knowledgeable about howand when to be compliant withhand hygieneEliminate health careassociatedinfections (HAI)Global aimImplement hand hygienechampion program to identify &mitigate non compliance (LOR 2)Unit leaders committed and serveas role models & educators (LOR 1)Post visual reminders on unitsregarding compliance protocols(LOR 1)Leadership committed toimproving hand hygienecomplianceEmpower families to remind HCWsto perform hand hygiene byposting educational reminders(LOR 2)Fig. 4.1 Hand hygiene project key driver diagram. (Reproduced with permission from Hospital Pediatrics, McLean et al. [2] © 2017 by the AAP)

58lems, not the least of which is that team members might be working to accomplish different tasks, and sometimes these can be at odds with each other.Teams benefit when they spend time docu-menting exactly what they want to accomplish. Though creating this aim itself can take con-siderable time for some groups to achieve, this diligence will help prevent scope creep, and the long-term benefit toward the team’s goals will be considerable. Goals that use the SMART aim mnemonic [3] provide a concise, easily understandable goal for team members and non-team members alike. These goals are Specific, Measurable, Achievable, Relevant, and Timely. The handwashing SMART aim for the hospital in this chapter is to increase health-care worker compliance with hand hygiene protocols from 87% to ≥95% within 9 months in two pediatric inpatient units, leaving very little doubt to anybody who knew their work exactly what they were striving to do. Question 4.2: How Will We Know That a Change Is an Improvement?In quality improvement, measurement and data analysis are paramount. To determine if a change results in improvement, a team needs to know exactly what it is they are trying to improve and what is the unit of measurement for success. Often, the main measure of interest is articulated in their SMART aim statement which provides the team with some guidance. However, some-times the measurement requires some clarifica-tion, and an operational definition is needed [1, 4]. For example, in our hand washing example, what does it mean for somebody to properly wash their hands? Does it have to be with foam? Can it be soap and water? When does hand wash-ing have to occur in relation to donning a gown and gloves for patients on isolation? Such defini-tions provide clarity to the team to ensure they are comparing “apples to apples” during their improvement cycles with data collection and pro-vide a concrete definition of what is to be improved. The operational definition for measurement also provides the staff with a stan-dard work process expectation as they enter and exit patient rooms.Once a team knows what to measure and how to measure it, the methods of analyzing the data become important. Since improvement, by defi-nition, occurs over time, it is necessary that the data be tracked as such, with more frequent data collection (daily or weekly) being preferred over longer periods of time. Multiple, successive data points provide near real-time information to teams as they test changes, implement proven changes, or work toward sustainment. When ana-lyzed with run or statistical process control charts, teams use specific statistical rules to understand variation in their data, separate data signals from noise, and quickly learn the impact of their tests of change on their systems. Identifying common cause variation (variation inherent to a system) and special cause variation (variation that is not expected within the system) What are we trying toaccomplish?How will we known that achanges is an improvement?What changes can we makethat will result in improvement?ActPlanStudyDoModel for improvementFig. 4.2 Model for improvement. (Reproduced with per-mission from Associates in Process Improvement [9])

59provides teams valuable insight into each inter-vention’s impacts and the actions they should take [1, 5, 6]. Question 4.3: What Changes Can We Make That Will Result in Improvement?For many novice teams, this is where quality improvement work both begins and ends. Everybody wants to provide a solution, and team members jump to answer this question before answering questions 1 and 2. A group describes the problem, people say with some certainty what should happen to fix the problem, discussion ensues, action items are identified, and the meet-ing adjourns. No real goals. No measurement plans. And there will likely be frustration at the follow-up meeting because the only “proof” of whether or not something helped is personal anecdotes. When using the Model for Improvement, this question should only be addressed after questions 1 and 2 are answered. With a unified goal, the team knows exactly what measure they are following to determine whether or not their interventions are altering their system.The key driver diagram (Fig. 4.1) for the case vignette visually depicts all three questions of the Model for Improvement. A key driver diagram can quickly anchor a team, answering the first two questions very clearly in the aim statement and with the key drivers. The team is then ready to brainstorm some ideas for question 3 and can easily see how any potential ideas relate to the key drivers and the SMART aim, resulting in more focused discussions and preventing scope creep. The diagram is also a living document, changing as the project progresses with new knowledge and potential interventions. Testing Changes: The Plan-Do-Study- Act (PDSA) CycleThe Plan-Do-Study-Act (PDSA) cycle might be the most well-known quality improvement acro-nym but might also be the most poorly under-stood conceptually. There are multiple misperceptions of the cycle itself [7], and, per-haps as a result, the medical literature is full of instances where the term PDSA was invoked, but there is no evidence that PDSA cycles actually took place [8]. The PDSA cycle, which is informed by and used in conjunction with the first three questions for the Model for Improvement, enables teams to learn quickly about the feasibility and effectiveness of the pro-posed interventions. The PDSA cycle is based historically in the scientific method with the intention of producing new knowledge based on hypothesis testing [4]. Therefore, a true PDSA cycle requires not just putting a change in place but also deliberately studying the results in rela-tion to the team’s hypothesis. The bidirectional arrows between the first three questions and the PDSA cycle in the Model for Improvement are deliberate and vital to its purpose (Fig. 4.2) [9]. In essence, each PDSA cycle builds the team’s knowledge of their process over time by gaining insight from data through these sequential tests of change (Fig. 4.3) [1].Having answered the first three questions of the Model for Improvement, a team now has a specific SMART aim statement and understands how they will know if the changes result in an ActPlanStudyDoWhat changes areto be made?Next cycle?Complete theanalysis of the dataCompare data topredictionsSummarize whatwas learnedCarry out the planDocument problemsand unexpectedobservationsBegin analysisof the dataObjectiveQuestions andpredictions (why?)Plan to carry outthe cycle (who, what,where, when)Plan for data collection••••••••••••Fig. 4.3 The Plan-Do- Study-Act cycle. (Reproduced from the Improvement Guide Fig. 5.2 with permission from Wiley Books; Clifford et al. [1])

60improvement (their chosen measure), and they will have brainstormed ideas that might improve their system. Teams are now ready to conduct their PDSA cycles. PlanTeams first plan a small test of change based on their team’s predictions. Taking one of the pro-posed interventions, the team can try to incorpo-rate that on a limited scale to begin to understand the effectiveness. For example, in our case vignette, the team would learn much faster by testing interventions on a single room than “roll-ing out” a new policy to an entire hospital. The team should be explicit about the details of their test – where it will happen, what they will do, what data they will collect and how, and even predict what might happen. The team needs not to have consensus before testing an intervention as there might be significant resistance to change. In fact, allowing team members to predict failure can assist teams’ cohesiveness and encourage everyone to share their thoughts and concerns. DoThey then do the test exactly as it is laid out. The intervention might be done by a single provider or in a single room. As the intervention is tested, data are collected to inform the next steps. StudyUsing feedback from the person or people doing the test or those impacted by the test, they then study the results. The study of the results can be either in the form of qualitative data or a quantita-tive measure related to the SMART aim. Did it go as planned? How were the staff impacted by the change? Did it have the intended effect? Do the results move the team closer to their intended goal? How do the data change the perceptions (if at all) to those that were resistant to the change? If their concerns were borne out by the data, what other changes would they suggest? If the test resulted in signs of improvement, how can these data be used to begin to assuage their hesitancy for change? ActThrough the new knowledge gained in this test-ing, the team then acts. They choose to either adopt, adapt, or abandon this test. In rare cases when the first test achieved its desired effect per-fectly, the team may adopt the test and attempt it on a larger scale, ramping up to conduct another PDSA cycle with multiple providers or multiple rooms. More commonly, there are mixed results from which important lessons are gleaned about the intervention’s potential effectiveness. Modifications to the test of change are done, and a new PDSA cycle is conducted again on a small scale. Finally, in some scenarios, the tested inter-vention does not have the desired effect or is not well-received by those whom it will affect, and the team chooses to abandon the intervention altogether. Regardless of whether the team chooses to adopt, adapt, or abandon, the PDSA cycle is a success because they gained important insight into their system without disruption. The Strength of the Model for ImprovementThe Model for Improvement brings structure and discipline to any quality improvement project and applies to all organizational levels [5]. This focused stepwise learning process based on test-ing theories through the iterative PDSA cycles allows teams to learn from their tests, use accepted statistical methods, and improve their process faster than other approaches [5]. This model is also easy to teach and can be adopted by frontline staff who can begin to work as teams and solve everyday issues that might not rise to the level of management or leadership. Allowing staff to solve their own problems and have early wins can improve morale and resilience. Finally, the structure of the Model for Improvement can be utilized as the framework alongside many other quality improvement methods such as Lean and Six Sigma.In our vignette, the previous system was not achieving the desired results despite hard work

61and the best intentions. We needed to use a struc-tured problem-solving approach. The creation of a SMART aim statement provided a unified vision for the team, and each person knew what they were striving for and how to play their part (Question 1 in The Model for Improvement). Data analysis using the control chart (Fig. 4.4) allowed them to analyze their data in real time, providing important insight as to whether or not the tested changes were making a difference (Question 2 in The Model for Improvement). The key driver diagram allows them to propose inter-ventions that would result in improvement (Question 3 in The Model for Improvement) and then proceed with PDSA cycles to inform their decisions. With this new problem-solving struc-ture in place, the frontline staff could now address concerns using the Model for Improvement and be empowered to voice larger concerns to their leadership.Computer screensavers visual reminderUnit leaders serve as role models and educatorsWeekly feedback of dataStandardizing supply systemIsolation signs visual remindersDoor signvisual remindersSafetyChampion program Empowering patients & families with posted informationData reported monthly50%60%70%80%90%100%8/05/13 (n=039)8/19/13 (n=026)9/02/13 (n=030)9/16/13 (n=021)9/30/13 (n=090)10/14/13 (n=158)10/28/13 (n=041)11/11/13 (n=058)12/01/13 (n=014)12/15/13 (n=032)12/29/13 (n=031)1/12/14 (n=032)1/26/14 (n=028)2/09/14 (n=035)3/02/14 (n=055)3/16/14 (n=047)3/30/14 (n=042)4/13/14 (n=033)4/27/14 (n=038)5/11/14 (n=023)5/25/14 (n=023)6/30/14 (n=106)8/31/14 (n=114)10/31/14 (n=265)12/31/14 (n=317)2/28/15 (n=159)4/30/15 (n=222)6/30/15 (n=206)8/31/15 (n=143)10/31/15 (n=257)12/31/15 (n=264)2/29/16 (n=155)Percent compliant (%)Date endingPercent compliant (%)Average proportion of compliant encountersControl limitsGoal line (≥ 95%)reminderStaWeeklyfeeddbbbaadatareminddDesired directionof changeDatareporteFig. 4.4 Statistical process control chart (percent or p-chart) showing percent compliant hand hygiene encoun-ters on two inpatient units with annotations of test of change. The x-axis is labeled with every other week or month, and data points are weekly until June 2014 when they are measured monthly. (Reproduced with permission from Hospital Pediatrics, McLean et al. [2] © 2017 by the AAP)Vignette 4.3Now that the initial phases of the project were complete, the team is excited to start testing the interventions they had planned. First, the team decides to develop new posted paper signs and computer screen-savers to remind staff to clean their hands to see if fresh new ones placed in different areas would help nurses who no longer noticed the old ones. In addition, unit lead-ers (nurse managers and medical directors) decide to leverage their roles by discussing hand hygiene performance during meet-ings and to provide regular, frequent feed-back to nurses, nursing care assistants, and physicians. The team uses a control chart (Fig. 4.4) to display weekly hand hygiene compliance data and posts them in

62In this case vignette, the team has so far focused on using the Model for Improvement framework to improve a process. The team is missing two key elements and needs to incorpo-rate them into their work if they are going to achieve and sustain the results they are seeking – developing a more reliable process and resilient safety culture. Both are needed for the team to achieve the goals of this project. So what does this mean? How can the team apply these principles to improve hand hygiene? (Key Point Box 4.2).First, let’s understand how to develop a more reliable process. The term reliability, as it applies to healthcare, as described by Berwick and Nolan in 2003, is defined as “the measurable capability of a process, procedure, or health service to per-form its intended function in the required time under commonly occurring conditions.” [10]. Reliability can be quantified as a ratio of failures or errors per number of opportunities (Fig. 4.5). Most healthcare processes operate at levels of reliability with error rates of 10 or more per 100 opportunities (10–30% or 10−1 failure rate or level of reliability of 1 [LOR 1]) as compared to high reliability organizations (HROs), such as the nuclear power industry or commercial aviation, which have failure rates of 0.0001% or 10−6 (LOR 6) [12]. The hand hygiene failure rate the Ratio oferrors/opportunitiesReliabilityFailure percent rateFailure rateExamples1/100.91010-1Hand hygiene compliance1/1000.99110-2Pediatric adverse drug events1/10000.9990.110-3General surgery deaths1/10,0000.99990.0110-4Road safety1/100,0000.999990.00110-5Giving wrong blood to patient1/1,000,0000.9999990.000110-6Nuclear industryMore reliableFig. 4.5 Measures of reliability displayed as ratios of failures per number of opportunities, reliability, failure percent rate, and failure rate with examples from health-care and industry for each level to illustrate these differ-ences mathematically. (Adapted from Pediatric Clinics of North America, Luria et al. [12], © 2006, with permission from Elsevier)workrooms and distributes them in emails so the performance is viewed by all nurses, physicians, and other staff working on the units. Performance improves >95% at first but, unfortunately, drifts back down to the mid- 80s in a few weeks. The team feels frustrated with these results and decides to take a step back. A quality improvement coach suggests the team needs to under-stand the principles of reliability science before proceeding with the project. The team agrees that the interventions used so far will not sustain a high level of perfor-mance and want to learn more about how they can design the process in a different way.Key Point Box 4.2 Reliability and ResiliencyReliability – the measurable capability of a process, procedure, or health service to perform its intended function in the required time under commonly occurring conditions [10]Resiliency – the safety culture of an organization is able to systematically understand failures that occur and make adaptations to improve over time [11]

63team observes is 10–15% or level of reliability of 1 (LOR 1). We know from our vignette that this is expected since the process in place includes only training, feedback, and reminders. The Chief Medical Officer in our case is asking the team to design a process with a higher level of reliability equal to or better than 1 or fewer failures per 100 opportunities (1% failure rate or 10−2). In order to achieve this failure rate, the team in the vignette will need to incorporate additional interventions into the project design if they are going to achieve this level of reliability. Studies of human factors engineering and design show us that the team needs to consider interventions such as incorpo-rating decision aids, redundancy, and taking advantage of habits and patterns in order to achieve this level of reliability [12]. Put another way, the team will need to “hard wire” the pro-cess by using these types of tactics to create a more reliable design. Use of a visual trigger placed at the entrance of the patient room that notifies the healthcare worker of noncompliance in real time is an example of a human factors engineering intervention that could be used to improve hand hygiene compliance results. Smart process design is critical, but without changing the culture or behaviors of the people working in the area, the team will not be able to achieve and sustain the results they are seeking. To under-stand more about coupling reliable process design and resilient culture into a healthcare improvement project, we can learn from indus-tries that are high reliability organizations (HROs).High reliability organizations, such as nuclear power and commercial aviation, achieve both a reliable process and resilient culture with error rates in the order of 1 in 10,000–100,000 oppor-tunities. Weick and Sutcliffe examined HROs and described key features that can be applied to complex healthcare processes, measure perfor-mance, and design interventions to achieve desired results. These authors identified five prin-ciples of high reliability that are common to HROs shown in the box below (Key Point Box 4.3) [11]:In summary, the team can use the Model for Improvement framework for the overall project design and implement both reliable process and resilient safety culture change concepts as inter-ventions that are indicated on the key driver dia-gram (Fig. 4.1). Using PDSA cycles and tracking the impact of these multimodal changes over time on the control chart (Fig. 4.4) will help the team understand when they have achieved special cause variation and reached their goal of ≥95% compliance (less than 5 failures per 100 opportu-nities) with hand hygiene protocols.Vignette 4.4Empowered with a new understanding of concepts of reliability and resiliency, the team reviews the key driver diagram (Fig. 4.1) and decides to test interventions with a level of reliability (LOR) greater than 1. Now it is clear that the reminders, education, feedback of data, and engage-ment of leaders were examples of level of reliability 1 (LOR 1) interventions and these alone will not give the team the results they desire of less than 5 failures per 100 hand hygiene opportunities. The team Key Point Box 4.3 Five Principles Common to High Reliability Organizations (HROs) [11] 1. Preoccupation with Failure – small fail-ures are noticed, reported, and learned from continuously by the organization 2. Reluctance to Simplify – embrace com-plexity and welcome diverse experience 3. Sensitivity to Operations – attentive to frontline workers’ expertise 4. Commitment to Resilience – ability to learn and bounce back after failure 5. Deference to Expertise – authority migrates to the person with most exper-tise regardless of rank

64 Value of Problem-SolvingProblem-solving involves an intentional process to break down complex issues into actionable components in an effort to create solutions. The Model for Improvement represents only one problem-solving technique, but many more methods exist that can help inform or augment a team’s problem-solving strategy. For example, the “5 Whys” can help determine the root cause of a problem [13] and inform a team’s plan to test changes through the Model for Improvement. Toyota Production System’s 8 Steps of Problem Solving, discussed elsewhere in this text, pro-vides another established framework to solve problems and ultimately improve outcomes.In healthcare, problem-solving often requires altering the fundamental way a system operates, impacting the frontline staff much more directly than management and leadership who classically are the ones determining how to solve the prob-lem. Quality improvement and patient safety in healthcare require a different approach – an approach in which the experience, expertise, and knowledge of the frontline staff are valued and in which they are given the freedom to improve also realizes they need regular interaction with frontline staff and real-time observa-tion of the unit practice to get to the root of the problem. It is clear to the team that engaging the true experts (deference to expertise and sensitivity to operations) in the testing and implementation is the key to achieving and sustaining the goal.During observation of hand hygiene practice on the units, the project team learns from frontline staff that hand sani-tizer canisters in and outside of each patient room are not replaced consistently. As a result of this problem-solving, the team decides that it is important to standardize the hand sanitizer resupply process for both the environmental services worker and the nursing care assistant roles. The idea is to not only create standard work but also to have each worker role responsible for the hand sanitizer and soap resupply process. Therefore, if one individual fails to replace a hand sanitizer canister, then the other one will catch it so that it would be a rare occur-rence for there to be no sanitizers available at the point of care, thereby incorporating the design concept of redundancy into the system. During additional observations and discussion with frontline staff, the team recognizes the value of when a healthcare worker gently reminds another person to clean his/her hands before entering or leav-ing the patient room. In order to foster a culture in which compliant hand hygiene practice is the norm, the project team decides to implement a multidisciplinary hand hygiene champion program to pro-vide real-time mitigation across the units. The idea is to have a knowledgeable peer recognize noncompliance (separate from the hospital-wide auditing process) and provide gentle and respectful feedback, therefore turning a noncompliant encoun-ter into a compliant one. Not only will this practice improve hand hygiene compliance results, but it will also create an environ-ment where people feel comfortable raising concerns that foster a resilient safety cul-ture. Finally, during observations and dis-cussion with patients and families on the units, the decision is made to involve the patients and families as partners in this pro-cess. Since care is centered around the patient and family, empowering them to speak up provides an additional layer of accountability and further strengthens the culture of safety. Implementation of all three of these change concepts positively impacts safety culture and process design by incorporating the high reliability prin-ciples described by Weick and Sutcliff (See Key Point Box 4.3).

65their environment for the better of their patients. This embodies the essence of Weick and Sutcliffe’s five principles of HROs [11]. When done well, most problems can be solved quickly by those doing the work to reduce “workarounds.” Management and leadership personnel are then freed to spend more time with forward-thinking exercises and less time “putting out fires.”W. Edwards Deming’s theory of profound knowledge focused intently on people’s abili-ties and their innate desire to feel like an impor-tant contributor to their workplace [14]. A key component of this desire is its ability to solve daily problems and see immediate results. Fostering a sense of cooperation instead of competition will raise the level of performance of an entire team, resulting in better results than the sum of each team member’s abilities [14]. When teams can harness these abilities and use a disciplined framework such as the Model for Improvement, the frontline team members’ understanding of their process, observation of the issues at hand, and ideas for improvement can be harnessed to problem-solve efficiently and effectively.Importantly, as teams work together, they must not only think of their own results but the goals of the entire organization. Russell Ackoff, a revolutionary systems thinker, wrote: “If each part of a system, considered separately, is made to operate as efficiently as possible, the system as a whole will not operate as effectively as possi-ble.” [15] In essence, working in “silos” might help one team meet a metric, but that team’s “win” may hinder the system as a whole. This issue is not unique to healthcare. As an example, General Stanley McChrystal led the Joint Special Operations Task Force in Afghanistan and had to rethink how his teams worked together. He dis-covered that traditional military hierarchy was not nimble enough to effectively accomplish his Task Force’s goals. By creating a “team of teams” (also the title of the book), he was able to empower the frontline members on his units to solve problems efficiently and effectively. This approach also created relationships between the teams such that the broader mission’s goals were taken into account as decisions were made in the field [16]. In effect, he harnessed and magnified each team members’ ability as they worked within and between teams. This approach empowered team members to speak up and to problem-solve within the boundaries to their stated mission, fostered a sense of self-worth and cooperation, shattered silos, flattened hierarchy, and led to efficiencies and successes that the Task Force had not previously seen.The ability and desire of people to problem- solve based on their knowledge of their system propelled McChrystal’s model to success. This approach essentially established a high reliability organization by building reliability and resilience where it was needed most – in the people who were carrying out the important work. Healthcare can harness problem-solving in a similar manner. Frontline workers, based on their knowledge of the system, can provide ideas to lead to improve-ment through structured approaches such as the Model for Improvement. When teams discover changes to the system that are successful, “cross- talk” between silos can lead to larger improve-ments through more reliable process design. And, perhaps most importantly, this cross-talk between silos can lead to profound resilience as teams around an organization are able to speak freely to each other and to leadership, identify and verbal-ize a problem, and propose action knowing that their voice will be heard.Vignette 4.5Brainstorming and feedback from staff now regularly occur during safety and operational meetings as well as during intermittent, unannounced visits to the units. Use of the multidisciplinary champi-ons helps to sustain the results and continue to incorporate the principles of high reli-ability. Following standard statistical pro-cess control chart rules, the centerline shifted twice during the project when special- cause variation occurred (Fig. 4.4). The project control chart now shows the

66 ConclusionIn this chapter, the example of the challenges faced, and successes achieved, by an actual improvement team highlights the importance of using a structured improvement approach (the Model for Improvement) and in harnessing the knowledge and ability of frontline staff to problem- solve. This, in turn, creates a resilient safety culture that is coupled with reliable process design. The Model for Improvement propelled the team to the next level, assisting them in identify-ing a SMART aim and producing theories that would help them test and measure interventions to determine if intended changes were occurring. These small wins achieved through using the Model for Improvement improved morale, pro-vided frontline staff a voice, and engaged staff in identifying solutions that could be tested. Each organization must make incremental changes, using examples of small wins gained through quality improvement methods to reinforce front-line problem-solving. With this, reliability and resilience become symbiotic with the quality improvement methods, each building on the other, creating an upward spiral toward any healthcare organization’s goal of becoming an HRO and bringing them closer to a goal of “Zero Harm.”results they desire with hand hygiene com-pliance sustained ≥95%. Project results are reported to executive leadership of the hos-pital with an emphasis on pairing reliable process design with a resilient safety cul-ture that is needed to give these two units the results they need. The hospital Chief Medical Officer celebrates the results of the project and helps the team plan for spread to other units in the hospital.Editors’ CommentsReliability, resilience, and problem-solving are the core of improvement science. This chapter highlights the difficulties we face in healthcare using a vignette of hand hygiene. The vignette demonstrates that something as simple as washing one’s hands prior to caring for a patient is com-plex to perform reliability and consistently. There is no better exemplar than hand hygiene; if we cannot deconstruct this issue into its constituent parts and perform it with reliability and with resilience, then we will fall short of major improvement initia-tives, which are sorely needed in health-care, such as reducing readmissions, decreasing length of stay, and optimizing patient throughput.We would like the reader to appreciate the significance of the Model for Improvement and strategies to approach change (PDSA cycles); the authors go in depth on these concepts to ensure that the reader will have the requisite knowledge to try and use these approaches for their improvement. This chapter espouses the traditional surgical mantra of see one, do one, teach one; the chapter is fundamental and written at an appropriate level to serve as a primer or toolkit for a novice to under-stand the techniques and try these on a small scale in their span of control.The important concept of reliability is further developed in this chapter with the authors once again pulling from Weick and Sutcliffe’s five principles of high reliability organizations. We feel it crucial for the reader to continually hear about these five principles and see how they are applied to various situations; it is in this way that the reader will develop a profound respect and understanding of the power of these prin-ciples as an overarching framework for improvement science.Conceptually, the hardest part of the chapter is to describe and attempt to reach problem-solving. We believe the authors convey this very well toward the end of the chapter. Once we understand reliability and resilience, the difficulty is how to develop

67 Chapter Review Questions 1. What are the three questions the Model for Improvement asks teams to address in the design of a project?Answer: (1) What are we trying to accom-plish? (2) How will we know change is an improvement? And (3) what change can we make that will result in improvement? 2. What is the difference between the concepts of reliability and resiliency?Answer: Reliability is the measurable capa-bility of a process, procedure, or health ser-vice to perform its intended function in the required time under commonly occurring con-ditions [10]; resiliency is the safety culture of an organization and its ability to systematically understand failures that occur and make adap-tations to improve over time. 3. What are the five high-reliability principles that are described by Weick and Sutcliffe?Answer: (1) Preoccupation with Failure – small fail-ures are noticed, reported, and learned from continuously by the organization (2) Reluctance to Simplify – embrace com-plexity and welcome diverse experience (3) Sensitivity to Operations – attentive to frontline workers’ expertise (4) Commitment to Resilience – the ability to learn and bounce back after failure (5) Deference to Expertise – authority migrates to the person with most expertise regardless of rank 4. True or false: “Zero harm” results in patient safety can be achieved by incorporating reli-able process design into a healthcare system alone.Answer: False (need to use both reliable process design and resilient safety culture con-cepts in order to achieve “zero harm” results). 5. True or false: Engagement of frontline staff in the PDSA cycles for improvement can be essential for successful problem-solving and positively impacts the safety culture of the organization.Answer: True.References 1. Langley GJ, Moen RD, Nolan KM, Nolan TW, Norman CL, Provost LP. The improvement guide: a practical approach to enhancing organizational performance. 2nd ed. San Francisco: Jossey-Bass Publishers; 2009. 2. McLean HS, Carriker C, Bordley WC. Good to great: quality-improvement initiative increases and sustains pediatric health care worker hand hygiene compli-ance. Hosp Pediatr. 2017;7(4):189–96. 3. Ogrinc GS. Joint commission resources Inc., Institute for Healthcare Improvement. Fundamentals of health care improvement: a guide to improving your patient’s care. 2nd ed. Oak Brook Terrace/Cambridge, MA: Joint Commission; Institute for Healthcare Improvement; 2012. 4. Perla RJ, Provost LP, Parry GJ. Seven propositions of the science of improvement: exploring foundations. Qual Manag Health Care. 2013;22(3):170–86. 5. Provost LP, Murray SK. The health care data guide: learning from data for improvement. 1st ed. San Francisco: Jossey-Bass; 2011. 6. Wheeler DJ. Understanding variation: the key to man-aging chaos. 2nd ed. Knoxville: SPC Press; 2000. 7. Reed JE, Card AJ. The problem with plan-do-study- act cycles. BMJ Qual Saf. 2016;25(3):147–52. 8. Taylor MJ, McNicholas C, Nicolay C, Darzi A, Bell D, Reed JE. Systematic review of the application of the plan-do-study-act method to improve quality in healthcare. BMJ Qual Saf. 2014;23(4):290–8. 9. Associates in Process Improvement. 2019. https://www.apiweb.org. Accessed 17 June 2019. 10. Berwick D, Nolan T. High reliability health care, presented at the Institute for Healthcare Improvement’s 15th annual national forum on quality improvement in health care. 2003. http://www.ihi.org/resources/Pages/Presentations/HighReliabilityHealthCarePresentation.aspx 11. Weick KE, Sutcliffe KM. Managing the unexpected: sustained performance in a complex world. 3rd ed. Hoboken: Wiley; 2015. 12. Luria JW, Muething SE, Schoettker PJ, Kotagal UR. Reliability science and patient safety. Pediatr Clin N Am. 2006;53(6):1121–33.and sustain a problem solving culture. The authors draw from their experience and the literature to provide approaches for this difficult part of quality improvement.

68 13. Improvement IfH. 5 whys: finding the root cause. 2019. http://www.ihi.org/resources/Pages/Tools/5-Whys-Finding-the-Root-Cause.aspx. Accessed 25 Feb 2019. 14. Deming WE. The new economics: for industry, gov-ernment, education. 2nd ed. Cambridge, MA: MIT Press; 2000. 15. Ackoff RL. Creating the corporate future: plan or be planned for. New York: Wiley; 1981. 16. McChrystal SA, Collins T, Silverman D, Fussell C. Team of teams: new rules of engagement for a complex world. New York: Portfolio/Penguin; 2015.

69© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 R. K. Shah, S. A. Godambe (eds.), Patient Safety and Quality Improvement in Healthcare, https://doi.org/10.1007/978-3-030-55829-1_5Building an Engaging Toyota Production System Culture to Drive Winning Performance for Our Patients, Caregivers, Hospitals, and CommunitiesJamie P. Bonini, Sandip A. Godambe, Christopher D. Mangum, John Heer, Susan Black, Denise Ranada, Annette Berbano, and Katherine StringerAbbreviationsCMS Centers for Medicare and Medicaid ServicesDART Days Away, Restricted, or TransferredED Emergency DepartmentEHR Electronic Health RecordFTA Fast-Track AreaHRO High-Reliability OrganizationIHI Institute for Healthcare ImprovementIOM Institute of MedicineLVN Licensed Vocational NurseMIT Massachusetts Institute of TechnologyMVV Mission, Vision, and ValuesNASEM National Academies of Sciences, Engineering, and MedicineNICU Neonatal Intensive Care UnitPDCA Plan-Do-Check-AdjustPDSA Plan-Do-Study-ActPIs Pressure InjuriesPO Per Oral or By MouthRN Registered NurseTPN Total Parenteral NutritionTPS Toyota Production SystemTSSC Toyota Production System Support CenterWOT Wound Ostomy TeamJ. P. Bonini Toyota Production System Support Center (TSSC), Toyota Motor Corporation, Plano, TX, USAe-mail: [email protected] S. A. Godambe (*) Children’s Hospital of The King’s Daughters, Norfolk, VA, USAe-mail: [email protected] C. D. Mangum Children’s Hospital of The King’s Daughters, Department of Quality, Norfolk, VA, USAe-mail: [email protected] J. Heer Toyota Production System Support Center (TSSC), Toyota Motor Corporation Australia, Melbourne, Australiae-mail: [email protected] S. Black · D. Ranada · A. Berbano · K. Stringer Kaizen Promotion Office, Harbor-UCLA Medical Center, Torrance, CA, USAe-mail: [email protected]; [email protected]; ABerb[email protected]; [email protected] vignettes in this chapter are fictional. A glossary can be found at the end of this chapter.

70This fictitious example demonstrates timely problem-solving that those in healthcare are try-ing to emulate, where a defect or problem is quickly identified and analyzed, while the evi-dence is fresh. The frontline team temporarily stops the production line so that the cause of the defect can be uncovered and immediate group problem-solving can occur. The countermeasures were rapidly implemented which prevented any future recurrences. The alignment of the assem-bly plant with its suppliers can, at times, facilitate this rapid and joint problem-solving. The organi-zational culture that supports this took years to develop through shared experimentation and learning. It is not unique to this particular Toyota plant. Rather, this culture can be found at any of their plants around the globe. The frontline work-ers and Toyota leadership know that no defect is to be passed forward. Pushing defects through the system results in increased costs from muda (waste) such as rework, recurrence of defects, customer and employee dissatisfaction, and pos-sible safety concerns.Systems, both human and computerized, need to be in place to rapidly identify defects, devia-tions from the standard, or abnormal conditions. Healthcare providers, patients, and leaders of other industries often wonder if similar systemic Chapter Objectives• To share perspectives and learnings from the early years of applying Toyota Production System (TPS) principles to healthcare• To show how TPS principles align with high-reliability organization (HRO) principles• To make TPS principles relatable and understandable to people with varying backgrounds, especially in healthcare• To show the value of creating frontline problem-solvers to improve performance• To share perspectives and learnings on building a successful, high-performing TPS culture in healthcareOpening VignetteIt is early morning at a Toyota plant. The morning huddle with the various line teams has disbanded, and the production output has been increased to meet greater demand, so a new Toyota vehicle will come off the line every 55 seconds. At the morning hud-dle, everyone was notified that new loading fixtures will be installed to better assist with positioning large dashboards during installation into the new production vehi-cles. As production occurs, the team tasked with attaching the dashboards notices that several loading fixtures seem to cause slight blemishes on the dashboards, which are visible only under certain lighting conditions.A team member immediately pulls a cord (called an andon cord) which acti-vates a flashing light and musical tone to signal that a problem has occurred. A problem- solver, who is also the team leader, hears the musical tone and arrives within 20 seconds to assist with the situa-tion and immediately begins investigating the problem.Not too far behind are other plant lead-ers who coach a problem-solving exercise with the local team. The team follows the process upstream from the point at which the problem occurred and finds that the new loading fixture lightly rubs some dash-boards during the loading process. The clearance between the fixture and the dash-board was insufficient. The countermea-sures are immediately deployed, including increasing the clearance from 1/8 inch to 3/4 inch and placing tape around the fix-ture – which prevented scuffing if the fix-ture accidentally contacted the dashboard during placement. After the countermea-sures were operationalized, no further defects were noted.

71cultural changes can be instilled in their organi-zations. Such an ideal state employs all four parts of Deming’s System of Profound Knowledge [1] – Appreciation of System, Theory of Knowledge, Understanding Variation, and Psychology of Change. All of these parts are interrelated (Fig. 5.1). An examination of the fig-ure shows that the ability of the parts to interre-late and work well together is dependent upon the values of the organization (represented by the handle of the magnifying glass).Paul O’Neill has discussed the value of habit in changing culture. As the CEO of Alcoa, his focus on employee safety aligned his frontline workforce around a universally acceptable and popular agenda along with delivering improved organizational profitability. The encouragement of habit formation was key in his empowerment of the frontline [2]. Such principles have been applied by Toyota in the Toyota Production System (TPS) since the 1950s. Similarly, through their focus on organizational safety and develop-ment of frontline team members, Toyota’s finan-cial outlook has improved. More importantly, they have developed a sustainable organizational culture focused on frontline development. It will become apparent that TPS is much more than habit creation, but rather an organizational culture and quality management framework that can help an organization become a learning system, a high-reliability organization, and a desired place to work that achieves and sustains rigorous safety, quality, value, and financial goals.Weick and Sutcliffe [3] described the need for increased organizational “mindfulness” in the quest for high reliability. A high-reliability orga-nization (HRO) operates under trying conditions but nonetheless manages to have fewer than their share of adverse events. We will return to this dis-cussion of high reliability at the end of this chap-ter as TPS is a problem-solving, culture-centered improvement system that embodies and facili-tates the successful implementation of the five HRO principles: 1. Preoccupation with failure 2. Sensitivity to operations 3. Reluctance to simplify 4. Commitment to resilience 5. Deference to expertiseAs mentioned in other chapters, preoccupa-tion with failure refers to the constant vigilance about seemingly small or inconsequential issues being signs of bigger problems. Sensitivity to operations refers to the focus on what is happen-ing on the “shop floor” or where production of goods or delivery of services is occurring. A Appreciationof a systemTheory ofKnowledgePsychologyUnderstandingVariationValuesFig. 5.1 Deming’s system of profound knowledge. (Reprinted from by Langley et al. [1] with permission from John Wiley & Sons)

72reluctance to simplify interpretations encourages diversity in opinions, experiences, inputs, and perspectives. Finally, the latter two principles are most applicable when an error or defect occurs, as no system is perfect. There will need to be anticipatory processes in place that facilitate learning when failures do occur. Commitment to resilience refers to an organization’s ability to contain problems and create rapid solutions after errors are investigated. Deference to expertise involves people with the most relevant expertise, regardless of their position in the organizational hierarchy, in any post-event assessments or problem- solving. The introductory vignette dem-onstrated these principles. Problems are to be expected, so we need to design systems to rapidly detect and react to these problems and prevent recurrence.From our travels to various hospitals that are supposedly implementing “Lean” as their improvement methodology, the core values that the Toyota Production System represents are often misrepresented and/or misinterpreted.Here are a few common misconceptions:• Misconception 1: “LEAN is an acronym that stands for Less Employees Are Needed.” Leaders and consultants can be quick to assume that a reduction in workers is the answer to cost reduction – which is entirely contradictory to Toyota’s philosophy of respect (will be discussed further in the sec-tions, “TPS Triangle: Philosophy Arm” and “TPS Approach to Delivering Value”).• Misconception 2: “You need to spend large amounts of money for consultants to success-fully implement TPS in your organization.” Toyota considers its people as its most valu-able asset. Building an organizational culture of highly engaged and empowered individuals starts from within (will be discussed in the proceeding sections).• Misconception 3: “Implementing TPS in healthcare means we’re all going to work like robots.” When created and implemented cor-rectly (i.e., developed by the people who do the work and validated continuously at the genba or workplace), standardized work is one of the most powerful tools in TPS that keeps processes and practices safe, reliable, and evidence-based. In healthcare, the goal is to standardize around the patient, so that team members can do what they are trained to do – which is to care for people and patients (will be discussed further in “TPS House” section). TPS Approach to Delivering ValueI will say again: the only way to generate a profit is to improve business performance and profit through efforts to reduce cost. This is not done by making workers slave away, to use a bad expres-sion from the olden days, or to generate profit by pursuing low labor costs, but by using truly ratio-nal and scientific methods to eliminate waste and reduce costs. – Taiichi Ohno [4]Over the past two decades, the healthcare industry (especially in the United States where costs are among the highest globally and out-comes are not necessarily the best overall) has been challenged to improve value in its care deliv-ery systems. Some US healthcare professionals and administrators have suggested that the increased costs are related to the increasing com-plexity of the procedures or the use of more advanced and, at times, more expensive technol-ogy. However, compared to other countries per-forming comparable procedures, the US health systems remain costlier with poorer outcomes [5].Value is defined simply as quality divided by cost [6]. Healthcare leaders are often asked by their senior leaders and board members for the return on investment (ROI) for quality and safety. This discussion is difficult, at best, as some ben-efits cannot be readily measured [7]. Many indus-tries, including healthcare, determine the price of their services using the following equation:Selling Price, as set by the companyCost of Goods or Se[]=rrvicesProfit[]+””

73Therefore, organizations need to control costs to assure a reasonable profit to reinvest and sur-vive. In healthcare, fruitful partnerships must occur with our patients, their families, insurers, communities, school systems, other health sys-tems, and pharmacies to ultimately reduce costs and deliver value. Toyota proposes reducing costs using TPS as described by the TPS Triangle (Fig. 5.2) and the TPS House (Fig. 5.3). Reduction of costs through the reduction of workforce is not congruent with TPS principles and is detrimental to workforce morale and advancement of corpo-rate production and quality goals:Cost reduction must be the goal of consumer prod-uct manufacturers trying to survive in today’s mar-ketplace…there is no magic method. Rather a total management system is needed that develops human ability to its fullest capability to best enhance creativity and fruitfulness to utilize facili-ties and machines well, and to eliminate all waste – Taiichi Ohno [4]In healthcare, the focus is on preventive and proactive care (e.g., routine physical exams, immunizations, proper diet and exercise) to pre-vent the more expensive care like emergency department visits. Improving operational effi-ciencies is the desired result. This includes waste reduction, outcomes, and costs all while increas-ing workforce and customer satisfaction. Scientific Method and Becoming a Learning OrganizationTaiichi Ohno, the former Vice-President of Toyota Motor Company and TPS leader who helped develop TPS in the 1950s–1970s, often spoke of the intelligent frontline team members who surface problems, work to quickly create countermeasures, and solve these identified prob-lems through testing and application of scientific methodology. In this intentional process of creat-ing and testing hypotheses, a robust learning sys-In this equation, as the costs of goods and services will increase over time due to increases in raw materials or staff costs, the selling price is usually increased to achieve the needed profit. We know that the healthcare market will only bear small increases in costs, if any, given the amount of gross domestic product already allocated to overall population medical needs, including direct care, preventa-tive care, technology, research and develop-ment, and pharmaceuticals [8]. Similarly, Toyota has long believed that its customers and market conditions limit the price that can be charged. The automotive market is very competitive and will not bear high prices. To survive and reinvest in the future, a company must be profitable by reducing its costs. Toyota rewrites this equation asToyota Production SystemFig. 5.2 Toyota Production System (TPS) Triangle. (Used with the permission of Toyota)Selling Price, assetbythe marketCost of Goods or Servic[]–eesProfit”[]=”

74tem results. The National Academies of Sciences, Engineering, and Medicine (NASEM), formerly the Institute of Medicine (IOM), has recom-mended in its numerous publications [9–11] that health systems emulate this very environment, as it greatly enhances organizational agility while creating a system that is most desired by patients, their families, and society as a whole. The role of senior leaders or administrators is to lead, coach, and facilitate the work of frontline members and their development of problem-solving expertise. Simultaneously, these leaders should increase their visibility to the frontline team members and regularly visit the shop floor or areas (e.g., clini-cal and nonclinical) where the improvement is desired, during which time they can observe, receive input, and provide guidance. The value of local or unit-based huddles, especially with senior leaders present, cannot be overemphasized to drive frontline engagement with TPS and improvement efforts. Decades after the creation of TPS, the Institute for Healthcare Improvement (IHI) High-Impact Leadership framework espouses these very concepts [12].TPS places tremendous value on the develop-ment of the frontline worker and the creation of a corporate culture where people are trained to become problem-solvers or scientists. The appli-cation of the scientific method in real time on the automotive shop floor allows learning to occur rapidly, which in turn leads to innovation. This corporate approach supports the development of teams of problem-solvers who are empowered to drive change and innovate. The frontline Toyota workers are vital corporate assets and, by invest-ing in their growth (a concept known as people development), they help create a learning factory where knowledge is gleaned from planned experimentation. This new knowledge is applied and shared throughout the organization – corpo-rate agility results, employees feel valued, every-one wins, and a competitive corporate edge arises. Taiichi Ohno once said that “knowledge is something you buy with money. Wisdom is something you acquire by doing it [13].” You learn by doing! History’s Effect on the TPSHistorically, Toyota started out by making auto-matic looms. Some principles of TPS were intro-duced during this time of Toyota’s development. The founders of Toyota wanted to provide a greater service to society through automotive manufacturing [4]. The automotive arm of Toyota started in the 1930s, well after other global auto-makers. Not surprisingly, they had to overcome specific challenges when competing with these larger volume, more technologically advanced, global competitors like General Motors and Ford Motor Company.Fig. 5.3. Toyota Production System House. (Used with the permission of Toyota)

75Additionally, post-World War II Japan had some challenges not seen in the United States: 1. Geography, especially given its island loca-tion off the coast of Asia, with space chal-lenges and limited natural resources. 2. Impaired industrial infrastructure. 3. Limited market for automobiles. 4. Only 2% of automobiles sold were Japanese in origin, and, therefore, the market was domi-nated by foreign manufacturers. 5. Vehicles were much more technologically complex when compared with Toyota’s for-mer business line, automatic looms [4].Due to these challenges, Toyota further refined TPS through practical trial and experimentation in the 1950s and 1960s. Its founders realized early on that their people, especially their front-line workers, were most capable of learning, cre-ating, and problem-solving. For this reason, they were the most valuable resource and needed to be treated with respect. The value of Toyota’s front-line workers is emphasized by the fact that they are always referred to as team members. In a 1988 New England Journal of Medicine article [14], Donald Berwick, President Emeritus and Senior Fellow at the IHI and former Administrator of the Centers for Medicare and Medicaid Services (CMS), advocated that healthcare adopt the continuous improvement (kaizen) approach to healthcare, which engages people’s minds by applying the scientific method to problems. Suddenly, defects are positively looked at as opportunities to learn and improve rather than punitively as a way to identify potential “bad apples.”Steven Spears in The High-Velocity Edge [15] fondly described that “Toyota’s success is attrib-utable to its ‘velocity of discovery’ – the speed with which the company improves, innovates and invents.” Toyota’s founding fathers achieved this by “ensuring that pieces of a larger whole are har-moniously synchronized rather than discordant.” The downstream needs and processes paced work further upstream, creating the feeling of a syn-chronized orchestral piece with all units linked together to deliver the product or service to the end customer. The concept reduced wasted inven-tory and improved efficiency and quality. Toyota discovered how to do more work, rapidly and more reliably, without using more labor.The 1973 global gas crisis brought attention to Toyota Motor Company. They were producing high-quality, safe, small cars efficiently in the quantities needed by their customers with very little waste (muda), and remained financially sta-ble during this economic downturn. James Womack and his colleagues at the Massachusetts Institute of Technology (MIT) had been studying Toyota and published The Machine That Changed the World in 1990, which highlighted the suc-cessful principles of TPS and used the words “Lean production” to refer to TPS [16].Lean has taken on a wide range of meanings to different organizations due to the misunderstand-ing of TPS principles. Additionally, Lean can be unfortunately mistaken by the workforce to be a job elimination tool – a way to match staffing to hourly demand, sending people home early when deemed necessary, or to staff light daily [17]. This is contrary to the value that Toyota places on the development of its team members.Steven Spears and H. Kent Bowen [17, 18] describe four rules that need to be followed in the application of Lean principles which are congru-ent with TPS principles:• Rule 1: “All work is highly specified regarding content, sequence, timing, and outcome.”• Rule 2: “Every customer-supplier connection must be direct, and there must be an unam-biguous yes-or-no way to send requests and receive responses.”• Rule 3: “The pathway for every product and service must be simple and direct.”• Rule 4: “Any improvement must be made in accordance with the scientific method, under the guidance of a teacher, at the lowest possi-ble level of the organization” [18].The aforementioned rules have built-in signals to highlight problems automatically and rapidly and to make organizations adaptable to changing situations. These principles require organiza-tional commitment, but, when adhered to closely,

76will best align with TPS philosophy. This will become evident in the proceeding sections. The TPS TriangleThe TPS Triangle (Fig. 5.2) has been used to describe TPS outside of Toyota for greater than 25 years, which coincides with the founding of the Toyota Production System Support Center (TSSC). TSSC is a nonprofit subsidiary that shares TPS with people, companies, and nonprof-its outside of Toyota to contribute to society. At first glance, the simplicity of the TPS Triangle is evident. The people of any organization drive its excellence, so at its center is the emphasis on people development. TPS is an organizational culture of highly engaged people solving prob-lems or innovating to drive performance. This culture is created and sustained by a three-part system of (1) philosophy, (2) technical tools, and (3) managerial roles.The philosophy side has four key points: 1. Customer first – Understand the customers deeply and provide exactly what they want, only when they want it, and in the amount wanted. 2. People are the most valuable resource – Our employees, staff, and volunteers are our most valuable resource and should be engaged and treated as such. 3. Continuous improvement – The sum of many, many small improvements by many people accumulates to significant overall perfor-mance improvement and innovation. 4. Shop floor (gemba or genba) focus – Focus atten-tion on where the customer value-added work is done. In healthcare, this is typically in clinical areas, such as the emergency department, operat-ing room/theater, inpatient unit, or outpatient unit, but it does not need to be, as a project can extend into finance or other nonclinical areas.These philosophies also fit well for healthcare. First, customers are the priority. In healthcare, the most obvious customers are our patients and their families. We have other customers as well. A hospital unit or team member who receives a patient from another unit or team member is the customer of the upstream unit or colleague. No defect shall be passed on to the next customer. Customers can be internal or external to an orga-nization, including insurers. We should strive to meet or exceed our customers’ expectations. Customer and workforce safety are most impor-tant. For instance, an emergency department team needs to stabilize an ill patient to the best of their ability before admitting them to the inpa-tient medical-surgical or intensive care unit. They need to answer the questions of the patient and their family. Similarly, the accepting unit and medical team should expect a patient to be stabi-lized as much as possible, to receive a proper sign-out from the upstream team, and to have all of the needed chart documentation completed in a timely fashion. This allows for the excellent, team-based clinical care to continue and decreases the possibility of the patient becoming susceptible to a medical error. For this reason, we need to be cognizant of who our many customers are. To reiterate, we do not pass on defects to our customers as this creates customer and workforce dissatisfaction, increased costs from rework of defects, and potential safety problems.Second, people are the most valuable resource. Only people, after all, are capable of continual learning, especially problem-solving and inno-vating. For this reason, they must be treated as an organization’s most valuable resource and be provided a safe working environment, job secu-rity, intellectual challenges, and jobs that add value. The effectiveness and commitment of an organization depend on the motivation and capa-bility of its people. The role of management, or senior leaders, is to motivate and develop these frontline people. In healthcare, we generally think of our caregivers – physicians, nurses, and other allied health professionals who care for patients – as the core frontline people.Third, these motivated team members move forward to drive continuous improvement and associated problem-solving, also known as kai-zen, which occurs in small manageable steps. All team members come to work to both do and improve their work. Finally, key improvement

77activities occur on the shop floor with the follow-ing assumptions: (a) the shop floor (gemba or genba) is constantly changing; (b) one must be on the shop floor to understand the current state; and (c) the input from the members on the shop floor is invaluable to understanding the current conditions, feasibility of change, and goal of any change, and to set SMART (Specific, Measurable, Aggressive yet attainable, Relevant, and Time- sensitive) targets.By accepting that all humans learn the most by doing, organizational leaders need to provide frontline team members and managers opportuni-ties to learn, practice, and also fail. The role of a manager, as defined by the managerial side of the Triangle, is to engage and develop all team mem-bers into problem-solvers. On the technical side of the Triangle, team members use many TPS tools and methods (reviewed shortly in the TPS House discussion), to expose problems correctly. In the TPS culture, problems are also brought to the surface quickly as discussed in the initial vignette. We cannot solve problems we cannot see. In healthcare, organizations are investing in better training of their team members in improve-ment science methodologies to promote problem- solving as soon as a problem is identified. By teaching team members a common institutional standard way to approach problems, they have a common language through which they can imme-diately describe their initial problems, their ongo-ing progress, and resolution. This reason for the common language is no different than the reasons that have supported the need for common resus-citation methods, such as basic life support or advanced cardiac life support (BLS/ACLS), in the clinical setting or the use of the scientific method in the laboratory setting.Toyota places considerable value on customer input and satisfaction. They strive to provide cus-tomers with exactly what they want, when they want it. By encouraging patients or their families to provide feedback or speak up, health systems can design desirable services for their patients. By incorporating these family members into the discussion or improvement project involving the care of their loved one, more informed decisions can be made, increasing the likelihood that the project will be successful, lead to meaningful change, and ultimately increase patient and fam-ily engagement and satisfaction. Similarly, by encouraging families to initiate rapid response teams, problems can be brought to the surface sooner [19]. Rapid response teams are comprised of hospital team members that respond to the bedside of a patient with early signs of deteriora-tion in response to staff or, in some situations, family member concerns. This is also a perfect example of the application of the aforementioned HRO principles – preoccupation with failure, sensitivity to operations, and deference to exper-tise. Our frontline workers and families are very aware of the minute-to-minute changes in the clinical status of their loved ones.By now, Toyota’s obvious focus on connecting production to customer preferences and demand, and the focus on the development of frontline team members, is apparent. This reduces waste, promotes the rapid identification and resolution of problems, and ultimately creates a learning system.Key Learning Points 1. TPS is an organizational culture of highly engaged people solving prob-lems or innovating to drive perfor-mance. This culture is created and sustained by a three-part system, as described by the TPS Triangle, of (1) philosophy, (2) technical tools, and (3) managerial roles. 2. In healthcare, this culture must be a win for patients and their families, a win for caregivers, a win for hospitals, and a win for the community. 3. An organization’s people are the best learners and advocates that can help drive excellence. 4. Problems need to be brought to the sur-face quickly as we cannot fix things that we cannot see. Problems detected early are often smaller and more manageable. Missed problems, or delayed detection of problems, can permit problems to

78 Toyota Production System HouseThe Toyota Production System House (Fig. 5.3) depicts the key technical elements of TPS. These concepts will be discussed in detail followed by vignettes from various healthcare organizations that have applied TPS-based improvement sci-ence to local problems with direct guidance from TSSC.The TPS House is covered by a roof which represents the performance that TPS is designed to deliver – very high quality, low cost, and short lead (or wait) time. Safety comes above all else. To achieve high performance, there are two main pillars: just-in-time (JIT) and jidoka (building in quality at the source). JIT and jidoka require some foundational elements starting with the 4Ms. This stable foundation enables stable opera-tions. Specifically, the foundation requires Manpower (People),1 Machine, Material, and Method – which need to be of high quality and reliability, and properly chosen. Manpower (People) need to perform reliably with good work habits, proper skill level, good attendance, and low turnover. For instance, a common challenge in nursing and other healthcare roles is managing the rotating shifts over the 24 hours of a day, 7 days per week, and the associated turnover. Finding the correct people for these roles is crucial, as is assuring everyone is working to the top of their licensure and expected competency. 1 Manpower is mentioned here but this refers to humans of all genders.Machines need to be available in the right num-ber and location, and be reliable (not break down or create defects). This is especially true in criti-cal areas such as the operating room/theater or intensive care units where key machines such as ventilators must be dependable. Materials (such as references, standardized work documents, and manufacturer guidelines) need to be easily acces-sible to the people who do the work. Materials in healthcare also refer to the patient, their EHR, and their specimens. We want material to flow. The methods are the best, optimal practices for delivering care and services and are often the result of local continuous improvement efforts. At times, they can be best practices developed at other organizations but adapted and perfected locally through small tests of change. Often included in the foundation is the environment, which can also be referred to as Mother Nature (or the fifth M). The environment needs to be clean, clutter-free, and organized so that it can facilitate high quality, lower costs, and shorter lead times without the introduction of defects.The 4Ms foundation is required to support and enable the layer immediately above it, which is comprised of kaizen, heijunka, and standardized work.Kaizen refers to continuous improvement and problem-solving. TPS encourages continuous improvement since it is small steps of change that, when added together, can result in great innovation. Kaizen is the bridge that brings cus-tomers and improvement team members together. This has been especially impactful in healthcare when caregivers can experience their processes through the eyes of patients and their family members. Engaging customers/patients as we address imperfect processes helps to create an environment where respect for people, a key TPS concept, is realized. This concept is referred to as “humanize.” It helps reinforce why we need to improve and can help provide the motivation to support change remembering that 100% of what we do ultimately impacts our customers/patients 100% of the time.Heijunka refers to the leveling of work or pro-duction. By leveling work, you prevent process bottlenecks or the buildup of inventory in the evolve to those that are larger, less man-ageable, and detrimental. 5. Defects are not to be passed on to our customers, as this creates customer and workforce dissatisfaction, increased costs from the rework of defects, and potential safety problems. 6. Problem-solving is a crucial skill set. 7. The voice of the customer is important and needs to be incorporated into any improvement project.

79industrial setting. In the healthcare setting, you can distribute the work evenly so as not to over-burden any single person, preventing safety and quality issues. For instance, hospitals have applied this to their operating room scheduling process by distributing the types of cases evenly to the various operating rooms, optimizing work, and balancing the overall flow throughout the week.Standardized work (a step-by-step document written by the people who do the work outlining the current best thinking on how to perform the process, including step sequence and timing) needs to be defined to maintain changes. Standardized work at Toyota is a framework for maintaining kaizen improvements. Once the cur-rent practice is known, efforts are made to docu-ment and train to this standard until a better way is developed. When a better way is discovered, new work method standards are created. One key point is that standards are a starting point with the expectation that they will be improved. At times, healthcare providers are resistant to standardiza-tion without fully understanding that it is a start-ing point for the improvement process. ThedaCare in Wisconsin has therefore coined the term “flex-ible regimentation” where regimentation refers to the creation of a common standard process for “performing a specific service based on the best available evidence,” and flexible refers to the ongoing work to improve this standard [20]. Standardized work also reduces variation in sup-plies and instruments used in the operating the-ater since uniform predetermined supplies and instruments are used for each type of surgery among the various medical providers. As a result, it also plays a critical role in surfacing problems. When abnormal conditions occur, the behavior of following standardized work allows members performing tasks to identify problems rapidly.The two pillars of the TPS House are just-in- time (JIT) and jidoka. The JIT pillar advocates continuous flow, takt time, and the use of pull systems. Production is tightly run, where the key components reach an assembly line at the time needed and only in the quantity desired. Everyone in the production process works in sync and is aware of takt time. Takt time is cal-culated by dividing the operable time per day by the required number of units of a particular product per day (output). With a high level of JIT, any disruption to flow is immediately visi-ble, so immediate problem-solving can be initiated.Pull production is important to the concept of continuous flow. Toyota, from its earliest years, realized that extra inventory was disadvanta-geous. Toyota had very limited financial resources and space in its early years to afford the storage of inventory, so they had to be innovative and find alternative manufacturing solutions. Taiichi Ohno once said, “manufacturers and workplaces can no longer base production (from) desktop planning alone and then distribute, or push, their products onto the market. It has become a matter of course for customers, or users, each with a dif-ferent value system, to stand in the front line of the marketplace, and, so to speak, pull the goods they need, in the amount and at the time they need them.” This reduced inventory has given Toyota the ability to tightly regulate their processes to uncover defects when problems arise. The prob-lems become easier to find, and this, in turn, reduces problem-solving time. Ohno further elo-quently stated that the goal of “Toyota Production System is to produce what you need, only as much as you need when you need” [4, 13]. He realized that mass production without a linkage to the true customer needs would not work long term.Jidoka (automation with a human touch) refers to processes with built-in quality that immediately signal when a problem occurs, so that a person does not have to monitor a process just looking for defects. An everyday example of jidoka is the seat belt alarm that beeps whenever a seat belt is not properly fastened. Back when Toyota was originally an automated loom manu-facturer, jidoka referred to a loom’s stoppage if a string broke, alerting the worker, which, in turn, prevented the manufacture of cloth with defects. At a Toyota plant, jidoka may refer to a sensor that stops the line and brings attention to a defect or process abnormality that is then immediately rectified so that the defect or abnormality is not passed on further, leading to a larger problem. Any of the Toyota visual defect detection systems

80(andons), which can stop an assembly line at their plants, can facilitate the problem-solving process, since problems are immediately pin-pointed to their respective microsystem. These can also be manually pulled by a team member. The only reason to stop the line is to ensure that it will never have to be stopped again for the same circumstances. Quick fixes or stopgaps are never a solution. However, they may be tempo-rarily utilized when recurrence prevention takes time to complete.Andons and poka-yoke are important parts of jidoka. As mentioned in the introductory vignette, andons are tools for visual control. Poka-yoke refers to a built-in quality that prevents defects from occurring. In car manufacturing, parts may be created that only fit one way to prevent incor-rect assembly. At home, our riding lawn mowers automatically turn off if the rider gets off the seat while the mower is still running. Similarly in healthcare, the various anesthesia gas connectors only fit specific gas lines, thus preventing the accidental mixing of medical gases (Fig. 5.4). This type of human factors integration has elimi-nated the accidental fatal administration of gases other than oxygen during operative cases.In short, jidoka prevents the continued propa-gation of defects and reduces the chance that they will reach the customer, as well as signals prob-lems so that people can immediately investigate their causes to then devise improvements to pre-vent recurrence. For jidoka to succeed, much effort must be placed on work standards, as only once “normal” is defined and made visual can “abnormal” exist. This high level of standardized work has proven to be a challenge for many healthcare organizations. Additionally, the struc-ture of an organization must include people who respond quickly to an andon and have the time and mindset to solve problems, so they never recur. Without such a structure, andons will not be effective.Fig. 5.4 Use of poka-yoke to prevent the accidental mixing of anesthesia gases. Note each gas hosing and connector is different to prevent incorrect connections, which, in turn, can lead to an error and potential patient harm

81Looking back at the TPS House (Fig. 5.3), when both the JIT and jidoka pillars are balanced above the two lower levels of the TPS House, the roof of the house is level so that the House’s goal of producing high-quality and low-cost products with a short lead time can be met. JIT and jidoka both deliberately signal and highlight problems during operations. As these problems are solved to prevent recurrence, performance for safety, quality, cost, and lead time improves. Problem-SolvingAs previously mentioned, problem-solving through the use of the scientific method as part of kaizen is the essence of TPS. Clarifying and nar-rowing a problem is crucial, as represented by the funnel (Fig. 5.5). Problems are barriers to prog-ress for an organization but need to be antici-pated. There are some problems that require a deeper and more focused approach, such as the eight steps of problem-solving (Fig. 5.5) as described in Lean Hospitals: Improving Quality, Patient Safety, and Employee Engagement [17]. Yet others can be quickly resolved using a “just do it” approach.Generally speaking, the determination of what is a problem requires the definition of a standard of practice or care. Often when problems are uncovered, they are due to the following issues: there is no standard; the standard is not known; the standard is ignored; or normalized deviance results from standards not being completely fol-lowed. By creating standards and tracking the variations from the standard, the deviations are readily visible and can be targeted by Toyota’s “disciplined, yet flexible and creative community of scientists” [18] who help Toyota move toward a zero defect rate, similar to a health system’s analogous journey to zero harm. By having stan-dards in place, experiments, or rapid cycle tests of change, can occur to see if the standard can be improved further. However, without standardiza-tion, experimentation cannot occur in a way where its effects can be measured or appreciated.The eight-step problem-solving method (Fig. 5.5) breaks down problems through the use of a didactic approach in a manner analogous to the scientific method, which is only mastered through practice [17]. The eight steps are often captured on A3-sized (11-inch × 17-inch) paper, which forces teams to stay focused, concise, and Fig. 5.5 Eight steps of problem-solving. (Figure Courtesy of Eric Cardenas and adapted from Graban [17])

82simplify the problem. This A3 problem-solving document is portable, can be used to articulate the goals of the project and how they were devel-oped, and can become the expectation for all improvement projects. The most important part of the A3 is the problem-solving and continuous improvement thinking behind the template. While the A3 is a useful summary document, using the template enables and supports teams’ thought processes as they work through a system-atic approach to problem-solving rather than sim-ply filling in the boxes on a form.In his 2011 book, Thinking, Fast and Slow, Daniel Kahneman [21] describes how we as humans are wired for automatic, rapid interpreta-tion of input with little or no effort or voluntary control. Dr. Kahneman refers to this as System 1 thinking. In other words, we are quickly able to move from a problem to a solution. In healthcare, this thinking serves us particularly well in life-saving situations. However, not every problem we face in healthcare is a dire emergency. Many of the long-standing problems that we have been unable to solve in healthcare today require us to deliberately seek objective alternative interpreta-tions of data/events or what Dr. Kahneman refers to as System 2 thinking. Anyone who has been on the sharp end of “standard” solutions based on assumptions to problems (including endless e-mail reminders to “just be more careful,” count-less “read and sign” policy attestations, and redundant in-service education), as a means to “solve” the same issues over and over, can attest that there must be a better way. The eight-step problem-solving provides a structure supporting the System 2 thinking necessary to make sustain-able improvements that can transform healthcare. The eight steps using Plan-Do-Check-Adjust as a familiar framework are reviewed in Fig. 5.5; of note, Plan is inclusive of the first five steps of the eight steps.In step 1, the problem is clarified through fact- based quantifiable data. The current situation is compared to the ideal situation, and the gap is identified. In step 2, the problem is broken down into smaller concrete problems by asking the fol-lowing questions of the data: what, where, when, and who? When breaking down a problem, it is important to avoid “why” questions that prema-turely lead to root causes, as this can mislead-ingly stop the strategic breakdown of data. Usually, based on the frequency or relevancy of an occurrence, the prioritized problem is chosen. This point of occurrence is identified on the pro-cess map. This is confirmed by walking, or observing, the shop floor (also called gemba or genba) in a process called genchi genbutsu (to go look, to see, to understand, to take action). In step 3, we set a target for the prioritized problem which is measurable and concrete, yet challeng-ing. The SMART acronym is often used to lead teams through target setting. SMART stands for Specific, Measurable, Aggressive yet Attainable, Relevant (to the problem), and Time-sensitive. In step 4, the root cause is sought after by looking at all of the possible causes. Facts are gathered through genchi genbutsu and the “5-Why” approach is used to uncover the root cause. The 4Ms (Manpower, Machine, Material, and Method) can provide a structure when seeking root causes, and it can also ensure that the prob-lem is looked at systematically without prejudice. By purposefully asking “why” several times, and validating information through genchi genbutsu, facts are separated from opinions and assump-tions, thus resulting in true root cause(s). Most experts consider step 2 (breaking down the prob-lem) and step 4 (analyzing the root cause) crucial for problem-solving to occur.In step 5, many potential countermeasures need to be considered. A countermeasure is a set of actions that seeks to prevent the problem from arising again. Countermeasures are different from “solutions” that may just seek to deal with the symptom of the problem vs. the root cause(s). For every root cause, at least one countermeasure should be identified, understanding that one countermeasure may address more than one root cause.Countermeasures will need to be prioritized based upon costs, ease, feasibility, and other fac-tors. Countermeasures need to be in line with the ultimate goal and organizational priorities. These, in turn, are used to create a clear and concrete plan of action. Consensus needs to be reached around these countermeasures through discussions among

83stakeholders, especially those with upstream and downstream process owners, to ensure the imple-mentation of selected countermeasures will not negatively impact other processes.In step 6, efforts are aligned to implement countermeasures with speed and persistence. When creating the action plan, consider the following:• Who will be involved and affected (e.g., stakeholders)?• What is to be achieved and how will it be achieved?• When are potential completion times?• Where will the work occur?• Why is it important work?• How is it going to be messaged throughout the organization?Also consider all costs involved (e.g., poten-tial downtime, manpower hours). The improve-ment team’s efforts are messaged to the entire organization to inform and garner support. Monitor progress through the tracking of prede-termined metrics. Be persistent and in line with the aforementioned HRO principles. Multiple tests of change may need to occur before success is achieved. The value of the Plan-Do-Study-Act (PDSA), also known as Plan-Do-Check-Act (PDCA), cycle, which has been extensively dis-cussed in other chapters, cannot be understated to test countermeasures. Through the data-driven eight-step process, proper predictions for ideal solutions or countermeasures to problems are made, which consequently increases the likeli-hood that the ensuing planned tests of change (PDSA/PDCA cycles) will be successful.Step 7 emphasizes the importance of evaluat-ing results based on the SMART target set in step 3. Evaluate all results from the perspective of the customer, the team members, the organization, and society, seeking to understand the reasons behind the successes and failures. In addition, identify and celebrate potential return of investment(s), or ROIs. This can include cost sav-ings and immeasurable benefits such as people development, team engagement, and a renewed commitment to kaizen.Step 8 stresses the importance of standardiz-ing successful interventions and creating new standards. Share and spread the improved stan-dards with other parts of the organization or other organizations. Plan the next round of continuous improvement. By developing a standard method for problem- solving , through the eight-step process and A3 document, Toyota has created a procedure for communicating within a team and across its orga-nization. This method allows innovative solu-tions to spread across teams in a more understandable way. It incorporates PDCA/PDSA cycles for running small tests of change. This data-driven approach requires discipline and fact-based root cause analyses. The direction of an organization is not left to conjecture or the whims of a few strong personalities. In short, at its core, the Toyota Production System is:• An integrated approach to problem-solving that creates an organizational culture of highly engaged people, solving problems to drive performance. High levels of JIT and jidoka expose and signal problems to solve.• A way to achieve sustainable improvements that help foster a culture of continuous improvement and support the transformative change needed in healthcare.• An organizational culture created and sus-tained by a three-part system, as described in the TPS Triangle of (1) philosophy, (2) techni-cal tools, and (3) managerial roles.Key Learning Points 1. The implementation of TPS requires the creation of a stable foundation which incorporates the four (or five) Ms – Manpower (People), Machine, Material, Method (and Mother Nature). 2. The two pillars (just-in-time and jidoka) and all of the foundation levels of the TPS House need to be equal so that its roof can remain level and deliver high- quality goods at a low cost with short lead time. This emphasizes the impor-

84 Vignettes with Relevant DiscussionThe next sections describe vignettes from actual TPS-driven improvement projects from several health systems, followed by a discussion of the TPS concepts relevant to each vignette. The fic-tional patient cases are based on actual cases that have occurred at many hospitals but have been modified to protect the anonymity of each case.tance of all the components of the TPS House to achieve sustainable improvement. 3. The creation of standardized work is an important basis for measuring and driv-ing improvement. 4. Just-in-time (JIT) focuses on customer demand and refers to the production and conveyance/transportation of only what is needed, when needed, and in the quantity needed. It meets the exact demand of the customer in terms of product, timing, and volume. 5. Building quality into a process (jidoka), so that defects become readily visible, is crucial to uncovering defects (andon). This has proven to be difficult for healthcare delivery systems to install for a multitude of reasons. Facilitating problem detection is the best way to ensure its rapid resolution. 6. The most important part of the eight- step process is the problem-solving and continuous improvement thinking behind the template. While the A3 is a useful summary document, using the template enables and supports teams’ thought processes as they work through a systematic approach to problem- solving rather than simply filling in the boxes on a form. 7. Team member problem-solving skill development is critical and should be facilitated by all leaders and managers. These same leaders and managers need to be problem-solving experts themselves. 8. The shop floor (gemba or genba) is where all improvement occurs and, for this reason, local team members need to be incorporated into, and at times lead, improvement teams. Leaders need to visit the shop floor often to be visible to team members and better understand any problems they may face (genchi genbutsu).Vignette 5.1 Improving the Delivery of Critical Nutrition to Our Most Vulnerable PatientsA 500-gram baby boy is born prematurely at 25 weeks and is cared for by the neonatal intensive care unit (NICU) team. He cannot breathe on his own since his lungs are not fully developed, so he is intubated and placed on a ventilator. At this point, his odds for survival may not be good as a majority of his organ systems are not mature, especially his respiratory, immune, renal, and neurologic systems. His caloric expenditures are high and they will need to be continually replenished, as his energy reserves have not been built up. The baby is immediately started on intravenous fluids, and the decision is made to start him on total parenteral nutrition (TPN). The TPN is ordered at 11 AM and will be delivered in the evening. It will likely be hung at the patient’s bedside and the infusion started 9. The aforementioned summary points complement the definition of TPS. As a reminder, the Toyota Production System is an organizational culture of highly engaged people solving problems or innovating to drive performance. This culture is sustained by a three-part sys-tem, as described in the TPS Triangle, of (1) philosophy, (2) technical tools, and (3) managerial roles.

85A process that has taken hours, rather than minutes, can hinder a clinical team’s ability to render excellent care and meet the changing needs of a critically ill premature infant. TPN is produced with the hopes of mimicking the nutri-tional supplementation pathway available in utero from an infant’s mother. The members of the TPN process improvement team sought to improve the TPN ordering, production, and deliv-ery processes and reduce the time from TPN order to TPN infusion for an infant [22]. Figure 5.6 shows the process map from the order-ing to delivery of TPN for a single patient on a single day.There was considerable non-value-added time (or 80% waste) built into the original process, as shown by the areas in blue, red, and yellow. Figure 5.7 illustrates the different types of waste (muda). Figure 5.8 shows the same process after the various changes were implemented.Multiple changes were implemented. The TPN production areas were reorganized to maxi-mize efficiency using 5-S concepts (5-S = Sort, Set in order, Shine, Standardize, and Sustain; Fig. 5.9). Within the pharmacy, the technicians’ workflows were streamlined by placing supplies at the point of use, decreasing par levels (and therefore, on-hand inventory), and decreasing the automated TPN compounder’s changeover time by standardizing its setup and breakdown (Figs. 5.9 and 5.10). The latter was created by using a video to demonstrate the standard setup and breakdown procedures, and technicians were then trained to this standard. This training was routinely repeated to ensure that there was no normalized deviation from this standard.Within the NICU, medical team rounding, which involved the physicians, nurses, and phar-Fig. 5.6 Process map for TPN flow – initial state for a single patientby 9 PM, 10 hours after the order was placed and 17 hours after his blood was first drawn to assess the various serum electrolyte levels. The parents inquire whether this TPN delivery time is the norm, and they are told that this is, in fact, the case at most organizations.

86Fig. 5.7 The 7 + 1 types of waste. The seven (7) types of muda (waste) are motion, rework (that lead to defect), waiting, overprocessing, inventory, conveyance (transpor-tation), and overproduction. In healthcare, wasted time and potential of people is commonly referred to as the eighth waste, which includes the inability to support peo-ple to function to the highest of their licensurePre-improvement flow chartPost-improvement flow chart05:0011:0014:0016:0020:3004:5905:0011:0014:0016:0020:3004:59Start processStart processLabresultsreadyWaitingforTPNordersTPNorderswrittenTPNorderssent topharmacyTPN orderswaiting to be sentto pharmacy(batch deliveries)TPN waiting to bedelivered to NICU(batch delivery)TPN waiting to beadministered(infused)TPNcompoundedTPNdeliveredto NICUTPN administered(infused)LabresultsreadyWaitingforTPNordersTPNorderswrittenTPNorderssent topharmacyTPNcompoundedTPNdeliveredto NICUTPN administered(infused)Blooddrawn forlabsEnd processBlooddrawn forlabsEnd processFig. 5.8 Process map for TPN flow – pre-improvement (top) vs. post-improvement (bottom). The top figure shows the waste in the system pre-improvement. The red boxes in the first map indicate waiting time (representing waste or muda). The bottom figure represents TPN flow post-improvement

87macists, was standardized over all 7 days of the week. The goal was to have most TPN orders sent to the pharmacy for compounding before the end of the morning.Additionally, the daily TPN initiation times were changed, so this task fell to the day shift (7 AM–7 PM) team which, in turn, releveled the work (heijunka), since many more tasks tradi-tionally fell onto the evening shift (7 PM–7 AM) team. Job instruction sheets (JIS) were created to teach the day shift nursing team how to start the TPN infusion and related standardized work. Since all patients who were receiving TPN also had central venous catheters, the efforts to stan-dardize TPN delivery also required the creation of standards as to how TPN was infused using these central lines (Fig. 5.8). The JIS showed the “what” of each step in the “Key Point” column and the “why” behind each step in the “Reason for Key Point” column (Fig. 5.11). Additionally, to improve TPN delivery times, a TPN ordering software program was created within the elec-tronic health record (EHR) with built-in algo-rithms that prevented ordering errors. Since this program communicated directly with the TPN compounder, errors from the re-transcription of Standardization: Setup and useCompounder setup kitManual additives kitBefore• Products at point of use• Standardized inventory• Time managementAfterFig. 5.9 Standardization of the TPN materials and additivesStandardization (compounder setup)StartFinish• Enables continuous clean technique• Training video and standard operating procedure createdFig. 5.10 Standardization of the TPN compounder setup

88paper orders into electronic orders were elimi-nated. This is a great example of jidoka!The TPN delivery improvement project resulted in a 45% reduction in the average time of TPN delivery to the patient after the initial order. All previous transcription errors which resulted from the rework that had been part of the initial process were eliminated as well. By addressing the TPS House’s roof, its JIT (especially with the creation of a pull system) and jidoka pillars, lead time reduction and the maintenance of high qual-ity were achieved (Fig. 5.12).Fig. 5.11 Job instruction sheet (page 1 of 4) for TPN administrationKey Learning Points 1. Understand the current state of a pro-cess before implementing change, iden-tifying value-added and non-value-added times. Value-added time refers to time that improves a process and is important to your customers (patients, in this situation). 2. Organize the work areas to maximize efficiency while minimizing inventory. These concepts of organization are referred to as the five Ss – Sort, Set in order, Shine, Standardize, and Sustain. 3. Processes need to be designed to focus on the customer first, not what is easiest based on layout, machines, or old habits. 4. Standardization is a critical first step for quality improvement. Without a stan-dard, kaizen cannot take place. 5. Heijunka, or leveling of work, is critical to prevent team member burnout, improve patient safety, and improve efficiency. 6. Building in quality, or jidoka, as with the TPN compounding software pro-

892018-12017-92017-52017-12016-92016-52016-12015-92015-52015-1700600500400300200Year-monthAvg.time (minutes)__X=279.3UCL=297.6LCL=260.9BaselineINT 1INT 2INT 31Tests performed with unequal sample sizesAverage time lapse between TPN order written and start of infusionJanuary 2015 – January 2018Fig. 5.12 Average time between the TPN order written and start of the TPN infusion. The X-bar chart shown below displays the reduction in average TPN delivery times from approximately 500 to 280 minutes through the changes described in this vignette. Mean values for each phase are denoted by the green linesVignette 5.2 Improving the Pain Medication Reassessment Process in the Emergency DepartmentIt is the first Wednesday of the month, and Margo, the nurse manager for the emer-gency department (ED), is in her office pre-gram, can improve the ability to detect defects. 7. Use job instruction sheets (JIS) to share and teach standard work to frontline team members. The JIS show the “what” of each step in the “Key Point” column and the “why” behind each step in the “Reason for Key Point” column. The pictures provided in the “visual” col-umn provide further clarification and guidance for each major step.paring to meet with her supervisors. As she looks over the ED Quality Data Metric Report she just received, she shakes her head in disbelief. Once again, the ED is below the target for pain reassessment – a key measure of pain management for her department. Not just a little under the tar-get, data showed that only 55% of ED patients were being reassessed by their nurse timely (per hospital and regulatory requirements) after receiving pain medica-tions. “Barely half!” she exclaimed to her-self as she glanced back at the file cabinet that held all the pain management in- service education provided to each shift for the past six (6) months. She remembered that she even had kept a copy of the color-ful poster created by the unit secretary to remind staff of the importance of reassess-ing patients for pain – a staff member’s

90A pain reassessment team was formed and included Margo, the nurse manager; Dr. Beverly Chase, an emergency medicine physician; Randy, an RN (registered nurse); Lisa, an LVN (licensed vocational nurse); and Walter, the newly appointed team leader for this initiative. They called themselves, “The A-Team” and agreed to meet every Wednesday for 1 hour just before the weekly staff meeting. The following is a sum-mary of the team’s improvement efforts (Table 5.1) which walks through the eight steps of problem-solving (Fig. 5.5).This vignette illustrates the robust methodol-ogy outlined in the eight steps of problem- solving. Normally all eight steps are captured on a single A3-sized document, but were formatted here to meet the publication needs of this text-book. As mentioned earlier, the use of the A3 document as a standard permits easier communi-cation, idea sharing, and standardization and spread of successful change ideas across an organization.idea to help improve their compliance. “After everything we have done, how can our compliance be so low?” she thought to herself as her management team begins to fill the room for their weekly meeting.Per usual, each supervisor provided a brief update on their areas of responsibility. Walter, Margo’s newest supervisor, was just finishing his update when he shared a flyer from the Lean Department offering an A3 class. “I really am interested in taking this class. I just have to get your approval and bring a real problem for which we have data. I will need your help to identify a small team, including line staff that we can pull offline for 1 hour a week, dedicated to solv-ing the selected problem for the next few months. What do you think?” Margo sighed, “Here’s the most recent pain reassessment compliance data. Let’s do this – I’m all in!”Key Learning Points 1. The eight-step (A3) problem-solving process can be successfully applied to solve long-standing problems in healthcare. 2. Breaking down the problem using data (step 2) is key to helping the team pri-oritize and focus their improvement efforts on the most problematic area first. 3. In step 4, the 5-Whys analysis is used to arrive at the root cause. For instance, a team member asks “why” moving down the causal analysis tree to arrive at the root cause. To double-check the ratio-nale, one can state “therefore” to move upwards from the root cause. 4. Generally in step 4, we look for one root cause to a problem. In some cases, there will be a root cause with additional con-tributing causes. In these vignettes, the main and contributing causes are being classified as root causes for the sake of simplicity. 5. Developing standardized work is a criti-cal first step for quality improvement. Without a standard, kaizen (continuous improvement) cannot take place. 6. Building in one-piece flow into the pro-cess where patients are brought back into the FTA, stay until pain is reas-sessed, and re-medicated for pain if indicated, decreases the waste of motion and waiting for the patient and improves care and experience. 7. Commitment, support, and humility are modeled by the nurse manager as she encouraged the new supervisor to lead the improvement team, allocated dedi-cated time for team members to do improvement work, and supported the team by joining as a member and not the

91(continued)Table 5.1 Eight steps of problem-solving for Vignette 5.2Step 1: Clarify the problem – background data/informationPain is the most common reason patients come to the emergency department [23, 24]. Inadequate and untimely pain management in the ED is a global problem – despite the availability of resources, protocols, and effective interventions [25]Internal audits revealed that only 55% of ED patients were being reassessed within the appropriate timeframe (per hospital and regulatory requirements) after pain medication was administered Ideal situation: 100% of patients who receive pain medications (meds) are reassessed for pain timely in the ED Current situation: Only 55% of patients who receive pain medications (meds) are timely reassessed for pain in the EDThere is a need to improve pain reassessment in the ED which is a high patient volume unit with considerable acuity and, as a result, prone to problemsStep 2: Breaking down the problemBreaking down the problem: Ideal Situation:100% of patients arereassessed for paintimely after giving painmedication in the EDCurrent Situation:55% of patients arereassessed for paintimely after giving painmedication in the EDGap:45%What is thereassessment rate bytype of pain medicationroute?43% of patients whoreceived PO (oral) painmeds are reassessedtimely97% of patients whoreceived IM(intramuscular) painmeds are reassessedtimely99% of patients whoreceived IV(intravenous) painmeds are reassessedtimely97% are reassessedtimely in theTrauma ED91% are reassessedtimely in thePediatrics ED85% are reassessedtimely in theAcute ED16% are reassessedtimely in theFast Track Area (FTA)What is thereassessment rate forPO (oral) painmedications in eachED area?In the FTA, whendoes the problemmostly occur?14% of patientsare reassessedtimely in theAM Shift(First Shift)24% of patientsare reassessedtimely in thePM Shift(Second Shift)Identify the point of occurrencePatient checks inTriageTeamevaluatespatientFast TrackevaluationprocessstartsNurse givespain med topatientPatient issent to thewaitingroomNurseperformsmedication re-assessmentPatient issent tonextprocessPrioritized problemOnly 14% of patients who received PO (oral) pain medication in fast-track area (FTA) and who are sent back to the waiting room during the first shift are reassessed for pain timely (the ED staffing is traditionally based on two 12-hour shifts. The first shift refers to the 7 AM to 7 PM shift. The second shift refers to the 7 PM to 7 AM shift)

92Table 5.1 (continued)Current conditions/genba observationsTo validate the internal audit, the team conducted genchi genbutsu and observed FTA first shift’s processesThe FTA team observed:Manpower: Nursing staffing – i.e., registered nurses (RNs) and licensed vocational nurses (LVNs) – is frequently at a bare minimumMachine: No issues identifiedMaterial: The EHR alert for pain reassessment is set at 45 minutes, leaving the nurses only 15 minutes to complete the taskMethod: There is a misunderstanding on who should reassess a patient’s pain level. When RNs were asked, most responded that it is the responsibility of the nurse (RN or LVN) who gave the medications to reassess the patient, while most LVNs stated that it is the responsibility of RNs to reassess a patient’s pain level after pain medications are given Workflow – Below is the current process and diagram of the flow of the patient visiting FTA needing pain management: Task 1: Patient checks in at the router desk Task 2: Patient is called into FTA to be evaluated by the triage team Task 3: Patients are sent to either room 4 or room 5 to have their medical evaluation process initiated Task 4: During the evaluation of the patient, if pain medications are needed, they are sent to room 6 or room 7 to receive pain medications Task 5: After a patient receives pain medications, they are sent to the waiting room Task 6: If applicable, a patient is called from the waiting and directed to room 8 for further workup Task 7: Patients are called back to room 7 to obtain discharge informationDiagram of workflow:Step 3: Target settingImprove timely reassessments of patients who receive oral medications in the FTA during the first shift from 14% to 100% by December (3 months)

93Table 5.1 (continued)Step 4: Analyze the root or main cause(s)/identify root or main causes (RC)Why?ThereforeOnly 14% of patients who received PO (oral) pain medication in FTA first shift are reassessedfor pain timely.ManpowerMachineMaterialMethodNo designated“PainReassessmentRN” isavailable everyshiftNot enoughRNs to assign adesignated painreassessmentnurseX(Out of Scope)No machineissues identifiedXThe painreassessmentnursing e-charting(EHR) doesn’tenable timelyreassessment ofpatientsEHR alert is set at45 mins aftermedicationadministration(RC 1)Pain reassessmentis not part of aspecific nurse tasklist to complete(RC2)Not enoughrooms in FTA toroom everypatient after painmeds are given.(RC3)Patients are lostto follow-up forreassessmentNot clear who(which nurse) isresponsible forperforming painreassessmentPatients are sentback to thegeneral EDwaiting room afterreceiving painmedicationThe team identified three root causes (green boxes below, RC 1–3) by asking “why” the prioritized problem happens, followed by checking the logic of the root causes by stating “therefore” for each root cause up to the prioritized problem (Generally in step 4, we look for one root cause to a problem. In some cases, there will be root cause with additional contributing causes. In these vignettes, the mean and contributing causes are being classified as root causes for the sakes of simplicity)For this step, ask “why” as you move down the casual analysis tree and arrive at the root cause. One can double-check the reasoning by moving up, or backwards, on the tree and state “therefore”.(continued)

94Table 5.1 (continued)Step 5: Develop countermeasuresFor each root cause, the team develops at least one countermeasure, understanding that one countermeasure may address more than one root causeRoot Causes (RC)Countermeasures (CM)RC 1: EHR alert is set at 45 minsafter medication administrationChange electronic health record to support reassessment policyrequirements (CM1)Modify FTA reassessment workflow by developing standardizedwork (CM2)Reconfigure ED set up to allow patients to remain in the FTAafter pain medication administration (CM3)RC 2: Pain reassessment is not partof a specific nurse task list tocompleteRC 3: Not enough rooms in FTA toroom every patientStep 6: See countermeasures throughThe team needs to develop a plan for how the countermeasures will be implemented. The plan includes who will be accountable and when actions are to be completed:Countermeasures (CM)PlanWho?When?Change EHR to supportreassessment policyrequirements (CM1)Modify FTA reassessmentworkflow by developingstandardized work (CM2)and reconfigure ED set upto allow patients to remainin the FTA after painmedication administration(CM3)1. Change the EHR alert from 45 minutes to 30 minutes to allow more time for the nurse to complete pain reassessment.2. Meet with the EHR workgroup to gain consensus3. Implement changes to EHR4. Train staff and establish go-live date1. Dedicate a room for patients to remain in FTA post pain med administration for reassessment2. Validate process, update standardized work.3. Obtain equipment (e.g., 6 chairs)4. Infrom team–in all shifts–of new process5. Update standardized work and audit standardized work; provide updates at weekly visual management board meetingWWWWWWWWMK/WWBC/WWDRMK/WWBC10/110/1510/1510/1510/2510/25–10/3010/3010/3010/20

95Table 5.1 (continued)Step 7: Evaluate both results and processesCompliance to pain reassessment during the FTA first shift improved from 14% to 94%, and overall ED compliance improved from 55% to 90%. While significant improvement was achieved, the 100% target was not met.Barriers: Despite the new processes in place, dedicated time for improvement work (e.g., auditing of standardized work) became limited due to high staff turnover and slowed progress for a period of time.Some of the additional benefits and return on investment (ROI) ignited enthusiasm for a team-based approach to develop standardized work for other processes in the ED. The team was propelled to create workflows for other areas within the ED, including the transition of care to the urgent care center. Improved psychological safety (achieved 100% positive response in a staff engagement survey) and human development (staff acknowledging how their work, ideas, and involvement support the organizational goals) created a structure platform for weekly multidisciplinary problem-solving discussionDate:12/15Step 8: Standardize and spread successful processesPain reassessment data for FTA is displayed on a visual management board – now monitored and analyzed weekly for all shifts to ensure improvement is sustained and ideas to further improve are captured. Part of the huddle in front of the visual management board is the discussion of the instituted continuous observations of standardized work and how the process can be further improved.The new electronic health record alert change (from 45 to 30 minutes) was adopted throughout the healthcare network.Further, causal analysis is now performed on individual cases not in compliance, and results are communicated back to individuals and the team for further ideas to improve and for immediate recurrence prevention discussion. Improvements are shared within the ED and across the organization.Reflections: While the target of 94% timely pain reassessment was not achieved, the team became empowered and encouraged to exceed this target in the coming months – not only on the first shift but also on the second shift. Having current data displayed and readily available to all has ignited shared accountability and healthy competition to improve care and meet compliance across the unit.Date: Ongoing

96leader (HRO principle of deference to expertise). 8. The 4Ms provide an excellent frame-work for systematic root cause analysis. 9. Visual management boards help to provide a forum for communication of key performance metrics, building staff engagement and knowledge of departmental goals, and their individ-ual role in helping to meet them. 10. The implementation, standardization, and resulting spread of the new elec-tronic health record alert for pain reas-sessment were successful since it had first been tested on a small scale.Vignette 5.3 Improving the Clinic Cycle Time for Orthopedic PatientsA 38-year-old male motorcyclist (Mr. M) was brought to the ED after he was acciden-tally hit by a car. His chief complaint was that his right wrist was painful. The patient stated that, when he fell off his bike, he landed on his right wrist. Diagnostic tests were performed, and the orthopedics trauma team (abbreviated ortho trauma) was con-sulted. Based on the X-ray, the patient was diagnosed with a new acute distal radial bone (wrist) fracture. Ortho trauma stabi-lized, reduced, and splinted the injured wrist. The patient was sent home and was instructed to go to the ortho trauma clinic the next day (Monday) when it opened at 7:30 AM to be seen by a hand specialist.Scrambling for transportation, Mr. M had to take two separate buses to make it to the clinic by 7:30 AM. On the way, Mr. M called his boss to let him know what hap-pened to him and that he would be into work immediately after his appointment was finished. The clinic was packed with patients – all with some sort of cast or ban-dage on one limb or the other. At 7:45 AM, Mr. M was relieved when his name was called, and he was escorted back into an exam room by a nurse. After asking him a few questions, and performing a brief assessment, the nurse informed Mr. M that the hand specialist team would soon be reviewing his case and would be in as soon as possible. After about 30 minutes of wait-ing, Mr. M fell asleep in the chair, exhausted for having spent the entire evening in the emergency department the night before. He was awakened a few times as the nurse re- entered the exam room to check on him and, each time, she reassured him he would be seen as soon as possible. Around noon, Mr. M peaked his head out the door asking the nurse for directions to the nearest rest-room. When he returned, the nurse informed Mr. M that he may want to get something to eat in the cafeteria as he most likely would not be seen until after 1:00 PM. Hungry, tired, and frustrated, Mr. M left the clinic, quickly ate, and called his boss to let him know he still hadn’t seen the doctor yet and would most likely not make it into work at all. Mr. M returned to his exam room at 1:00 p.m. as instructed by the nurse. Mr. M was seen by the hand specialist at 1:30 PM – 6 hours from the time he arrived at the clinic that morning! While relieved when he was informed by the specialist that he did not need to have surgery on his wrist, Mr. M couldn’t believe he had lost a whole day of work – a day he wasn’t going to get paid for. Mr. M was given clinic discharge instruc-tions by the nurse that included a follow-up appointment the following Monday. As he left the clinic at 2:15 p.m. he wondered, “How am I ever going to pay for all this and get my bike fixed? I have to work – I just can’t take another day off from work to sit here all day.”Table 5.2 walks through the 8 Steps of Problem-Solving for this vignette.

97Table 5.2 Eight steps of problem-solving for Vignette 5.3Step 1: Clarify the problem- background data/informationThe organization’s guiding principles include putting patients first, improving their experience, and respecting people. In the ortho trauma clinic, patients report staying for their visit longer than anticipated. Along with overall patient experience, “perceptions of information, instructions, and the overall treatment provided by physicians and other caregivers” are also negatively influenced by prolonged clinic times [26].An internal review of data revealed that some patients who were scheduled for morning appointments were staying until later in the afternoon. The clinic termed these patients “carryovers” – patients scheduled in the morning clinic with cycle times so long they were “carried over” to the afternoon clinic, negatively impacting the clinic cycle time throughout the entire day.As part of an overall improvement initiative to reduce the ortho trauma clinic cycle time (time from when patients are checked into the clinic until the time they are discharged), the team prioritized understanding and reducing the number of “carryovers” (high volume, high risk, and problem-prone).Ideal situation: 0 patient “carryovers” from morning appointments to the afternoon clinicCurrent situation: In a month, 107 patients were unable to be seen timely and were “carried over” to the afternoon clinic.Step 2: Breaking down the problem and method of measurementsBreaking down the problem Ideal situation: 0 morningclinic patients: carried over’to the afternoon clinicCurrent situation: In amonth, 107 patients withmorning clinic appointmentswere not seen timely andwere ‘carried over to theafternoon clinicGap:107/monthWhen do most ‘carryovers’happen?On Mondays, Which clinictype has the most‘carryovers?’53 patients onMondays22 patients onWednesdays17 patients onTuesday8 patients onThursdays3 patients onFridays49 patients fromOrtho Trauma4 patients from OrthoSpecialty Hand0 patients from OrthoSpecialty Pre-OpJointsIdentify the point of occurrenceAppointment bookedPatient checks infor theirappointmentPatient completesdiagnostic testingPatient is roomedby the nursePatient is seen bythe providerPatient isdischarged by thenurse/sent home(continued)

98Prioritized problem at the point of occurrenceIn a month, 49 ortho trauma clinic patients who were booked for a Monday morning appointment did not complete theirclinic visit by noon (49 Monday morning appointment visits were carried over to the afternoon clinic).Current conditions/genba observationsBased on the prioritized problem, genba observations were conducted by ortho trauma clinic staff for the next two Monday mornings and found the following to be consistent:Manpower: There are enough providers and nursing staff to see patientsMachine: No issues identifiedMaterials: Trauma patients, who come in through the emergency department (ED) and urgent care clinic (UCC) over the weekend, are told to report to ortho trauma clinic at 7:30 AM Monday for follow-up (patients are booked and overbooked into the morning appointment slots). This was identified when the team visited the ED genba Hand patients are prioritized and are traditionally given the earliest morning appointments. This was also learned from the visit to ED genba A list of patients with non-emergent hand injuries is created by the trauma consultant. The list prints out at 8:00 AM in the ortho trauma clinic on Mondays Each case is reviewed by the hand specialist team to determine the treatment plan and whether surgery will be required when they meet from 8:00 to 9:00 AM on Mondays If the hand specialist team determines a patient will need to be scheduled for urgent surgery, the patient is prioritized to be seen regardless of their clinic appointment timeMethod: ED and UCC traditionally book and overbook all non-emergent trauma cases from the weekend to early Monday morning appointments. This was garnered from ED genba and validated in the ortho trauma clinicStep 3: Target setting: set *SMART goalDecrease the monthly number of ortho trauma clinic patients with Monday morning appointments who are carried over to the afternoon clinic by 50% (from 49 to ≤25 patients) in 1 monthTable 5.2 (continued)

99Step 4: Analyze the root or main cause(s)/identify root or main causes (RC)ThereforeWhy?In a month, 49 ortho trauma clinic patients who were booked for a Monday morning appointment did not completetheir clinic visit by noon (49 Monday morning appointment visits were carried over to the afternoon clinic)ManpowerxThere areenoughproviders &nurses to seepatientsNot validatedin the genba.xThere are noequipmentissues identified.There are not enoughexam rooms for Traumapatients beginning at8:00 AMPatients who areroomed first (at 7:30AM) are still occupyingthe majority of theroomsPatients are waiting tobe seen by the HandSpecialist TeamThe Hand SpecialistTeam is not available tosee their patients untillafter 9:00 AMBetween 8:00 AM and9:00 AM the HandSpecialist Team isreviewing their newcases from a consultlist generated over theweekend.The consult list for handtrauma cases is notavailable until 8:00 AMon Monday Mornings.(RC1)Patients whocome in on theweekend needto be seenxOUT OFCLINIC SCOPECertain patient’s surgeryis more urgent in naturexOUT OF CLINICSCOPECurrent Mondaymorning clinicscheduletemplate doesnot matchcustomer/patient(RC2)There are too many overbooked patientsfor Monday morning clinicOlder practicein the ED &Urgent Care(UC) areas ofbooking andoverbooking allnon-emergentortho traumafollow-uppatients intoMondaymorningappointments(RC3)MachineMaterialMethodBased on the resultsof reviewing theconsult list, patientswho need surgery areseen first regardlessof appointment orroom-in timeThe team identified three (3) root causes (green boxes below) by asking “why” the prioritized problem statement is happening, followed by checking the logic of the root causes by stating “therefore” for each root cause up to the prioritized problem. (continued)

100Step 5: Develop countermeasuresRoot causes (RC)Countermeasures (CM)RC 1: The consult list is onlyavailable to the hand traumaspecialists on Monday morning at8:00AM• Increase the number of times the hand trauma specialist team is given the consult list in a week (CM1) and standardize expected time (i.e., before 7:30AM) (CM2)• Revise hand trauma specialist appointment time slots to 8:20• AM(CM3) Standardize completion time of trauma specialists’ review of the consult list (i.e., by 8:20AM) (CM4)RC 2: Current Monday morning clinicschedule template does not match customer/patient demand• Identify types of appointments that are getting overbooked (CM 5) and create a new appointment types (e.g. Trauma Hand) to match patient/customer demand (CM6)RC 3: Older practice in the ED &Urgent Care (UC) areas of overbooking patients into Mondaymorning slots• Provide copies of new schedule to areas who schedule patients (CM7)Table 5.2 (continued)

101Step 6: See countermeasures throughThe team needs to develop a plan for how the countermeasures will be implemented. The plan includes who will be accountable and when actions are to be completed:Countermeasures (CM)PlanWho?When?Increase the number of times thehand trauma specialist team isgiven the consult list in a week(RCl/CMl) and standardizeexpected time (i.e., before 7:30AM)(RC1/CM2)Revise hand trauma specialistappointment time slots to 8:20AM(RC1/CM3)Standardize completion time oftrauma specialists’ review of theconsult list (i.e., by 8:20AM)(RC1/CM4)1. While EHR restructures electronic consult process, consult list will be sent by the Ortho Trauma team by 07:30AM2. Notify hand team of new changese1. Communicate go-live date (8/23) of updated template to team2. Reschedule current patients who still have 7:40AM appointments and inform patient of rationale of changes (i.e., to decrease waiting times)1. Reach out to services and areas who schedule non-emergent hand trauma appointments (i.e., the emergency department, urgent care, other clinics)2. Provide copies of new schedule (both printed and electronic copy)3. Develop electronic calendar system to improve sharing of future updates1. Follow up with scheduling department on request for updated EHR appointment type (to add “Ortho Trauma Resource: Hand Trauma) based on data2. Reallocate Trauma Clinic resources (i.e., 12 appointment 20-min slots, 2 rooms, manpower) to Hand Trauma based on data3. Schedule and commnunicate go-live date (8/23) of updated template to team1. Communicate updated template to Hand team2. Revise schedule of Hand team to be at the clinic in time to review consult list starting at 7:30AM.Dr. H8/22Dr. H8/20Dr. M8/23Dr. M8/20NP8/20Identify types of appointments thatare gettiing overbooked (RC2/CM 5)and create a new appointmenttypes to match demand (RC2/CM6)Provide copies of new schedule toareas who schedule patients(RC3/CM7)Step 7: Evaluate both results and processesThe monthly number of morning ortho trauma clinic patients carried over to the afternoon decreased from 49 to 21 patients – a 57% improvementSome of the additional benefits and return on investment (ROI): Mondays’ clinic cycle time decreased from an average of 3 hours and 36 minutes to 2 hours and 15 minutes, which is a 37.5% improvement Several staff ideas for improvement continue to be discussed and small tests of change are initiated during weekly team meetingsDate: 9/30Step 8: Standardized successful processesCarryovers are tracked and analyzed daily (through ortho’s visual management board). The team developed routine data-based discussion (at least weekly) with the ortho clinic team members to analyze and address clinic issues to make small incremental improvements each day to optimize capacityReflections: While the 57% improvement is a positive change, the team is looking forward to their next PDCA to continue to level out the clinic schedule and daily operations to achieve their ideal state to see all their patients in a timely mannerDate: On-going

102Vignette 5.4 Reduction of Pressure Injuries in Patients and Days Away, Restricted, or Transferred (DART) Days in Their ProvidersAn 8-year-old complex medical needs patient born with a large omphalocele (open abdomen associated with a chromo-somal defect during prenatal development where parts of the intestine and liver grow outside the abdominal cavity) was placed on mechanical ventilation in a pediatric intensive care unit as part of the postopera-tive clinical pathway associated with her plan of care following the surgical reduc-tion of an intestinal obstruction secondary to adhesions. The clinical team managing the care of this patient was afraid to turn her to the lateral or prone position for fear of disrupting the recent repair. Given the patient weighed 30 kg and had multiple attached devices, including monitoring equipment, this patient could not be turned by a single staff member without risking employee back injury. The nurse completed her Braden Q assessment just after the beginning of her shift (8:00 AM) but scored the patient a 22 (low risk of developing a pressure injury on a scale of 0–26), not rec-ognizing the high risk due to the patient’s immobility. The nurse is called to the care of another patient in respiratory distress and does not complete her head-to-toe skin assessment on this patient. The other patient is finally stabilized, and the nurse begins to document her care of both patients in the EHR, noting that it was now 4:39 PM and that she was administering scheduled medications for her patients. The nurse ends her shift at 7:00 PM, and the oncoming nurse completes a head-to-toe skin assessment along with the Braden Q assessment. She finds an advanced (stage 3) pressure injury (PI) on the patient’s occiput and proceeds to treat the patient based on recommendations from the wound Key Learning Points 1. The TPS eight-step process and the problem-solving thinking of the A3 pro-cess can be successfully applied to out-patient clinic problems by an interdisciplinary team of clinicians and surgeons, nurses, and ancillary staff with administrative support. 2. The balancing of the clinic schedule by designating specific hand clinic slots to match customer demand, and revising/streamlining the hand specialist work-flow, eliminating unnecessary batching of case reviews, is an excellent demon-stration of heijunka (level loading or balancing of the workload). 3. The importance of genba (shop floor) and genchi genbutsu (go look, go see, to understand and take action) was demon-strated especially in steps 2 and 4 (breaking down the problem and root cause analysis) of the eight-step pro-cess. Data told the team that “trauma” patients were the most problematic. However, through genchi genbutsu the team was able to go beyond the avail-able data. Through direct observations in the genba, the team identified that patients required to be seen by hand trauma specialists within “trauma” were the early morning bottleneck which led to long waits for all trauma patients and a significant factor in causing carryovers. 4. The importance of involving a team of experts, including members from other areas of the hospital (ED) and reinforc-ing the HRO concept of deference to expertise, is highlighted in the vignette. 5. The changes implemented by the ortho team led to a significant reduction of muda (waste) for patients (waiting) and clinic staff (rework of having to recheck on patients multiple times).

103Members of the pressure injury reduction team sought to address the number of PIs that were developing across the organization. They used the eight-step problem-solving methodol-ogy (Fig. 5.5) and quickly walked through the various steps of identifying the gap in perfor-mance, breaking the problem down to a manage-able scope. Figure 5.13 shows, by breaking down the problem (step 2 of the eight steps of problem- solving; Fig. 5.5), that Unit I had the highest occurrence of PI. By carefully examining the PI cases in Unit I, they found that PIs developed in patients that were not turned regularly and in patients with multiple devices used for complex medical treatment. In accordance with step 3 (tar-get setting), they chose to address the patients who were inconsistently turned. While focusing on that cohort of patients, they uncovered several staff injuries related to lumbar strain and were able to use data gathered from direct observation, the EHR, and occupational health to identify that the problem was bigger than initially anticipated. This team set a target (step 3 of the eight steps of problem-solving; Fig. 5.5) of reducing the occur-rence of PIs located on the occiput of patients. While working through the 5 whys (step 4 of problem-solving; Figs. 5.5 and 5.14), the team found that patients were not turned when two things were present: (1) the lack of perception of the risk of the patient’s ability to develop a pres-sure injury (identifiable when the Braden Q Scale is used appropriately) and (2) the patient was per-ceived to be too heavy to turn alone. As the first countermeasure, the team worked with frontline staff to create a simplified standard for assessing patient risk of developing pressure-related inju-ries by simplifying the verbiage of the Braden Q Scale (a risk assessment tool used to identify patients at risk for developing pressure injuries, where the lower the score, the higher the patient’s risk of developing a PI). Using the wound ostomy team (WOT), they tested inter-rater reliability between frontline staff and the WOT using the newly developed modified standard tool for assessing risk, the modified Braden Q [27] (Fig. 5.15).Once the gap between the WOT and the frontline staff’s Braden Q Scale results was narrowed, the team turned their focus to creating a standard used to train staff on turning patients alone, using a safe and simplified method (countermeasure 2). All Unit I staff were trained using JIS, and the standard was maintained using random audits by peers, WOT, and local leadership. As a result of staff using the new standard for turning patients alone, employee ostomy nurse. After applying the pre-scribed treatment for the PI, she remembers that her coworker injured his back turning a similar patient 2 weeks prior and has yet to return to work. She asked the charge nurse for assistance to turn the patient.89 Pressure injuries(11 month time period)Moderate harm(DTI, stage one, stage two)(69)Severe harm(stage 3, stage 4, unstageable)(20)Unit A(1)Unit B(1)Unit C(2)Unit D(2)Unit E(2)Unit F(7)Unit G(8)Unit H(9)Unit I(37)Fig. 5.13 Step 2 of problem-solving – breaking down the problem

1049 hip, trunk andgluteal regionpressure injuriesPatient notturned in a timelymannerPatient’s notbeing turnedcorrectlyPatients tooheavy; caregiverneeds assistanceto turn patientsCaregiver notavailablePatients areasleepNo dear standardfor turning heavypatientsNo set time forturning patientsToo many ways toturn a patientNo clear standardon how to turnthe patientToo manysupplies to utilizeNo clear standardfor supply useThereforeWhy?Fig. 5.14 Step 4 of problem-solving – root cause analy-sis using the 5 whys. The question “why” is asked repeatedly to arrive at the root cause(s). The lowest green boxes represent the root causes. To double-check the analysis, “therefore” can be applied as shown in step 4 of Vignettes 5.2 and 5.3282420Average Braden Q score16128412*WOT = Wound ostomy teamBaseline Data Source: Electronic Health RecordIntervention 1: Use of a modified Braden Q scale3456Days of pilot study78910Avg. staff Braden Q (baseline)Avg. *WOT Braden QAvg. staff Braden Q (INT 1)Braden Q score pilot testing2 week testing period(Before)averagescore of 22(After)averagescore of 18Fig. 5.15 Narrowing the gap between staff and WOT Braden Q scoring after training

105injuries measured by the number of days away, restricted, or transferred (DART) has been reduced.This vignette demonstrates the challenges of patient care. Ill and immobile patients are prone to PI, which often are subtle before they become larger and more obvious. The TPS encourages the creation of processes that bring these problems readily to the surface. Unlike the automotive assembly line, it is difficult to create processes that automatically uncover a PI and stop hospital pro-cesses, as described earlier for the jidoka pillar. In healthcare, there is value in setting up auditing processes and assigning accountability to identify PI in a more timely fashion, such as regular clini-cal skin assessments and creation of wound care teams that routinely audit at-risk patients.Similarly, prevention strategies are helpful. The team’s problem-solving exercise revealed that PIs were related to the absence of standards on how to take care of at-risk patients, especially the use of standard preventive methods (e.g., regular patient turning) and bedding materials (those that would facilitate turning or reduce pressure on at-risk body surfaces). Frontline team members had not been trained to adequately assess a patient’s skin to detect and classify these pressure injuries as they occurred, so they were trained by the WOT. As shown in Fig. 5.15, the frontline staff responded well to their training on the use of the Braden Q assessment tool, so much so that their assessment scores nearly mirrored those of the expert WOT (compare days 1–5 vs. days 6–10).The problem-solving exercise (Fig. 5.14) also revealed that heavy patients posed a challenge to the staff with regard to turning. The use of the Turn and Positioning System (TAPS) (Fig. 5.16) enabled patients to be turned with minimal risks to the frontline staff. This equipment was stocked on all units that cared for heavier patients, and front-line staff members were trained to the newly cre-ated standard using a JIS (Fig. 5.17) and simulation (Fig. 5.18), which improved their ability to care for patients at risk for PI through the implementa-tion of prevention and early detection strategies.Countermeasure: Standard supplies• 1 Taps• 2 Z-FloLarge Z-FloLarge Z-FloAnchor wedgeLumbar wedgeTurning and positioning system(TAPS) with wedgesFig. 5.16 Standardized supplies to enable easy turning (Z-Flo pillow, wedges, and TAPS) of patientsKey Learning Points 1. The phrase “if the student has not learned, then the teacher has not taught” emphasizes the value of teaching a stan-dard process or method. If frontline team members are not taught a standard and the learning is not reinforced, varia-tions in practice as well as normalized deviation will occur. This can have dev-astating consequences. 2. The creation of standardized work, which emphasizes best practices and the use of JIS to teach to the standard, can help improve outcomes. 3. The eight steps of problem-solving, when done properly, can discover hid-den root causes. 4. In healthcare, the use of timely, unbi-ased, and robust auditing processes can be an alternative to jidoka, which is used extensively at the Toyota manufac-turing plants.

106Fig. 5.17 Job instruction sheet shows the use of TAPS to position the patientUsed medical simulationFig. 5.18 Using simulation to teach the new standard described in the job instruction sheet

107 Building a Successful TPS CultureBuilding a TPS culture takes planning, consider-able culture building, and training, similar to our medical education processes. For instance, when physicians, nurses, and other allied health profes-sionals are trained, they go to a school where they are taught key concepts and fundamental princi-ples. While in school and during their internships and residencies, they are exposed to practical concepts and procedures and learn through obser-vation while under the guidance of their teachers or coaches. During these training years and early part of their careers, they are paired up with good coaches who provide continued guidance. Learning TPS is no different (Fig. 5.19) in that most learning is by doing, or practice, under the auspices of a good coach. There are three phases – education by concept, exposure/observation, and practice with a good coach. Key to this success are good coaches who can provide guidance to TPS teachings. These phases can occur in one of the two likely ratios (10:10:80 or 20:20:60). In other words, 10–20% of TPS can be learned with formal training/classroom exposure and 10–20% from exposure or seeing TPS in action. However, the greatest learning is from hands-on experience or direct involvement with team problem- solving, a key aforementioned point from the teachings of Taiichi Ohno – learning by doing. Compared to medicine, TPS concepts are relatively simple. In fact, they are so deceptively simple that people sometimes skip the learning by doing.As with any project dealing with change, the goal is to start with small tests of change. TSSC also embraces the model line concept, where building the TPS culture should first occur within a single service line or program. That single area is developed fully to the point where it can serve as a model of successful TPS implementation for others within a system to look to for advice, sup-port, and leadership. The newly trained unit members and leaders can also be redeployed to coach similar improvement projects elsewhere in the organization.Chandrasekaran and Toussaint [28] recently described a set of best practices that can help sustain a TPS culture within a health system. First, instill TPS behaviors in managers at all levels of the organization. Senior leaders need to be present and visible at regular intervals in the various organizational huddles. There will 80%10%10%Practice with a good coachExposure and observationEducationby conceptPractice with a good coachExposure and observationEducationby concept60%20%20%Alignment of Learning TPS with Medical EducationLearning TPS(10:10:80)Learning TPS(20:20:60)Fig. 5.19 Three phases of TPS training and implementation and their alignment with medical education

108need to be succession planning for the senior leaders, especially the CEO and various board members, with specific preference to those who understand and embrace TPS. Stories of suc-cess need to be created and shared. Finally, the quality and cultural management system needs to be a TPS-based operating system. All of these aforementioned concepts will permit problems to come readily to the surface to be resolved in a timely fashion, since problem-solving is part of the daily culture and expectation.The application of TPS principles to a health-care organization requires a new mindset that might at first appear foreign, especially with regard to the role of leaders. Kim Barnas, while a senior leader at ThedaCare in Appleton, Wisconsin, best described this mindset or busi-ness improvement system as comprising of eight key elements which are similar to TPS or their lean principles [29]: 1. Status reports – local daily dialogues that occur throughout the organization which enable situational awareness. 2. Daily team huddles – enable teams to discuss opportunities for improvement, challenges, and ongoing improvement projects. 3. Managing or auditing to the established standard. 4. Problem-solving. 5. Transparency – defects and problems are brought forward along with accomplishments. 6. Advisory teams – advisors comprised of team members or leaders from across the organiza-tion are available to individual units to pro-vide knowledge and expertise where needed. 7. Scorecard – tracks actual monthly perfor-mance metrics against goals. 8. Leadership standard work – leaders round regularly and set standard work expectations for all team members, including themselves (see reverse fishbone diagram; Fig. 5.20).Leaderstandard workVisual controlProblem solvingcorrective actionLeadershipdisciplineLeaders havea frameworkby which theycan simply andeasily managetheir business,develop people,solve problems,and improveperformanceWHAT: Driving balancedperformance improvementat all levelsWHAT: Metric alignmentwith strategy deploymentWHAT: Managing toStandard WorkWHAT: Manage flowWHAT: Using a standardproblem solving processWHAT: Eye for improvementWHAT: Visual workplaceWHAT: Visual managementFig. 5.20 Reverse fishbone diagram that depicts the role of any leader. This is a cause-and-effect diagram [30] in which the desired leadership outcomes were first defined and then the actions needed to create the effect were tested and implemented if successful. (Reprinted with permis-sion from Kim Barnas [29])

109The latter point emphasizes the role of any leader in an organization committed to adopting TPS. Every organizational leader has a structured day which begins by assessing and understanding the current state and anticipating problems (Fig. 5.20). The goal is to move from a “firefight-ing” mentality to an anticipatory focus where problems can be solved before they become criti-cal. Leaders need to become more visible, more respectful, actively supportive of the organiza-tion’s improvement initiatives and daily work, and process improvement focused. As discussed in the Bundles and Checklists chapter (Chap. 13), some hospital leaders use kamishibai cards (K-cards) as rounding tools to improve compli-ance with best practice bundles. Problem-solving is everyone’s responsibility. Teams work together to solve problems using the scientific method and leaders encourage and facilitate this. All improve-ment projects must be aligned with corporate goals which are rigorously reviewed annually. Summary and Closing DiscussionThe application of the Toyota Production System to healthcare is a recent development. While there are many differences between manufactur-ing and healthcare, we believe that the principles are applicable. The TPS culture can be invaluable when properly inculcated into the daily main-stream operations of an organization and can especially assist with its cultural and quality transformation. TPS is, after all, an organiza-tional culture of highly engaged people solving problems to drive performance that is created and sustained by the three-part system described by the TPS Triangle (Fig. 5.2).Many healthcare organizations have started the “mindfulness” journey to becoming a high- reliability organization – and as a result, improve their quality and safety outcomes. Weick and Sutcliffe [3] referred to mindfulness as the qual-ity of attention. The agility needed to address the ever so changing opportunities, and threats fac-ing mindful organizations and their team mem-bers, is due to the constant refinement of existing expectations, continual improvement of cogni-tive foresight, and rapid learning from events as they occur. Toyota is one such mindful organiza-tion where their journey to sustained excellence has occurred through careful planning and the focus on the development of its team members – one member at a time. They have realized that the HRO journey takes time, may have occasional setbacks, yet have processes in place that pro-mote resilience. They have created a successful organizational culture that they have to reinvigo-rate every time a new plant opens or its team members retire or transfer. Toyota has accentu-ated the value of continuous process improve-ment and the related problem-solving. It has integrated the principles of the TPS Triangle (Fig. 5.2), as well as the technical tools described in the TPS House (Fig. 5.3), throughout its global operations.TPS requires senior leadership team and man-agement commitment and visible participation, especially with the modeling of desired behav-iors, new habit formation, problem-solving skills for all, and all of the HRO principles mentioned earlier in this chapter and throughout this text-book. The power of TPS is in the method which mandates constant demonstration of competence through the application of learned principles, par-ticipation in improvement projects, and account-ability for personal growth and that of your respective teams. In short, TPS leaders are visi-ble, known to all, and enable the success of their teams. They are lifelong learners, teachers, and coaches.Process visibility is also crucial. Only if the current state of pre-existing processes can be defined can problems be brought to the surface and processes improved, ultimately leading to better outcomes. Problem-solving is everyone’s responsibility, as is the resulting shared learning. Taiichi Ohno coached his disciples by drawing a chalk circle onto the floor (often referred to as Ohno’s Circle [31]) and then asking them to stand in it and thoughtfully observe the actual processes on the shop floor. His disciples then reported on the various problems observed and were asked to use data-driven and observation- confirmed problem-solving to arrive at solutions. Data was collected through simple observations

110initially. Later, more complex data collection was made possible from the various automated jidoka tools that had been implemented on the shop floor. The resulting problem-solving occurred quicker. The value of each team member’s learn-ing by doing cannot be overstated.As a corollary, healthcare teams are inappro-priately focused on the unavailability of auto-mated data rather than embracing the value of collecting data through simple, yet purposeful, observations from which to drive cycles of change. The “just do it” mentality is sometimes lost in the pursuit of perfection, but all improve-ment methods mentioned throughout this text will not be successful if they succumb to analysis paralysis. Toyota encourages small PDCA/PDSA cycles using simple data collection methods and austere, inexpensive countermeasures. The proper use of problem-solving permits better pre-diction to increase the likelihood of successful PDCA/PDSA cycles.Toyota prides itself on its safety record for its team members and customers, but this can only happen if issues are rapidly addressed through multiple test cycles of change. After all, it is the cumulation of small cycles of change that eventually lead to bigger changes and break-through innovation. Even if automated data were available, verification of the current state through observation of the shop floor through genchi gen-butsu is of utmost importance to breaking down any problem and analyzing for the root cause (steps 2 and 4 of the eight steps of problem- solving; Fig. 5.5). Also as described in the HRO principles – sensitivity to operations and defer-ence to expertise – Toyota’s frontline teams are the experts and always assist with any unit and even an interfacility-based problem-solving exer-cise. Toyota’s leaders are present, directly inter-acting with team members and coaches to facilitate the problem-solving process. This is quite the contrast from some healthcare organiza-tions where problem-solving may occur without the direct involvement of and guidance from their senior leaders. Ideally, there should not be any perceived or actual barrier to the bidirectional communication or flow of ideas and feedback between the organization’s leaders and its team members. Not surprisingly, Taiichi Ohno valued leaders who excelled at mentoring and teaching.TPS also provides organizations with a frame-work for sustaining results through the creation of a culture where organizational goals and expectations are evident to all team members and linked to the yearly organizational strategic pri-orities. Often the best judge of organizational culture is as an outsider looking in. Multiple cli-ents of TSSC, Toyota’s not-for-profit entity charged with sharing TPS outside of Toyota, have commented that Toyota’s team members “point in the direction that they will be walking before crossing a street” and “do not walk while talking or texting on their mobile devices” – both key safety behaviors they practice when in one of Toyota’s busy manufacturing plants. Clearly the value of modeling behaviors is not lost upon Toyota team members. Similarly, Toyota leaders and managers are required to demonstrate ongo-ing mastery of problem-solving methods. This continuous cultural reinforcement, facilitated by TPS, is paramount for sustaining and continually building upon past results that leads to new, improved, and innovative products and methods.Clearly, Toyota and its production system and its history deserve our attention. Toyota’s corporate DNA [18, 32] appears to have encoded the principles of the TPS Triangle which, in turn, has been engineered into the DNA of its leaders and team members. Healthcare and other indus-tries are trying to understand how a similar trans-formation can be facilitated within their respective realms. Toyota’s journey has been deliberate. It has been subject to its constraints in its initial development from the global economic climate facing post-World War II Japan, ongoing chal-lenges from the changing global landscape, and a result of the successful application and practice of the scientific method by all of its leaders and team members.As a final thought, healthcare systems are complex and problems are inevitable, especially with regard to human error. We need to simplify these complex processes, and eliminate faulty processes that make errors more likely to happen, by employing the TPS. TPS is a different way of thinking and can be the methodology to move

111any organization along its HRO journey. Its suc-cess requires commitment and internal reflection from an organization’s leadership and team mem-bers. A review of Toyota’s history reveals a well- orchestrated journey with the development of processes to address and learn from the unex-pected! A few healthcare organizations have suc-ceeded in the application of TPS, but they have been on a multiyear journey with ongoing com-mitment to becoming even better. It remains to be seen whether the application of TPS will start increasing the velocity of change and innovation in healthcare, as we try to attain the goals of delivering value to our customers, both patient and team member, with zero harm.Key Closing Points 1. The Toyota Production System is an organizational culture of highly engaged people solving problems or innovating to drive performance. This culture is sustained by a three-part system, as described in the TPS Triangle, of (1) philosophy, (2) technical tools, and (3) managerial roles. 2. The TPS philosophy consists of four key points: (1) customers first, (2) peo-ple as the most valuable resource, (3) continuous improvement, and (4) shop floor focus. 3. Bringing problems to the surface is important. Problem-solving skills, as part of kaizen, are important to teach team members. 4. Team members learn best by doing. 5. When done properly, culture driven by TPS is a win for patients and their fami-lies, a win for caregivers, a win for hos-pitals, and a win for communities! If it is not win, win, win, win…, then it is not TPS. 6. TPS adoption can assist with the high-reliability journey of any healthcare organization.Editors’ CommentsThis chapter represents a comprehensive overview of one of the most productive, efficient, and well-known improvement process frameworks historically: the Toyota Production System. The editors sincerely appreciate the efforts of Toyota in creating this thorough chapter aimed at describing their company’s journey to develop the Toyota Production System. We find the granularity of the chapter of significant value for the reader so that one can under-stand the nuances and broad applications of the Toyota Production System.The most exciting part of the chapter is the direct application to healthcare. The second half of the chapter focuses on the use of the Toyota Production System meth-ods and processes in healthcare; the authors accomplish this by using actual cases with the methods detailing the specific interven-tions with the resultant data. Without the specific information, the reader would have been left with a theoretical understanding of their system; however, the second half of the chapter brings the teaching full circle by showing the reader how the Toyota Production System has been applied and continues to be applied in healthcare – driving outcomes that heretofore were not able to be achieved. The value of its eight steps of problem-solving methodology cannot be understated. The TPS, its Triangle and House, and its problem- solving methodology can stand alone or be used in part with other methods including the IHI Model for Improvement (as dis-cussed in Chaps. 4 and 9).This chapter epitomizes the concept of this textbook: to take theory and demon-strate how to put it in action and the bene-fits that can be derived from such an application. The quest for zero harm was the impetus for the editors – we keenly

112Acknowledgments The authors would like to thank Adam Campbell, Maya Godambe, Michael Goss, Nathan Hurle, Elizabeth Martinez, Lisa Parker, Elizabeth Pittman, and Teresa Saulnier for the critical review of this chapter.Glossary of Relevant Terms5-S refers to a visually based process for orga-nizing the workplace to reduce waste, espe-cially time spent looking for supplies, and consists of the following components: Sort, Set in order, Shine, Standardize, and Sustain. 5-S becomes 6-S if you include Safety.Andon a signal which is automatic or manual that indicates to everyone in its proximity that a problem has been detected. It often also tells the nature and location of the prob-lem and, therefore, is critical to effective problem-solving.Fishbone, Ishikawa, or Cause-and-Effect Diagram a tool used to identify potential causes for an effect or problem. This is very effective when used in conjunction with problem-solving.Gemba/Genba refers to the shop floor or place of work being examined.Genchi Genbutsu refers to the purposeful pro-cess of walking and making humble observa-tions on the shop floor or where the work takes place. “To go look, to go see, to understand, to take action.”Heijunka refers to leveled work or production.Humanize to create an environment where respect for people, a key TPS concept, is real-ized remembering that 100% of what we do ultimately impacts our customers 100% of the time.Jidoka refers to “automation with a human touch” or the process of building in quality or quality at the source. Poka-yoke and andons are part of jidoka.Just-In-Time refers to the production and con-veyance/transportation of only what is needed, when needed, and in the quantity needed. It meets the exact demand of the customer in terms of product, timing, and volume.Kaizen refers to continuous improvement and problem-solving.Kamishibai card (K-card) a tool used to ascer-tain team member knowledge of a given best practice (often used to perform audits of stan-dardized work or as rounding tools to improve compliance with best practice bundles).Kanban refers to a signal, which usually is an information-laden card, attached to equipment or supplies that enhances a pull system by sig-naling upstream of the need for new produc-tion and delivery of a product to the point of need, i.e., usually the location of the card.Lead Time the time from initiation to comple-tion of a process.Muda refers to the waste in a process within an organization. There are seven catego-ries of muda: motion, rework/defects, wait-ing, overprocessing, inventory, conveyance/transport, and overproduction. The acronym MR. WOICO is often used to help teams remember the different types of wastes. In healthcare, wasted time and potential of people is commonly referred to as the eighth waste.One-Piece Flow refers to the continuous flow of goods or parts from step to step without any batching, no work-in-process intermediate product or any intermediate accumulation of inventory. Often, to facilitate, one-piece flow steps in a process are laid out in a cellular or U-shaped layout.Poka-yoke is a part of jidoka and refers to the hardwiring of a process so that errors can-not occur. This is also referred to as mistake proofing.Pull System refers to the integrated system of production and delivery from downstream to upstream processes where upstream suppli-ers deliver product to downstream processes realize that to get to zero harm, we will need to think differently and broaden our toolkits. This chapter achieves the trifecta of teaching a theoretical framework, apply-ing this to healthcare, and inspiring us with the case studies.

113only upon signaled need. This reduces excess inventory.Push System refers to operations where prod-ucts are made and inventory created based upon expert corporate forecasts.Shop Floor see gemba or genba.Standardized Work is a key framework for kaizen improvements. It is a step-by-step document written by the people who do the work outlining the current best thinking on how to perform the process. Once standard-ized work is established, planned tests of change can occur to eventually get to a better standard.Takt Time is the rate at which products or ser-vices should be produced to meet customer demand. Takt time is the total available pro-duction time divided by customer demand. For instance, if any emergency department is open 24 hours per day and sees approximately 240 patients per day on average. Its takt time is then 6 minutes.Value Stream (or Process) Map is a visual flow map that shows how activities or processes are interconnected to design, order, and provide a given product or service. Chapter Review Questions 1. Which is a key characteristic of the Toyota Production System (TPS)? A. All important decisions must be made only by the senior leadership team from the confines of their boardrooms or offices. B. TPS dedicates many resources toward developing and encouraging team member problem-solving skills. C. TPS was created in the 1980s. D. TPS was widely adopted by many US car makers in the 1950s and 1960s.Answer: B. If you examine the TPS Triangle, the core value of the TPS is to focus on the development of the frontline team member. TPS philosophy encour-ages a shop floor focus which is visited regularly by senior leaders. 2. True or false: Transparency is an important cultural trait that needs to be adopted by healthcare organizations to bring problems to the surface quickly.Answer: True. Healthcare systems cannot fix problems that are not known. Team mem-bers need to be given the authority and asked to be accountable to bring problems to the sur-face while they are small and manageable. Safety events at healthcare systems may be related to recurrent problems that were either hidden from the surface or not addressed com-pletely when they were noted the first time. Transparency builds trust with team members, customers, and other stakeholders. 3. What are the key TPS traits that are most ben-eficial for healthcare? A. Senior leaders are visible in kaizen activi-ties and model desired behaviors. B. Problems can be best visualized through genchi genbutsu. C. Andons are part of jidoka and can be used to identify abnormalities. D. Leveling the work (heijunka) can improve patient safety. E. All of the above.Answer: E. All of the answers listed are correct. Briefly, the TPS Triangle (Fig. 5.2) discusses the importance of senior leader modeling of desired behaviors and the value of the shop floor for visualizing and bringing problems to the surface. The TPS House (Fig. 5.3) discusses the key tools or technical aspects of TPS, including hei-junka and andons. 4. Who are your customers when you, as the emergency department physician, are admit-ting a 7-year-old male patient in the emer-gency department to the inpatient unit?Answer: The most obvious customers are the patient and his family who are with him during his emergency department and inpa-tient stay. Additional customers include the inpatient unit staff and physicians. As the ED clinician, you must stabilize the patient to the best of your ability. You must then prepare and give the best handoff to the inpatient unit. You must also call the primary care physician to let them know about their patient that you just admitted to the hospital.

114References 1. Langley GL, Moen R, Nolan KM, Nolan TW, Norman CL, Provost LP. The improvement guide: a practical approach to enhancing organizational performance. 2nd ed. San Francisco: Jossey-Bass Publishers; 2009. 2. Duhigg C. The power of habit. New York: Random House; 2014. 416 p. 3. Weick KE, Sutcliffe KM. Managing the unexpected: resilient performance in an age of uncertainty. San Francisco: Wiley; 2007. 4. Toyota OT. Production system: beyond large scale production. Boca Raton: CRC Press: Taylor and Francis Group; 1988. 5. Papanicolas I, Woskie LR, Jha A. Health care spending in the United States and other high-income countries. 2018 September 14, 2019; Available from: https://www.commonwealthfund.org/publications/journal-article/2018/mar/health-care-spending-united-states-and-other-high-income 6. Institute for Healthcare Improvement/National Patient Safety Foundation. Optimizing a business case for safe health care: an integrated approach to safety and finance. Cambridge, MA: Institute for Healthcare Improvement; 2017. p. 1–20. 7. Swensen SJ, Dilling JA, Mc Carty PM, Bolton JW, Harper CM Jr. The business case for health-care qual-ity improvement. J Patient Saf. 2013;9(1):44–52. https://doi.org/10.1097/PTS.0b013e3182753e33. 8. National Health Expenditures 2017 Highlights. 2018 [cited 2019 09/29/2019]; Available from: https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/NationalHealthExpendData/Downloads/highlights.pdf 9. Institute of Medicine (US) Committee on Quality of Health Care in America. In: Kohn LT, Corrigan JM, Donaldson MS, editors. To err is human: build-ing a safer health system. Washington, DC: National Academies Press (US); 2000. 10. Institute of Medicine (US) Committee on Quality of Health Care in America. Crossing the quality chasm: a new health system for the 21st century. Washington, DC: National Academies Press (US); 2001. 11. Institute of Medicine (US) Roundtable on Evidence- Based Medicine. In: Olsen LA, Aisner D, JM MG, editors. The learning healthcare system: workshop summary. Washington, DC: National Academies Press (US); 2007. The National Academies Collection: Reports funded by National Institutes of Health. 12. Swensen S, Pugh M, McMullan C, Kabcenell A. High-impact leadership: improve care, improve the health of populations, and reduce costs. IHI white paper. Cambridge, MA: Institute for Healthcare Improvement; 2013. (Available at ihi.org). 13. Ohno T. Taiichi Ohno’s workplace management: spe-cial 100th birthday edition. New York: McGraw Hill Professional Publishing; 2013. 208 p. 14. Berwick DM. Continuous improvement as an ideal in health care. N Engl J Med. 1989;320(1):53–6. 15. Spear SJ. The high-velocity edge. 1st ed. New York: McGraw Hill; 2009. 432 p. 16. Womack JP, Jones DT, Roos D. The machine that changed the world. New York: Free Press; 1990. 352 p. 17. Graban M. Lean hospitals: improving quality, patient safety, and employee engagement. 3rd ed. Boca Raton: CRC Press; 2016. 30 p. 18. Spear SJ, Bowen HK. Decoding the DNA of the Toyota production system. Har Bus Rev. 1999;77:96–108. 19. Bell SK, Martinez W. Every patient should be enabled to stop the line. BMJ Qual Saf. 2019;28(3):172–6. 20. Toussaint JS, Berry LL. The promise of lean in health care. Mayo Clin Proc. 2013;88(1):74–82. 21. Kahneman D. Thinking, fast and slow. 1st ed. New York: Farrar, Straus and Giroux; 2011. 499 p. 22. Mangum CD, Stanley AJ, Peterson CC, Biava L, Dice J, Khan J, Godambe SA. Use of the lean manufac-turing principles to improve total parenteral nutrition logistics and clinical outcomes in the neonatal patient population. Pediatr Qual Saf. 2019;4(6):e233. https://doi.org/10.1097/pq9.0000000000000233. 23. Tasdemir N, Celik S. Self-reported pain relief inter-ventions of patients before emergency department arrival. Int Emerg Nurs. 2016;28:20–4. 24. Downey LV, Zun LS. Pain management in the emer-gency department and its relationship to patient satis-faction. J Emerg Trauma Shock. 2010;3(4):326–30. 25. Sampson FC, Goodacre SW, O’Cathain A. Interventions to improve the management of pain in emergency departments: systematic review and nar-rative synthesis. Emerg Med J. 2014;31(e1):e9–e18. 26. Bleustein C, Rothschild DB, Valen A, Valatis E, Schweitzer L, Jones R. Wait times, patient satisfac-tion scores, and the perception of care. Am J Manag Care. 2014;20(5):393–400. 27. Noonan C, Quigley S, Curley MA. Using the Braden Q scale to predict pressure ulcer risk in pediatric patients. J Pediatr Nurs. 2011;26(6):566–75. https://doi.org/10.1016/j.pedn.2010.07.006. 28. Chandrasekaran AT, Touissaint JS. Creating a culture of continuous improvement. HBR. 2019:1–5. 29. Barnas K. Beyond heroes: a lean management system for healthcare. 1st ed. Appleton: ThedaCare Center for Healthcare Value; 2014. 216 p. 30. Ishikawa K. Cause-and-effect diagram. In: Ishikawa K, editor. Guide to quality control. Tokyo: Asian Productivity Organization; 1971. p. 18–29. 31. Nakane J, Hall RW. Ohno’s method: creating a sur-vival work culture. Target. 2002;18(1):6–15. 32. Spear SJ. Learning to lead at Toyota. Harv Bus Rev. 2004;82(5):78–86, 151.

115Additional ResourcesCatalysis is the first organization of its kind to exclusively focus on educational programs and resources designed to transform healthcare value. Their website is: www.createvalue.orgToyota Production System Support Center (TSSC) web-site has multiple examples of the application of TPS: www.tssc.comVideo highlighting the TPN project discussed in Vignette #1: https://www.youtube.com/watch?v=cekpkEYc2cY

117© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 R. K. Shah, S. A. Godambe (eds.), Patient Safety and Quality Improvement in Healthcare, https://doi.org/10.1007/978-3-030-55829-1_6What to Do When an Event Happens: Building Trust in Every StepMichaeleen Green and Lee E. BudinAbbreviationsAHRQ Agency for Healthcare Research and QualityETTO Efficiency-thoroughness trade-offFMEA Failure mode and effects analysisPSO Patient safety organizationRCA Root cause analysisVA Veteran’s AdministrationWAD Work as doneWAI Work as imagined Opening Question/ProblemWhen a potentially serious harm event occurs, there is a duty to complete a thorough analysis to understand the cause of harm and the opportunity to prevent a similar, repeat occurrence. Historically, accidents are routinely followed by a public statement from a visible leader making the promise to find and fix the problem and assure accountability. The promise to find out what hap-pened and fix the problem is a genuine commit-ment; however, the practical steps to fulfill this promise are intertwined with cultural nuances that shape the journey to prevent harm.If errors were only a result of predictable pat-terns of broken parts, the promise of a certain fix could be made with confidence. The limits to confidence in the planned solution come from the realization that event occurrences are varied and often involve human performance in the context of complex and dynamic socio-technical sys-tems. In this chapter we present a series of case vignettes that illustrate how the response to an M. Green Ann & Robert H Lurie Children’s Hospital of Chicago, Chicago, IL, USAe-mail: [email protected] L. E. Budin (*) Driscoll Health System, Corpus Christi, TX, USAe-mail: [email protected] Objectives• Clarify structures and methods to use when an event occurs.• Highlight decision points and applica-tion of methods in varied situations.• Demonstrate links between experience, learning, and improvement.• Consider how decisions can affect trust and how application of methods used in event management, analysis, and fol-low- up can influence safety and improvement culture.

118event, including but not limited to an event inves-tigation, is profoundly linked to culture. To examine the effectiveness of response in each situation, we have considered the Kirkpatrick framework which is a model suggested by Perry et al. to evaluate program effectiveness by assess-ing the impact on experience, learning, improvement, and outcomes [1]. The prevention of harm depends on reliable and resilient human performance in emergent situations that are not always predictable. Supporting resilient human performance is often not a quick find and fix but rather a journey through layers of culture that include accountability, leadership, learning, and improvement cultures. These are all critical com-ponents of the culture of safety needed for the prevention of harm and appropriate response to the occurrence of harm events. Said another way, the promise to investigate and ameliorate what went wrong is not a sufficient response to an event but rather, it is essential to consider event response in the broader context of building the culture of trust and continuous improvement.The key takeaway from this chapter is to understand not only what to do but to also con-sider how each action, the conduct of those involved, and the communication in response to a serious event will have an impact on safety cul-ture and outcomes with a particular focus on the impact on trust. As Berwick suggests:Because the improvement of health care is a team effort, the issue of trust comes to the foreground. Many forms of trust are relevant to improvement: trust that the future can be better than the present; trust in patients and families, allowing us to hear their needs as legitimate and reasonable; and trust in our own capacities to learn and change, even in a hostile environment. [2]While the relationship between actions, deci-sions, and trust may vary based upon the culture of an organization, fostering trust should remain top of mind while carrying out responsibilities to respond when a harm event occurs.In this chapter we highlight case vignettes that are noteworthy enough to warrant consideration for an investigation and response either through a root cause analysis (RCA), the most commonly used investigation approach, or an alternative response. Conducting an RCA has become a familiar standard in healthcare. The Joint Commission’s Sentinel Event policy indicates:…appropriate response to a sentinel event includes the completion of an analysis of the causal and contributory factors. Root cause analysis, which focuses on systems and processes, is the most com-mon form of comprehensive systematic analysis used for identifying the factors that underlie a sen-tinel event. A hospital may use other tools and methodologies to conduct its comprehensive sys-tematic analysis. [3]Expectations for conducting comprehensive systematic analysis and reporting findings are defined not only in accreditation standards but may also be defined by state law [4]. While spe-cific rules defined in these laws may vary by state, the intent to utilize learning from harm occurrences for improvement is consistent. Approaches used for comprehensive analysis are based upon accident models that have evolved over time including those described in Fig. 6.1 that are considered throughout this chapter [5–8]. Each of these models represents an effort to understand causal relationships resulting in harm. The healthcare industry has adopted RCA methodology that was well estab-lished in other industries for accident investiga-tion, and standard step-by-step approaches to conducting RCAs are well documented and eas-ily accessible. It is important to note that avail-able references on how to conduct an RCA have evolved and improved over the last two decades. Through experience and maturity in well-devel-oped patient safety programs, we have learned to not only consider cause and effect but to also focus on the engagement of and impact on the people involved in the event occurrence. For example, both the step-by-step guide published by the Veteran’s Administration [9] and the National Patient Safety Forum’s root cause analysis and action approach (also referred to as RCA2) [10] not only provide guidance on the steps needed for identifying a root cause but also provide consideration for those who need to be involved and how participants should be engaged to establish the basis for successful implementation of identified actions for

119improvement (see Box 6.1 for websites). That is, the success of the analysis is not in just find-ing a root cause but rather in the engagement in learning from error, appreciating risk, and sus-taining improvement. Both of these resources reflect the evolution of the RCA practices recog-nizing the value of methods that go beyond just asking “why.” This is particularly important in healthcare where errors are very likely to involve human performance within complex systems. Additionally, the evolution of our understanding of how errors emerge from complexity warrants a new and broader lens described by Hollnagel and others as a new view of safety including a stronger emphasis on resilience:Simple linear models, such as Heinrich’s (1931) Domino Model that is at the heart of Root Cause Analysis, later supplemented by composite linear models such as Reason’s Swiss Cheese Model, were soon adopted as the basic safety tools in health care. Few people noticed that the very same models were being progressively challenged by industrial safety outside healthcare as inadequate to the newer, more complex working environments.During the second half of the 20th century the focus of industrial safety efforts shifted from tech-nological problems to human factors problems DOMINO THEORYFAULT TREEEVENT TREESWISS CHEESE MODELSHARP END – BLUNT ENDBOW TIE MODELRisk represented as aligment of latent and activefailures that permeate weak layers of defense.The point where events occur is the sharp end whilethe blunt end (system, policies, culture) is where thecauses are typically rooted.Combines fault tree and event tree to consider bothprevention of the event and containment of hazardsfor events that are not completely prevented.Multiple outcomes represented as consequences of arisky event with consideration of barriers to contain risk.Risk represented as a weak link in a linear chain of events.Risk representedas combinationsof conditions andcauses followingmultiple linearpaths andconsiders barriersto prevent event.CausePointof CareSystemsCulturePoliciesabcdefEventEventEventEventEventEventFig. 6.1 Understanding events: accident models used to understand cause and effect relationships. (a) Domino theory; (b) fault tree; (c) event tree; (d) bow tie model; (e) Swiss cheese model; (f) sharp end-blunt end. Figures based upon descriptions of traditional accident models from several references including [5–8]

120and finally to problems with organisations and safety culture. Unfortunately, few of the models used to analyze and explain accidents and failures developed in a similar way. The result is that safety thinking and safety practices in many ways have reached an impasse. This was the primary driver for the development of resilience engineer-ing in the first decade of this century (e.g., Hollnagel, Woods & Leveson, 2006). Resilience engineering acknowledges that the world has become more complex, and that explanations of unwanted outcomes of system performance there-fore can no longer be limited to an understanding of cause- effect relations described by linear mod-els. [11]Since the absence of harm is likely dependent upon continued resilient human performance in complex environments and trying conditions, our mindset of how to respond when an event occurs must evolve beyond find and fix. The cases described in this chapter will consider how to respond when an event occurs but will also con-sider how we must take a broader lens and con-sider implications of the human experience before, during and after the event. The effort to understand what went wrong and how to fix it is not diminished in importance; however, the effect of event response decisions, behaviors, and com-munications on culture, trust, learning, and improvement must be elevated to the same level of importance to move the needle on prevention of harm (Key Points Box 6.1).Key Points Box 6.1 Root Cause Analysis in HealthcareToolsDescriptionWebsiteRoot cause analysis tools: VA National Center for patient safety’s root cause analysis (RCA) step-by-step guideDescribes the step-by-step approach utilized by the Veteran’s Administrationhttps://www.patientsafety.va.gov/docs/RCA_Step_By_Step_Guide_REV7_1_16_FINAL.pdf [9]RCA2: Improving root cause analyses and actions to prevent harm. National Patient Safety FoundationGuidelines based upon examination of best practices designed to standardize and improve investigation of errors, adverse events, and near misseshttp://www.ihi.org/resources/Pages/Tools/RCA2-Improving-Root-Cause-Analyses-and-Actions-to-Prevent-Harm.aspx [10]As we contemplate the impact of decisions in the response to each case vignette below, we are anchoring to the following preconditions that are presumed likely given the regulatory requirements for event investigation: 1. There are existing norms within the organiza-tion for a response to an event. 2. There are structures, policies, and defined resources that define some responsibility within the organization for event investigation and response. 3. The current practice is generally aligned with recommended RCA approaches (see refer-ences and Internet resources in Key Points Box 6.1 and at the end of this chapter).This chapter does not intend to declare a definitive best practice approach but rather recog-nizes that practices continue to evolve and are married to the culture within the organizations and sociopolitical environment where they emerge and are put into action. This chapter highlights the interconnectedness between cul-ture and event response that calls into question looking for a best practice and instead calls for assessment of how to better understand the impact on culture along the way. To establish a baseline for consideration of the impact of key decisions and their impact on culture, we offer an event response roadmap in Fig. 6.2 as a frame-work for considering event investigation and response. It is presumed that the basic compo-nents of each of these responsibilities exist in some form in most healthcare organizations with varying degrees of maturity and reliability. As evidenced by the evolving body of literature on harm prevention, the event response approach

121abcFig. 6.2 Event response roadmap. (a) Organizational responsibilities; (b) organizational structures; (c) analysis responsibilities

122will continue to evolve as long as harm events occur and as we broaden our lens in considering human performance within complex systems. Understanding the StoryWe start by considering a case with detection of harm that is likely preventable but not pre-vented in this case. In fact, the situation described in this case could occur intermittently without detection, prevalent error, or harm. Historically, a response to this type of event would be to disci-pline the person with the closest proximity to the error as it was initially recognized. The applica-tion of accident models (described in Fig. 6.1) to understand cause and effect relationships has advanced our understanding of underlying sys-tems, latent errors, and root causes. A first reac-tion, and perhaps the detail provided in an event report, may focus on why the clinical team did not question the inconsistent results sooner, com-plete a more thorough assessment, or recognize the pattern of abnormally low lab results. In this situation, a root cause analysis can be used for a more comprehensive analysis of the causal chain of events and an action plan that focuses on system- level improvements to address the cause and reduce the likelihood of recurrence.This is a straightforward case that may be suf-ficiently understood using simple deductive rea-soning to explore the causal chain of events to discover a root cause [12]. This approach is illus-trated in Fig. 6.3 by asking why each step in the causal chain occurred and by considering weak-nesses in layers of defense at both the sharp and blunt ends that, if strengthened, could have pre-vented the harm. This approach can be used to highlight a component failure in a causal chain of events, but the analysis is likely to only be effective with the engagement and candor of both the people involved in patient care and those that understand the underlying systems. Trust is essential to accurately clarify the chain of events through a review of data sources and interviews with staff involved. The patient expe-rience should also be fully represented in the construction of the story. The value from an RCA often results from combined insights from a group that would otherwise never convene to collectively understand how parts of a complex system are causally related. Further value results from clarifying preventable actions in concert with system-level improvements. Sustained value comes when trust and collaborative approaches to understanding risks, detection opportunities, and harm prevention methods become pervasive through collaborative learning and improvement.Vignette 6.1Josea was admitted to the hospital for treat-ment of diabetic ketoacidosis (DKA). On the second day of admission, the plan was for the nurse to follow the titration protocol for fluids and insulin based upon glucose and bicarbonate levels that had been ordered. Josea’s status began to deteriorate, so his nurse paged the physician who began to question the treatment plan and whether there was an additional cause of his DKA beyond the presumed viral infection. Rather than staying on the titration proto-col, the physician changed the plan and ordered a consultation from infectious dis-eases colleagues. When Josea’s bicarbon-ate levels continued to stay surprisingly low, a rapid response team was called. While the team was working through the diagnostic dilemma, uncertain of the cause of the surprisingly low bicarbonate levels despite the glucose levels normalizing, an intern noted that several of her patients had inexplicable changes in their electrolyte results. After calling the lab for clarifica-tion, it was discovered that a problem with an interface had resulted in errors in the lab result reporting for over 24 hours (correct lab results were reported with incorrect val-ues). Once this was detected and the care plan was established based upon the cor-rect results, Josea’s fluids were slowed down to an appropriate level. The harm incurred included increased monitoring and lab work and he had to stay an addi-tional day for extended monitoring.

123While this approach is likely to yield success in finding a system-level problem, sustaining success in fixing the problem is often not as sim-ple as perceived when analyzing retrospectively. Further, generalized assessment of needed changes in behavior, such as encouraging staff to speak up, can be difficult to implement and may not address the underlying risks that contributed to the event occurrence. Fixing broken parts still makes sense, but preventing harm in complex and dynamic socio-technical systems is not lim-ited to fixing broken parts but also requires atten-tion to the longer efforts to change culture and build resilience. Braithwaite, Wears, and Hollnagel call attention to the need for a shift in approach:Even staunch health care supporters have gradu-ally realized that real progress will require aban-doning the Taylorist approach. Indeed, Berwick (2003) has indicated that: ‘… prevailing strategies rely largely on outmoded theories of control and standardization of work.’ It seems to be a corner-stone of the human condition that people believe – or want to believe – that they will be able to solve today’s problems, improve things, reduce errors, and ameliorate harm – all with just a few more resources, a bit more effort, another set of recom-mendations from a wise enquiry, a little more knowledge of the amount and rate of harm being delivered, increasingly precise measurements of system features, tightening up practices or a new whizz-bang IT system that is just around the cor-ner. [13]Traditional approaches of retrospective review often result in new policies and reeducation of staff. This type of response may result in some immediate risk mitigation, but the benefits are typically short-lived. The effort to engage front-line staff early in the investigation fosters trust and promotes open discussion and discovery of strategies to prevent harm, considering Dekker’s insight that “the challenge is to create a culture of accountability that encourages learning. Every step toward accountability that your organization takes should serve that goal. Every step that doesn’t serve that goal should be avoided” [14]. The trust is further developed and utilized when creating solutions that do not add additional com-plexity but rather improve usability and strengthen relationships to reduce risk in all situ-ations and not only in situations involving the parties who otherwise would have been retrained or reprimanded. In this case, the reduction in risk relies not only on fixing the broken component but also on improving detection and awareness of this risk. With the complexity of caring for hospi-talized patients, nearly all providers fail to chal-lenge mundane things such as electrolyte reporting. In retrospect it might have been a clear cause, but the intensity of routine care does not make it plausible to challenge every result that Fig. 6.3 Finding root cause by asking why

124was not as the provider might have suspected. In fact, assuming all unexpected results are wrong can prompt trade-offs that could then further delay appropriate treatment or create distraction from important information needed for clinical decision-making.A corrective action plan to address a root cause can be satisfying but carries a separate risk of hindsight bias, tunnel vision, and a tendency toward blame. These analysis pitfalls are illus-trated in Fig. 6.4. In this case, the hindsight view may result in questioning why the lab was not called earlier because, with the benefit of hind-sight, that action may have helped solve the prob-lem more quickly. Efforts to mandate presumed solutions may result in adding complexity and burden that is ultimately not helpful in preventing the next event. Tunnel vision in the analysis of this case could be a result of focusing on the com-munications that did or did not happen and not recognizing some of the other factors in the socio-technical environment. The risk of blame is inherent in any event response that includes attributing an error to a specific cause. Even when there is no intention to blame, asking why they didn’t know, didn’t recognize, or didn’t act is likely to result in at least a perception of blame.In this case with a straightforward causal chain of events, the avoidance of blame can be a bit easier. That said, it is not uncommon for clini-cians involved in the event to have already con-sidered what they could or should have done to prevent harm from reaching the patient. This reflection is inherent in the culture of healthcare providers who have taken an oath to first do no harm. The effort to strengthen trust and limit the biases inherent in retrospective review calls for attention to the impact on second victims and the effort to ensure a just culture as emphasized by Dekker:Organizational justice does involve paying atten-tion to second victims – practitioners involved in an incident that (potentially) harms or kills other people, and for which they feel personally respon-sible. There is a relationship between resilient individuals (who are supported in recovering from or even growing in the face of such inci-acbFig. 6.4 Examples of analysis pitfalls: (a) hindsight bias, (b) tunnel vision, and (c) blame

125dents) and resilient organizations (which are able to face up to their vulnerabilities and learn from them). [15]Awareness of the risks associated with second victims and just culture help shape inquiry that avoids asking staff why they did or didn’t do something that is clear in hindsight but was not evident in the emergent situation. Questions that satisfy curiosity but do little to represent the real-istic experience of those involved can further cre-ate defensiveness and limit candor. As Berwick describes, “Trust is central to this entire endeav-our. Questions that are asked with distrust, jeal-ousy, or defensiveness will not be authentic. Also, the answers will not be listened to” [2]. An under-standing of how errors occur in complex systems and, in particular, how humans must adapt within complex and dynamic socio- technical systems helps to guide an effective inquiry process. In addition to authentically retelling the story of the event as it occurred, the safety analyst must be cognizant of the challenge of clarifying the causal relationship to harm while also showing the per-spective that was experienced by those directly involved at the time of the event when the emer-gent problem was not evident.Understanding the impact of the experience on patients and their families is paramount. The clinical team is often in immediate communica-tions with patients and families that are trying to understand changes in the plan of care. While disclosure of errors is important, it can be chal-lenging when the causes are not yet understood. In all cases, embracing reflective practice is important to clinician’s evolution of practice; however, the participation in event investigation must foster reflective practice and trust rather than exacerbate a tendency toward blame includ-ing self-blame. It is also important that the avoid-ance of blame most proximate to the event does not just shift the blame elsewhere. For example, a finding that the clinical team is not at fault but shifting blame to staff in other roles stops short of finding a path to sustainable prevention of future harm. The evolution of thinking about accident models (Fig. 6.1) has helped illustrate the impor-tance of not attributing blame at the point of care (the sharp end), but shifting blame elsewhere (the blunt end) is equally unproductive. Dekker goes even further to describe the risk of shifting blame to the system rather than the individual indicating that “at the sharp end, there is almost always a discretionary space into which no system improvement can completely reach. Rather than individuals versus systems, we should begin to understand the relationship and roles of individu-als in the system” [16].Increased attention to the experience of humans in the system has helped to improve the RCA approach keeping these analysis pitfalls in mind. As Dekker states, “of course we should look at the system in which people work, and improve it to the best of our ability. But safety- critical work is ultimately channeled through relationships between human beings (such as in healthcare), or direct contact of some people with the risky tech-nology” [16]. Insufficient attention to perceptions of staff regarding both the authenticity and fairness of the analysis may limit improvement and learn-ing opportunities and may also damage the trust relationship necessary for the prevention of future harm. Similarly, hindsight bias can further disrupt the learning and improvement journey and the effectiveness of the response. Hindsight bias can be so natural to the way humans respond once an outcome is known, that those involved with figur-ing out the find and fix may unwittingly predeter-mine the outcome of the analysis and event response. Dekker offers the reminder that “hind-sight gets you to oversimplify history. You will see events as simpler, more linear and more predict-able than they once were” [17]. That said, the goal is not to make the analysis more complicated or burdensome. The goal of the investigation process is to recreate the story of the event representing the authentic emergence of the event rather anchoring to a limited hindsight view.The limitations of the effectiveness of RCA corrective action plans were highlighted by Wu et al. as they challenged the reliance on root cause analysis as the central method to learn from mis-takes and mitigate hazards [18]. The commitment to learning and improvement must consider that a mindful approach to inquiry that extends beyond asking “why?” again and again can elucidate a far richer understanding of what happened and

126why, while further promoting the trust relation-ship. For this, the involvement of people that can distinguish between work as imagined (WAI) and work as done (WAD) is essential. Hollnagel clar-ifies the risk overlooking the distinction between WAI and WAD:The difference between WAI and WAD may well be unavoidable, but it is not unmanageable. It can, however, only be managed if we recognize its exis-tence and understand the reasons for it. The single most important reason is the human tendency to trade off thoroughness for efficiency. This is the reason why solutions often are incompletely thought through, and why we accept oversimpli-fied descriptions as the basis for our plans and analyses. But we do so at our peril. [19]This highlights the importance of involving frontline staff and diverse viewpoints. Also essential is a team with and knowledge in inquiry, investigation, and safety science and leaders will-ing to address conditions that limit human perfor-mance. Effective inquiry will yield more clarity than simply asking questions that begin with “why.” Even with a goal to understand what has happened and why, in inquiry process must con-sider Hollnagel’s clarification that “incidents and accidents do not only happen in a linear manner, but include emergent phenomena stemming from the complexity of the overall health system. Asking for ‘why and because’ does not suffice to explain the system in use and does not lead to an improvement in safety” [11].The skills of effective inquiry are not easily explained and may need to be honed over a life-time. At a minimum, effective inquiry involves listening and eliciting the story of an event from those that experienced it directly. That is, effec-tive inquiry and analysis are not limited to illus-trating a linear causal chain of events, but rather are an opportunity to recreate the story of an event as it emerged from the perspective of those involved and without the advantage of hindsight. The effort to elicit the story of the event, shown in Fig. 6.5, shows a nonlinear view that is not as tidy but may be a more realistic representation risk factors related to the emergent event. This approach also highlights the need for trade-offs at the time of care provision that may not be evident when focused only on the linear chain of events used to attribute cause and effect. Both the linear cause and effect relationships shown in Fig. 6.3 and the nonlinear relationships shown in Fig. 6.5 represent the story of this event, but the framing of the story can lead to different actions for improvement. The deductive reasoning used in Fig. 6.3 is used to identify a root cause to be addressed at the system level but overlooks some of the system complexities shown in Fig. 6.5.Fig. 6.5 Recreate the emergent event

127To be sure, deductive reasoning is an intuitive way to think about error and has helped reduce harm events. In fact, deductive reasoning is a well-practiced skill used by humans to solve straightforward problems in everyday situations. According to Dekker, “Newton’s and Descartes’ ideas have pretty much set the agenda for how we, in the West, think about science, about truth, about cause and effect. And how we think about accidents, about their causes, and what we should do to prevent them” [20]. This explains our reli-ance upon reductionism to understand how sys-tems work. That is, we can understand a complicated problem by taking apart the compo-nents and reduce to smaller components until the problem becomes understandable. In the case above, finding and fixing detected problems with the interface between information systems is important especially when this component failure could recur. This fix to this detected problem would prevent the same type of error that emerges in the same way. But is it also evident that issues excluded from this causal chain of events, includ-ing workload, user interface, and siloed work-flows are factors that could cause additional errors that would not be addressed if we look only at the linear chain of events. The reduction-ist approach helps break down the system com-ponents along a specific causal path and is often easier to complete; however, it runs the risk of overlooking other critical aspects of system com-plexity that are necessary to fully appreciate the risk of other emergent errors and opportunities to ensure the safest care possible moving forward. While using reductionist thinking is a familiar approach for understanding complicated prob-lems, it is not sufficient to understand complexity in adaptive systems [21] as clarified by Dekker in his examination of complexity and systems think-ing. This suggests that we should go beyond the simple linear chain of events to consider how errors emerge and how they can be detected in the complex environments where healthcare is delivered. Attributing a cause to a system-level error is not a sufficient application of systems thinking. To understand how errors emerge from complexity, it is important to not just attribute a cause to a system component but instead to understand the risk represented in the emergent event experience as these risks are likely indica-tors of future errors if not fully addressed.As we explore each additional case through this chapter, we will highlight both decisions and nuances that warrant some additional consider-ation while navigating how to respond to an event in varied situations. The emphasis on recreating the story from the emergent perspective is rooted in a recognition that trust will be lost if the story becomes infused with hindsight bias, tunnel vision, or blame. The emphasis on trust in this chapter also recognizes that that path of the next harm event may not follow the same causal chain of events. The event response and analysis expe-rience of those involved will also have an impact on engagement in the detection of other risks that emerge from similar situations. By focusing not just on the component failure but also focusing also on the human experience, we can build a mindset of resilience regarding additional risks. Moreover, it is possible that the experience of those involved in the event analysis and response, either positive or negative, will have longer- lasting impact on the culture of safety and safe practices than the specific corrective actions identified (Key Points Box 6.2). Staying Ahead of HubrisKey Points Box 6.2• Recreate the story from the emergent perspective.• Assess the entire situation including the effect on people rather than just cause and effect.• Make trust and authenticity top priori-ties in event response.Vignette 6.2In a busy primary care practice, the nurse called back a patient who coincidentally had a name very similar to another patient

128It is often reported that the occurrence of harm events in healthcare is likely underestimated. One cause of this diminution of reported events occurs when the harm is prevented by mere chance or things that are out of the system’s con-trol. In this case vignette, there was no harm because a parent asked the right question at the right time to prompt the detection of this poten-tial error. In the case of a no-harm or minimal harm event, there is no regulatory obligation to conduct an investigation or launch an improve-ment effort. Depending on the safety culture of the organization, this event may not be reported in a voluntary reporting system, or it may be reported, and the success of the avoidance of harm may be celebrated. While affirmation of positive safety behaviors contributes to building trust and resilience, the role of luck in this case should also be appreciated. When Weick and Sutcliffe describe characteristics of high- reliability organizations, they distinguish those that remain skeptical despite success indicating that “success narrows perceptions, changes atti-tudes, reinforces a single way of doing business, breeds overconfidence in the adequacy of current practices, and reduces acceptance of opposing points of view” [22]. Remaining wary despite a favorable outcome can mitigate the risk of over-estimating reliability while underestimating the role of luck in preventing harm.In this case the initial response to the event may be limited if the continued risk is not fully appreciated. The first reaction may focus on the atypical circumstances of the case and reassur-ance that the process works all the time, except of course, for this one unusual situation. Some may say that the error was caught because the system is working and speaking up just in time is an indi-cator of a strong safety culture. There may not be enough will to dedicate the time and resources to perform a timely and in-depth analysis in pursuit of reliability. Hollnagel considers the trade-offs in the resources spent digging deeper into causal chains:Since the purpose of an accident investigation is to find an adequate explanation for what happened, the analysis should clearly be as detailed as possible. This means it should not stop at the first cause it finds but continue to look for alternative explanations and contributing conditions, until no reasonable doubt about the outcome remains. The corresponding stop rule could be that the analysis should be continued until it is clear that a continu-ation will only marginally improve the outcome. [23]When determining if a no-harm or minimal harm event warrants further investigation, it is in the waiting room. They shared last names and their first names differed by only one letter. Further complicating mat-ters, these two young ladies were both 12 years old and were both there for well visits. Vanisha (the patient who was called back) had not completed her human papil-lomavirus (HPV) vaccination series, while Manisha had. Unfortunately, Manisha and her mother thought that it was Manisha who was called back, and the care team progressed with the visit of Manisha while charting in Vanisha’s record. The physician discussed the need for HPV and influenza vaccination, the nurse drew up the vac-cines, and just prior to inoculation, Manisha’s mother pointed out that she had already received both doses of HPV vac-cine. Initially, the nurse and physician chal-lenged the mother’s assertion, but eventually it became clear that there was confusion around the patient’s identity. In discussing what happened in the office lounge over lunch, the physician was boasting that they got lucky that they didn’t deliver an unwarranted vaccine. While the risk of side effects is low, there was no reason to accept any such risk. He was overheard by the practice’s charge nurse stating, “I guess it’s better to be lucky than good.” While she agreed it was great that Manisha didn’t have an unwar-ranted vaccine delivered, the wise words of Don Berwick, “Hope is not a Strategy,” echoed in her head. She saw this as an opportunity to prevent a similar mishap in the future and recommended contacting the system’s patient safety team who sanc-tioned further analysis.

129incumbent upon the organization to balance the potential lessons learned that prevent future harm against the effort and potential erosion of trust if the dedication of resources for an in-depth inves-tigation does not make sense to those closest to the situation. In this case, the decisions regarding how to respond to the event are less about the determination of actual harm and more about understanding the complexity of the situation and likelihood of recurrence of the risk. The resis-tance to conducting an in-depth analysis could be overcome with persistence but the intended result of sustaining reliable changes in process and behavior may become even more elusive without the burning platform that is usually associated with harm events. The goal is not to analyze more but rather to engage staff in learning and improve-ment that includes remaining sensitive to the inherent risks. The decision on how to respond in this case is best informed by considering the approach that is most likely to garner the resources and commitment to improve.Even with recognition of risk in this situation, the best path to learning and improvement may be found through less analysis and more attention to how clinical teams can partner with patients and families for better outcomes and experience for both staff and patients. A mature patient safety program has likely developed a portfolio of methods of event response that are not limited to root cause analysis. An alternative approach in this case is to consider is an apparent cause anal-ysis to understand what happened. In cases where the causes or contributing factors are apparent, a decision to spend less effort digging into the causal chain of events and more effort on the improvement approach may be warranted. The North American Electric Reliability Corporation (NERC) suggests that the why staircase approach is typically a good fit for apparent cause analysis and further suggests different analysis approaches for different situations. Alignment of the right approach for the right situation considers both efficiency and effectiveness and recognizes that not all cases warrant the same approach used for more complex cases [24]. The use of apparent cause analysis is an option for judicious use of resources and can still lead to rigorous improve-ment by applying improvement science to the implementation as described by Crandall et al. [25].The resistance to a full investigation for a near miss event does not have to be a barrier and can instead be used as leverage to focus on learning and improvement. If staff insist that the event is unlikely to recur because the process is usually reliable, there is also an opportunity to focus less on what went wrong and instead focus on under-standing what happens when the process goes well and the importance of relationships and communication in assuring the best outcome. Dekker clarifies that the search for the root cause, or broken part, can be limiting. “If we want to understand why it ended up broken, analytic reduction doesn’t get us very far. Instead we need to go up and out, rather than down and in. We have to begin to probe the hugely intertwined web of relationships that spring out and away from the broken part, into the organizational, the institutional, the social” [26]. This suggests focusing on creating the culture and environment that supports human performance in risky situa-tions. By shifting focus from the inquiry into what went wrong and instead considering the use of appreciative inquiry [27] as suggested by Trajkovski et al. as a way to understand what often goes right, there is a path to overcoming the resistance to both staff and patient and family involvement in the event response efforts. In some organizational cultures, this may yield bet-ter engagement in learning and improvement. Moreover, this process is likely to extend beyond fixing a process or technical component and, instead, extend further to consider relationships and how people adapt and collaborate for the pre-vention of harm.A focus on the positive is often well received but must not overlook a realistic perception of risk. The effort to accurately assess risk can ben-efit from considering not only an internal assess-ment but participation in learning communities that foster greater transparency to better appreci-ate the risk. Events that are relatively rare occur-rences are often perceived as unlikely to recur only because we lack perspective on occurrences elsewhere or lack appreciation for the severity of

130potential consequences. This case vignette describes similarities to the confusion that led to the wrong procedure performed on a young child reported by the Associated Press in 2000. The report also described an overwhelming sense of devastation that the error had not been prevented [28]. Appreciation for risk and motivation to improve should not be limited because we lack frequent devastating events or publicity for close calls. This blind spot can be ameliorated by par-ticipation in a patient safety organization (PSO) or other learning communities that foster trans-parency of information about the pervasiveness of risk in healthcare environments. Greater trans-parent learning about harm and risk of harm can inform the assessment of rare events as they are detected more frequently when information is shared within a larger community with similar risky environments. To better understand the complexity, risk, and likelihood of recurrence, hazard assessment may be a more effective approach than cause analysis in this case. Again, the decision on how to respond considers the trade-off between a more detailed approach such as conducting a failure mode and effects analysis (FMEA) [29] or a simple hazard assessment matrix that highlights relationships between the probability of occurrence and severity of impact [5]. Both methods shown in Fig. 6.6 are similar in the factors considered in the assessment. The event response should consider the likely effect on the engagement of staff in sustainable improvement when choosing between the more thorough and detailed approach or the simple and efficient alternative. In this case the approach of using the simple matrix may be a sufficient first step to inspire the engagement in learning and improvement with continued sensitivity to the risk of harm.As part of the event response process, it is common for events to be classified in terms of level of harm including near miss and minimal abFig. 6.6 Hazard assessment: (a) failure mode and effects analysis (FMEA); (b) hazard priority matrix

131harm events. This event is not likely to be classi-fied as a harm event because the patient did not experience actual harm. There could also be a debate on whether or not this event constitutes an error or if would be considered a near miss event. Hoppes and Mitchell summarize several harm classification systems in their white paper [30]. Regardless of what classification system or harm scale is applied, there could be some disagree-ment in how this event should be considered. While an effort is often made to assure consis-tency in classifying harm, consistency is elusive due to subjectivity in the classification process. Walsh et al. studied the reliability of harm clas-sification and found that “Unfortunately, evi-dence to date suggests that clinician ratings of severity for adverse events are highly variable, with Cohen’s Kappa coefficients ranging from 0.4 to 0.76. In spite of the importance of adverse event ratings, there has been little information on how to optimize the reliability of ratings” [31]. Williams et al. discovered similar challenges in reliability in assigning levels of harm [32]. Despite strategies to ensure inter-rater reliability, harm classification is often inconsistent. Pitfalls include distinguishing between potential harm and actual harm. There can also be subtle and subjective distinctions between levels of harm severity.Does this matter? Application of harm scales is often used to clarify safety outcome data and to capture error rates to measure safety over time. Dekker challenges the usefulness of error rates as safety outcome measures indicating that “most organizations which have suffered big calamities over the past decades had exemplary perfor-mance on incidents and injuries” [33]. He further clarifies that an “organization does not have a great safety culture because it has a low number of incidents. In fact, the opposite is true” [33]. That is, robust detection and reporting may be a stronger indicator of safety than error rates that rely upon a subjective classification of harm. The Agency for Healthcare Research and Quality (AHRQ) also highlights the limitations to current data systems recognizing that event data cannot be translated to error rates without population data [34]. This specific case of a no-harm event illustrates that, regardless of harm classification method or outcome metric utilized, the risk in the environment identified will most likely not be reflected in the harm classification and therefore will not be reflected in an error rate. If harm clas-sification data and error rates do not capture risks that show subtle changes in culture or gradual drift from safety-critical boundaries, they are unlikely to help with prediction or prevention of risks that are emergent from culture and com-plexity rather than a result of component failures.What becomes more important is the engage-ment of staff in cultural aspects of learning and improvement. If the attention to error rates is the key to engagement in learning and improvement, then establishing trust in the reliability of those data becomes particularly important. Since harm classification has poor inter-rater reliability, understanding the impact of over-calling or under-calling the resultant harm is also important. Some organization cultures may find a better path to engagement by focusing on the positive out-comes and collaborative relationships. In some cases, it is perhaps better to give less attention to the harm classification and error rates and focus more on the collaborative culture needed to pre-vent a drift away from safety-critical behaviors and safety boundaries. The key point is again, focus on understanding how the decisions made throughout the event response will impact the experience of people involved. Each decision that enhances trust and improves culture is likely to benefit the longer-term goal of sustained improve-ment. A broader lens considering how to respond to an event can be realized by recognizing that focusing solely on finding and fixing a problem often does not result in reliable adoption of a change in practice. This is especially true when culture and behavior are part of the solution. The focus on trust, culture, and engagement of front-line staff in designing new ways to work is better aligned with improvement science methods that do not start with solutions, or corrective actions, but rather rely upon disciplined methods to engage frontline staff in designing improvements that are realistic and sustainable (Key Points Box 6.3).

132 Managing the ConsequencesThis case describes a situation that may not be identified for analysis or follow-up until medical malpractice activity raises concern about possi-ble liability. Depending on reporting culture and surveillance systems, an unexpected outcome may be detected immediately; however, some complications may not be easily detected for fur-ther review or analysis. Once identified, if the outcome is not the expected or intended outcome, further consideration regarding preventability is warranted, but a presumption of error is not war-ranted. In this case, understanding cause and effect may not be as challenging as clarifying preventability and potential liability.There is substantial improvement momentum in eliminating hospital-acquired conditions as shown in the results of the collaborative effort described by Lyren et al. [35]. This represents an important shift that many conditions that used to be accepted as known complications are now considered preventable. The shift in considering hospital-acquired conditions as preventable was made possible by collaborative work to develop the evidence base that clarifies how these condi-tions can be prevented with changes in practice. Yet there are still many complications where pre-ventability is unclear. In their white paper, Hoppes and Mitchell offer a timeline that illus-trates the shift in thinking about harm over sev-eral recent decades and further state that “Learning and improvement should also occur from events that are classified as known complications or no harm, as there is often oppor-tunity for risk reduction in complications and no harm events and/or trends of events that may not be considered preventable at the time of occur-rence. Learning from near misses is one of the tenets of patient safety” [30]. For many compli-cations that occur in the course of care, clarifying preventability is not limited to understanding cause and effect but also understanding the cur-rent standard of care.Determining deviation from the standard of care may not always require a comprehensive root cause analysis but does warrant some retro-spective review. Hoppes and Mitchell emphasize the importance of understanding whether there was a deviation in practice standards and offer a decision tool that highlights both the reliance on evidence-based practice as well as assessing care decisions in the context of the situation when the outcome occurred [30]. Like the first case, the goal is to understand the story of the emergence of the outcome in the context in which it occurred. Response to this event should include a review of Key Points Box 6.3• Remain wary despite successful outcomes.• Do not equate harm and hazard.• Focus response to engagement in improvement.Vignette 6.3Jason underwent the removal of a large lipoma that had appeared on his face. Prior to his to procedure, Jason’s care team reviewed the procedure consent form with his parents. The procedure went as planned except the depth of the lipoma was greater than anticipated and closure was performed with unexpected difficulty. Jason was dis-charged after his outpatient procedure after the plastic surgeon chose to not disclose the unanticipated challenges as she thought there was an acceptable outcome despite the unexpected challenges. Notes in the outpatient chart indicated that Jason had a significant keloid development at the site. His parents expressed dissatisfaction with the results but stopped this part of the con-versation when the surgeon stated that scar-ring was addressed in the consent form. Despite sensing the family was not pleased, the plastic surgeon discussed the plan for additional scar revision procedures. Jason’s family agreed to the plan of care and left with the plan to return for a scheduled pro-cedure. Prior to the procedure, an attorney contacted the hospital to request a copy of his records.

133the literature and current standards defined by policies or protocols. Many cases like this include decisions that include a trade-off of risk and ben-efit. Ideally, the risk and benefit relationship and consideration of a known complication would be clarified through communication and consent prior to the care episode; however, there may not be adequate anticipation or documentation to clarify risk-benefit trade-offs as they occur. There also should be an assessment of decisions made in managing uncertainty and reasonably unfore-seen circumstances.This is another situation that calls for caution regarding hindsight bias. Once an outcome is known it can be easy to cherry-pick a possible cause without looking at the complexity of man-aging the trade-off between multiple concurrent risks. Dekker suggests that it is important to “rec-ognize that it is often compliance that explains people’s behavior with norms that evolved over time – not deviance. What people were doing was reasonable in the eyes of those on the inside of the situation, given the pressures and priorities operating on them and others doing the same work every day” [36]. This highlights the impor-tance of involving peers in the assessment of whether a deviation has occurred. The applica-tion of an algorithm such as the decision tree for unsafe acts [37] and applying the substitution test or a review of literature regarding management of known complications can be very helpful to assess if a decision was an error or a reasonable clinical judgment. The involvement of peers in the application of these tests is essential.The response to this event should also include learning that is not limited to understanding the clinical decisions and processes. In some cases, the liability is created through breakdowns in trust rather than errors in care. Again, the analysis in this case may benefit from less emphasis on digging into the causal chain of events and more about understanding the communication and management of the trust relationship. Additional perspectives regarding patient-family experience and patient relationships can be valuable in this assessment. Focusing learning and improvement in detecting and containing the trust problem may offer a greater benefit than a determined effort to attribute the clinical outcome to a spe-cific cause. In this case, the problem to be con-tained is the breakdown in communication resulting in the breakdown in trust.This case also highlights the importance of disclosure as part of the event response. It is com-mon for healthcare organizations to have policies that guide the communication of adverse out-comes. Policies may clarify the mechanics of the process but are likely insufficient support to those that must navigate these crucial conversations. The art of effective communication in disclosing adverse outcomes will be shaped by the organiza-tional culture and will reflect both risk tolerance and approach to just culture within the organiza-tion. The skill and discipline in effective disclo-sure continue to evolve in healthcare organizations along with the understanding of the benefits and ethics associated with disclosure. Resources such as the Communication and Optimal Resolution (CANDOR) toolkit [38] can be a place to start. The effort to bolster or restore trust through dis-closure of details of a complication, absent iden-tification of an error, relies upon effective communication and a specific understanding of the trust relationship in question.In this case, where the patient continues to seek care despite ongoing litigation, there is an urgency to restore the trust relationship. Regardless of the assessment of deviation or pre-ventability, the priority is to manage the immedi-ate needs of those involved. Dekker’s explanation of just culture highlights the importance of restorative justice and suggests that “Restorative justice asks very different questions in the wake of an incident: who is hurt? What are their needs? Whose obligation is it to meet those needs?” [39] The consideration of these questions is not only applicable in considering communication and support for the patient but also for the others involved in the experience of the adverse out-come. Dekker succinctly describes reporting obligations and ethics for adverse outcome and disclosure:The ethical obligation to disclose your role in an adverse events comes from a unique, trust-based relationship with the ones who rely on you for a product or service. Disclosure can be seen as a marker of professionalism. Disclosure means mak-ing information known, especially information that was secret or that could be kept secret. Information

134about incidents that only one or a few people were involved in, or that only professionals with inside knowledge can really understand, could qualify as such. [40]In this case there is not an error to disclose but rather an opportunity to strengthen communica-tion and understanding about the outcome and ongoing care. This case also reinforces the mes-sage that the event response is all about trust whether the focus is preservation of trust or resto-ration of trust. Both are important and the entire array of trust relationships, clinician to patient, patient to organization, and organization to clini-cian, should be considered when managing the response to this event.Lastly, this case highlights the need to coordi-nate an array of resources to manage the event response. It is necessary to consider not only event investigation and improvement resources but also to assure a collaborative and aligned approach involving communication, patient relations, and disclosure processes. The alignment of these con-current processes will help assure a cohesive expe-rience for the people involved in all aspects of this event response (Key Points Box 6.4). Responding with Unanticipated UrgencyThis case is similar to the Vignette 6.3 in that it may not be detected until notification is received from an outside agency. Also like the case above, this could result from a patient com-plaint to a regulatory body, or it could result from external quality surveillance systems with differ-ent sensitivities than those used internally. For example, some external quality groups may prompt review based upon a readmission or the use of a billing code that may or may not be related to an error. In all events, the timeliness of the response is important, but in this case, the obligation to report findings externally may exert additional pressure on both the efficiency and thoroughness of the response.To respond to this pressure, the initial investi-gation should consider the right type of analysis for the situation. It is important to recognize the Key Points Box 6.4• Event management is distinct from event analysis.• Engage and leverage resources to man-age and coordinate parallel processes (i.e., analysis, communication, patient relations).• Attempt to restore trust through effec-tive communication and disclosure.Vignette 6.4Jackson was thrilled to hear that he was going to be discharged after being treated for congestive heart failure. He expressed his delight while reviewing his prescrip-tions and plan for follow-up visits at the time of his discharge. The timing was great as he was going to join his family on a trip the following week. Just before leaving for the airport, Jackson realized he did not feel well and went to the emergency department instead. During triage he asked if he could also visit the pharmacy to fill his prescrip-tions as he had not filled them after his dis-charge. Jackson was admitted, returned to baseline, and was then discharged late the next day. Two weeks later the hospital received a notification from the Department of Public Health that indicated that con-cerns were raised that Jackson did not receive adequate care prior to his first dis-charge which resulted in his readmission. Since the readmission occurred within 7 days, this case met the criteria for further review, and the hospital leadership expressed concern about financial penalties associated with readmissions. The care team indicated that they believed the read-mission was related to nonadherence with his medication regimen.

135accreditation bodies share the goal of ensuring the response to the event advances understanding of how to identify hazards to be addressed to ensure safety and reliability in the system. For the case described here, there is less complexity in understanding the cause of the readmission and a need for greater attention on how to strengthen detection and containment. As Weick and Sutcliff indicated, while the unexpected is pervasive, “what is not pervasive are the well-developed skills to detect and contain errors at their early stages” [41]. While it is reasonable to assume that Jackson has culpability for having not obtained his prescriptions, stopping the review at this point will lead to missed opportunities to help future patients. Patients and families are often over-whelmed by the information at discharge and might not fully appreciate the importance of the timing of acquiring medications.In this case an apparent cause analysis is likely sufficient to understand the factors that led to the readmission as long as it clarifies opportunities for better detection and containment of risk. McLeod and Bowie also highlight the usefulness of a bow tie analysis to understand both causal and contributing factors and safeguards neces-sary for detection, containment, and management of hazards [42]. As shown in Fig. 6.1, the bow tie analysis combines the concept of a fault tree and an event tree with the top hazard placed in the middle of the two sets of branching logic. While the branches of the fault tree consider potential causes similar to a root cause or apparent cause analysis, the addition of the event tree is utilized to understand consequences after the event occurs. This approach helps to expand the analy-sis to include not only consideration of how to prevent the hazard but also consideration of mechanisms needed to contain problems while they are small and steps that can still be taken to prevent harm. In this case, an improvement focused on containment may emphasize follow- up communication after discharge rather than focus only on what happens prior to discharge. Using the bow tie analysis method in this case may offer an efficient alternative to a root cause analysis and may also offer a broader lens to focus on the learning and improvement needed.In this case, managing the event response requires a close partnership between resources that investigate events and resources that manage relationship and communication with the involved regulatory agency. This may be one in the same team or may include two groups work-ing in alignment. The response to the event is not necessarily different than what would occur if the event had been detected internally; however, the learning and improvement opportunity may include attention to improving detection and sur-veillance systems and may also include opportu-nities to focus on proactive hazard assessment to better appreciate risks that are currently not suf-ficiently detected or contained.This case also has similar challenges to the prior case in that the response to the event may highlight opportunities related to communication with patients and families. Subsequent care and communication rely upon restoring the trust rela-tionship even though the regulatory agency may now be a participant in communications between the organization and the patient and family. While all parties have the shared goal of the best possi-ble outcomes and experience going forward, the restoration of trust through all lenses will rely upon the effective coordination of communica-tion and due diligence in responding to this event (Key Points Box 6.5). Clarifying More than CausalityKey Points Box 6.5• Find balance between efficiency and thoroughness in response.• Engage resources to manage and coor-dinate parallel processes (e.g., event response and communication with accreditation body).Vignette 6.5Evelyn arrived at the emergency depart-ment via ambulance. She was stabilized at

136This is a case that presents unique challenges in understanding the best response. Traditionally, a potential error in diagnosis would only have been evaluated through a morbidity and mortality conference. This evaluation is still appropriate; however, with the increasing understanding of the complexity in healthcare, there is recognition of the benefit of understanding diagnostic errors through a systems view. The Society to Improve Diagnostics in Medicine highlights the National Academy of Medicine (formerly the Institute of Medicine) definition of diagnostic error “as the failure to (a) establish an accurate and timely explanation of the patient’s health problem(s) or (b) communicate that explanation to the patient” and additionally emphasizes “diagnostic error stems from the complexity of the diagnostic pro-cess, complexities in how health care is deliv-ered, and the same kinds of cognitive errors that we all make in our everyday lives” [43].An event report resulting in an analysis of a diagnostic event is likely to be prompted by an unexpected outcome that may or may not be caused by an error. This case is similar to the known complication case above in that the initial challenge is to clarify whether there was a devia-tion in the standard of care. Again, this determi-nation will likely rely upon the expertise of peers, a literature search, and application of tools such as the substitution test. In this case a comprehen-sive retrospective review through a root cause analysis can help assess not only potential errors in the diagnostic process but also contributing factors from the complexity of the system and environment. What can be particularly challeng-ing is understanding the factors that may be caus-ally related to the outcome when it is difficult to clearly distinguish whether the outcome resulted from these factors or progression of the disease.While conducting a root cause analysis meets the expectation to conduct a comprehensive anal-ysis of causal factors, the pitfalls of a root cause analysis, including the risk of hindsight bias and misattribution to a component failure, are partic-ularly evident in this case. Our understanding of why diagnostic errors occur and how to prevent them is developing but is not as well established as our understanding of process errors and com-ponent failures and how they can be fixed [44]. There is notable attention to learning in the medi-cal community about the risks of cognitive biases and some promising attention to the development of clinical decision support resources, but there is still only limited evidence on how to detect and prevent the array of diagnostic errors that occur but are largely unreported.an outside hospital, but given the complex-ity in Evelyn’s condition, admission at the hospital with the team of specialists involved in her care was justified and a plan that comforted her parents. Although Evelyn was stable, there were many hand-offs between hospitals, transport, and from the emergency department to the inpatient unit. During handoffs, it was unclear whether her antiepileptic medications had been administered, and at some point, it was assumed and reported that they had not. Subsequently, when she arrived at the unit where she was well-known, her nurse who endured several episodes of status epi-lepticus with her in the past and made sure she gave her medications as soon as she could get them on the floor. Unbeknownst to her, they had been given in the referring hospital, and this was a repeat and unneces-sary dosing. She was noted to have increased somnolence, and the repeat dos-ing was identified. The levels drawn when the error was noted were found to be at the lower end of the range known to be at risk for toxicity. She remained somnolent well beyond the expected timeframe, and this led to a prolonged admission and further test-ing. While it was presumed that her change in level of alertness was due to the dosing error, her lack of improvement was not explained by this error, and this led to a delayed diagnosis of viral encephalitis. While this delay did not cause interventions to be withheld, it was realized that this could be a real risk in other situations when attributing symptoms to the wrong cause.

137A starting point may be a recognition that the nature of a diagnostic error is fundamentally dif-ferent than a process error or a component fail-ure. That would suggest that our response to this type of event includes understanding the perfor-mance shaping factors present in the situation and environment and that may mean going up and out, as Dekker suggests, rather than going down and in to attribute the error to a single root cause [26]. Contributing factors may not be lim-ited to linear cause-effect relationships but may also include dynamic coincidences of perfor-mance variability with humans performing within complex adaptive systems. The analysis should appreciate the complexity in the system and con-sider the suggestion from Braithwaite, Wears, and Hollnagel’s that “adverse events increasingly needed to be explained as unfortunate combina-tions of a number of conditions, rather than as failures of single functions or components – including ‘human error’” [13]. Once again, the goal of the retrospective review is to recreate the story as it emerged from the perspective of those involved rather than through a limited hindsight view.Since humans are integral to the diagnostic process, an analysis to understand diagnostic error must be informed by knowledge of human performance, complexity, performance shaping factors that influence decision-making, medical knowledge, and understanding of the diagnostic process. This means that the mental model of understanding causal relationship to error must shift from linear thinking about resultant events caused by linear component failure to consider-ing emergent events that occur in dynamic and evolving situations where humans are adapting to unknowns and problem solving in the moment. Hollnagel’s description of the efficiency- thoroughness trade-off, or ETTO principle, is particularly helpful in understanding this case. The ETTO principle describes the balance between time to think and time to do in the con-text of time pressure and, at times, competing priorities. Hollnagel further clarifies:The ETTO principle refers to the fact that people (and organisations) as part of their activities fre-quently – or always – have to make a trade-off between the resources (time and effort) they spend on preparing an activity and the resources (time and effort) they spend on doing it. The trade-off may favor thoroughness over efficiency if safety and quality are the dominant concerns, and effi-ciency over thoroughness if throughput and output are the dominant concerns. It follows from the ETTO principle that it is never possible to maxi-mize efficiency and thoroughness at the same time. Nor can an activity expect to succeed, if there is not a minimum of either. [45]A tendency toward greater efficiency could result in the wrong action, and a tendency toward thoroughness could result in an action that occurs too late. In this case the correct diagnosis was in the differential but was not quickly identified as the correct cause of the symptoms seen. Thankfully, because her viral encephalitis only required supportive care, there was no interven-tion withheld; however, it is easy to think of situ-ations where this delay could result in an adverse outcome. The concept of managing the risk asso-ciated with trade-offs is particularly relevant for diagnostic challenges. While generally, thor-oughness is aligned with the caution that could benefit safety, there are some situations that the risk of delay of intervention outweighs the impor-tance of using thoroughness to assure certainty of the diagnosis and plan of care. While the trade- offs in diagnostic decision-making may be evi-dent in hindsight, increased reliability relies upon conditions that support the ability of clinical teams to assure timely interventions while mak-ing sense of uncertainty often in rapidly evolving situations. Weick and Sutcliff offer that “sense-making is about updating plausible stories, often by means of action, while looking for data that question initial hunches” [46]. Safety in light of diagnostic dilemmas relies upon constantly find-ing the right balance between efficiency and thor-oughness trade-off decisions that are needed for both the diagnosis and the delivery of timely interventions.A corrective action plan resulting from the analysis of a diagnostic error is likely to consider how to predict and prevent similar errors in the future. The limitation in predicting diagnostic error is, in part, due to the likelihood that

138diagnostic errors are more likely to be emergent from complexity rather than something that could be predicted by interpreting data on previous no- harm or low-harm events. The mental model that near miss event reporting data is predictive for this type of emergent error may not be realistic. The experience of a retrospective review of this event is likely to become integral to the reflective practice of the providers involved but difficult to spread. Transparent and broader learning from retrospective case review is valuable and impor-tant to increase awareness of risk. The challenge in error prevention is this case is the pervasive-ness of the ETTO principle in action in the com-plex high-risk environments where care is delivered. A corrective action plan in this case is not likely to be limited to quick wins or easily assignable tasks but instead is likely to include strategies to change collaborative culture and practices, to enhance team behaviors, and to sup-port human adaptation in complex changing environments filled with uncertainty and compet-ing pressures.This case also highlights challenges related to just culture and adverse impacts on second vic-tims that were described in previous cases. However, in cases with resultant harm, the effects on people involved are likely to be felt more acutely and profoundly. This event presents simi-lar disclosure challenges as discussed in the pre-vious cases, but based upon the outcomes, restoring trust in this situation can be difficult. The questions associated with restorative justice mentioned above are relevant in both cases: Who is hurt? What are their needs? Whose obligation is it to meet those needs? The event response in any case should also consider how to provide appropriate support to the patient, the patient’s family, and the care team (providers) in the after-math of the event experience.The attribution of a harm classification to this event can also be challenging especially if the causal relationships cannot be determined with certainty. This effort should rely upon peers with the medical knowledge to assess the plausibility and probability of the presumed causal relation-ships. The uncertainty regarding causality and preventability could result in a subjective classifi-cation decision ranging between considering the event a significant safety event or not a safety event at all. Considering the challenges with inter-rater reliability in the attribution of harm, it may be difficult to find a standard to establish which harm classification outcome is correct. As described throughout this chapter, each decision has an impact on trust relationships. Looking out-side of the organization either for external event review or event reporting to a patient safety orga-nization or similar learning organization can help to guide these decisions. Similarly, a better understanding of just culture and restorative jus-tice can also help shape these difficult decisions. Like all the cases considered in this chapter, what may be most important is how these decisions help guide improvement in the learning, improve-ment, and safety culture (Key Points Box 6.6). SummaryIn this chapter we have considered an array of situations that highlight the decisions encoun-tered while navigating the response to events that occur in healthcare organizations. Through each situation we have emphasized the importance of trust and authenticity; tritely stated – no two situ-ations will have the same response. Determining how to navigate event response requires under-standing the culture that exists within an organi-zation and the culture that is needed, and that the path to a better future is created through trust. This emphasis is rooted in the belief that experi-ence with harm events has a notable impact on the people involved, and that the people are the key to safety in complex and dynamic socio- technical systems.Key Points Box 6.6• Consider plausibility and probability when the causal relationship to outcome is uncertain.• Understand efficiency-thoroughness trade-off in a realistic context.• Attend to the second victim and just cul-ture risks.

139Safer care is reliant upon human ability to work collaboratively and adapt to complexity, problem solve, manage the unforeseen, and appreciate safety boundaries while balancing trade-offs. Because of this, the response to events must emphasize social relationships as much as causal relationships. Perhaps the greatest learn-ing from harm events comes from the appreciation that humans are uniquely able to adapt to com-plexity and do so more quickly and naturally than processes, protocols, or technology. The key takeaway from this chapter is to assess each deci-sion made while responding to harm events when they occur. Start by listening and understanding the experience of all the people involved in the event and continue to understand how each per-son experiences the key steps in the event response. Lastly, to increase the likelihood of sustained improvement, understand how each experience will influence the relationships, trust, and culture that are needed to support human per-formance needed to adapt and manage the risk of harm in complex environments. Internet Resources• Agency for Healthcare Research and Quality: AHRQ.gov [47].• Erik Hollnagel website: www.erikhollnagel.com [48].• Institute for Healthcare Improvement: IHI.org [49].• Society to Improve Diagnosis in Medicine: https://www.improvediagnosis.org [50].• Safety Differently: The Movie: https://youtu.be/moh4QN4IAPg [51].• Teaching and Assessing Critical Thinking: https://medicine.dal.ca/departments/core-units/cpd/faculty-development/programs/TACT.html [52].Editors’ CommentsEmbracing high-reliability principles can drive hospitals toward unprecedented out-comes in quality and safety. Many chapters in this textbook speak of the five principles of high reliability from Weick and Sutcliffe. Broadly the principles are grouped into anticipatory and containment; highly reli-able organizations focus on anticipating where, how, when, etc. problems can occur, and they also have systems in place to con-tain them once they inevitably occur. This chapter deals with the second grouping from Weick and Sutcliffe, containment. The title of this chapter summarizes the point of Green and Budin: “What to do When an Event Happens: Building Trust in Every Step.” The chapter, through the series of vignettes, demonstrates how trust in one another, our colleagues, and the sys-tem is the keystone of being ready for how to respond when an event occurs.The chapter thoroughly explains the role of root cause analyses (RCA) and how they can drive an understanding of an event as well as the response and action planning; our organizations com-plete approximately 12 RCAs a year between our 2 organizations (editors, RS-SG). We believe that for our organiza-tions with 500+ beds and 30,000+ pediat-ric admissions with large emergency departments and many ambulatory set-tings between our facilities, this number of RCAs is “healthy” for our organiza-tions; too many would be onerous and not value add, and doing less than this amount would not provide a robust safety and quality program. Each organization will ultimately decide on what healthy rhythm and amount of RCAs are best for their culture. Not all RCAs represent events that go poorly, but we also try to learn from events that go well and celebrate these moments.As the authors indicate in the fourth vignette, it is important for an organization to align their resources and responses to an event in parallel. This includes the event response (e.g., root cause analysis), inter-

140 Chapter Review Questions 1. Describe how hindsight bias can affect the attribution of root causes in a retrospective review of a harm event.Answer: When considering an event through retrospective review, those involved in the analysis have the benefit of already knowing the outcome. This can lead to a belief that the same error will occur in the same way and that prevention is simple. While this may be true for some errors that result from simple process breakdowns, this is often not true for errors that emerge in complex adaptive sys-tems. This can lead to selecting seemingly straightforward solutions that are not suffi-cient because they do not consider the chal-lenges of making inevitable trade-off decisions in complex environments where conditions include managing unknowns or emergent challenges. 2. Describe the differences in methods used for retrospective review.Answer: Most accident models used for cause analysis focus on understanding failures considering linear chains of events. The bow tie analysis method considers opportunities for containment as well as opportunities for prevention. Effective use of inquiry can broaden the perspective to consider the story of emergent events without reliance on hindsight. 3. What are alternative approaches to event response when a retrospective review is not required?Answer: Alternatives include the use of hazard assessment tools such as failure mode and effects analysis (FMEA) or a hazard assessment matrix (see Fig. 6.6). Appreciative inquiry is another alternative approach for engagement in the discovery of improvement ideas when a retrospective review is not required. 4. True/False – Effective inquiry means always asking why five times.Answer: False. While the approach of ask-ing “why?” five times is a structured approach to deductive reasoning, the limitations of this approach should also be recognized. Hindsight bias, tunnel vision, and attribution of blame are risks associated with this approach. The risks of these analysis pitfalls should inform an inquiry process that also considers com-plexity, competing priorities, and other factors that influence trade-off decisions. The best use of inquiry is to authentically re-tell the story as it emerged. Is it important not only to go “down and in” to understand linear causal chains but just as important to go “up and out” to understand trade-offs that occur in the com-plex adaptive system. 5. Identify the impact of key decisions in other situations such including: error affecting many patients, exposure to staff and patients, or staff injury.Answer: Each of these scenarios should include consideration of what resources are needed for event management. Similar to some of the cases described in this chapter, it is often not sufficient to focus on analysis and corrective actions but important to also con-sider management of the entire situation. An error affecting multiple patients will require consideration of how to manage the opera-acting with the family (i.e., disclosure, involvement of the ombudsman, etc.), the risk management component (reporting to the institution’s insurer), supporting the second victim if one exists, etc. There are myriad tasks that should occur in parallel once an event occurs; to wait and line them up to accomplish them one-by-one can per-haps be deleterious and result in worsening a culture and not properly addressing latent system defects.Ultimately this chapter moves us past simply performing rote steps in response to an untoward or unexpected outcome; the chapter implores us to use the culture of the organization and the trust that has devel-oped to deftly navigate an appropriate response.

141tions tasks of communicating, evaluating, and providing subsequent care to multiple patients. A situation involving exposure to both staff and patients may require separate but coordi-nated resources to urgently evaluate and care for both staff and patients. Lastly, response to cases involving staff injury may require the involvement of additional expertise and may involve different policies and information sys-tems that are not always aligned with resources used in response to patient events.References 1. Parry GJ, Carson-Stevens A, Luff DF, McPherson ME, Goldmann DA. Recommendations for evaluation of health care improvement initiatives. Acad Pediatr. 2013;13(6):S23–30. 2. Berwick D. Improvement, trust, and the healthcare workforce. Qual Saf Health Care. 2003;12(90001):2i–6. 3. The joint commission sentinel event policy [Internet]. Jointcommission.org. 2019 Available from: https://www.jointcommission.org/sentinel_event_policy_and_procedures/ 4. 2014 guide to state adverse event reporting systems. AHRQ patient safety network [Internet]. Psnet.ahrq.gov. 2019 Available from: https://psnet.ahrq.gov/resources/resource/29119/2014-guide-to-state-adverse-event-reporting-systems 5. Hollnagel E. The ETTO principle: efficiency- thoroughness trade-off. London: CRC Press; 2009. Chapter 6: ETTOs of the past; p. 103–120. 6. Dekker S. The field guide to understanding ‘human error’. Boca Raton: CRC Press; 2014. Chapter 5: Understanding your accident model; p. 123–141. 7. Reason J. Organizational accidents revisited. Boca Raton: CRC Press; 2016. Chapter 2: The ‘anatomy’ of an organizational accident; p. 9–11. 8. Reason J. The human contribution. Burlington: Ashgate; 2008. Chapter 5: Perceptions of unsafe act; p.69–103. 9. RCA step by step guide [Internet]. Patientsafety.va.gov. 2019 Available from: https://www.patient-safety.va.gov/docs/RCA_Step_By_Step_Guide_REV7_1_16_FINAL.pdf 10. RCA2: improving root cause analyses and actions to prevent harm [Internet]. IHI.org. 2019 Available from: http://www.ihi.org/resources/Pages/Tools/RCA2-Improving-Root-Cause-Analyses-and-Actions-to-Prevent-Harm.aspx 11. From safety-I to safety-II: A white paper. AHRQ patient safety network [Internet]. PSNET.AHRQ.gov. 2019 Available from: https://psnet.ahrq.gov/resources/resource/29228 12. 5 whys: finding the root cause [Internet]. Ihi.org. 2019 [cited 27 January 2019]. Available from: http://www.ihi.org/resources/Pages/Tools/5-Whys-Finding-the-Root-Cause.aspx 13. Braithwaite J, Wears R, Hollnagel E. Resilient health care. Burlington: Ashgate; 2013. Preface; p. xix–xxvi. 14. Dekker S. The field guide to understanding ‘human error’. Boca Raton: CRC Press; 2014. Chapter 1: Two views of human error; p. 1–26. 15. Dekker S. The field guide to understanding ‘human error’. Boca Raton: CRC Press; 2014. Chapter 8: Abandoning the fallacy of a quick fix; p. 191–207. 16. Dekker S. Just culture. 3rd ed. Boca Raton: CRC Press; 2017. Chapter 5: What is the right thing to do?; p. 127–152. 17. Dekker S. The field guide to understanding ‘human error’. Boca Raton: CRC Press; 2014. Chapter 2: Containing your reactions to failure; p. 27–43. 18. Wu A. Effectiveness and efficiency of root cause anal-ysis in medicine. JAMA. 2008;299(6):685. 19. Hollnagel E. Prologue: why do our expectations of how work should be done never correspond exactly with how work is done? In: Braithwaite J, Wears RL, Hollnagel E, editors. Resilient health care. Volume 3: reconciling work-as-imagined and work-as-done. Boca Raton: CRC Press; 2017. p. xvii–xxv. 20. Dekker S. Drift into failure. London: CRC Press; 2011. Chapter 3: The legacy of newton and descartes; p. 51–69. 21. Dekker S. Drift into failure. London: CRC Press; 2011. Chapter 6: What is complexity and systems thinking?; p. 127–168. 22. Weick K, Sutcliffe K. Managing the unexpected: resilient performance in an age of uncertainty. San Francisco: Jossey-Bass; 2007. Chapter 3: The three principles of anticipation; p. 43–64. 23. Hollnagel E. The ETTO principle: efficiency- thoroughness trade-off. London: CRC Press; 2009. Chapter 1: The need to understand why things go wrong; p. 7–21. 24. Cause analysis methods for NERC regional entities and registered entities [Internet]. Nerc.com. 2019 [cited 27 January 2019]. Available from: https://www.nerc.com/pa/rrm/ea/EA%20Program%20Document%20Library/Cause%20Analysis%20Methods%20for%20NERC,%20Regional%20Entities,%20and%20Registered%20Entities_09202011_rev1.pdf 25. Crandall KM, Sten M-B, Almuhanna A, Fahey L, Shah RK. Improving apparent cause analysis reliability: a quality improvement initiative, Kristen M. Crandall, MSN, RN, et al. Pediatr Qual Saf. 2017;2(3):e025. 26. Dekker S. Drift into failure. London: CRC Press; 2011. Chapter 4: The search for the broken compo-nent; p. 71–86. 27. Trajkovski S, Schmied V, Vickers M, Jackson D. Using appreciative inquiry to transform health care. Contemp Nurse. 2013;45(1):95–100. https://doi.org/10.5172/conu.2013.45.1.95. 28. National news briefs; Wrong operation done on young patient. [Internet] 2000. Available from: https://www.

142nytimes.com/2000/12/24/us/national-news-briefs-wrong-operation-done-on-young-patient.html 29. Failure modes and effects analysis (FMEA) tool [Internet]. Ihi.org. 2019 [cited 27 January 2019]. Available from: http://www.ihi.org/resources/Pages/Tools/FailureModesandEffectsAnalysisTool.aspx 30. ASHRM white paper serious safety events: a focus on harm classification: deviation in care as link [Internet]. Ashrm.org. 2019 [cited 27 January 2019]. Available from: http://www.ashrm.org/pubs/files/white_papers/SSE-2_getting_to_zero-9-30-14.pdf 31. Walsh KE, Harik P, Mazor KM, Perfetto D, Anatchkova M, Biggins C, et al. Measuring harm in healthcare: optimizing adverse event review. Med Care. 2017;55(4):436–41. https://doi.org/10.1097/MLR.0000000000000679. 32. Williams T, Szekendi M, Pavkovic S, Clevenger W, Cerese J. The reliability of AHRQ common format harm scales in rating patient safety events. J Patient Saf. 2015;11(1):52–9. https://doi.org/10.1097/PTS.0b013e3182948ef9. 33. Dekker S. The field guide to understanding ‘human error’. Boca Raton: CRC Press; 2014. Chapter 7: Building a safety culture; p. 161–189. 34. Lessons from PSOs on applying the AHRQ common formats for patient safety reporting. Program brief, network of patient safety databases. [Internet]. Pso.ahrq.gov. 2019 [cited 27 January 2019]. Available from: https://pso.ahrq.gov/sites/default/files/wysi-wyg/npsd-common-formats-brief.pdf 35. Lyren A, Coffey M, Shepherd M, Lashutka N, Muething S. SPS leadership group. We will not compete on safety: how children’s hospitals have come together to hasten harm reduction. Jt Comm J Qual Patient Saf. 2018;44(7):377–88. https://doi.org/10.1016/j.jcjq.2018.04.005. Epub 2018 Jun 6. 36. Dekker S. The field guide to understanding ‘human error’. Boca Raton: CRC Press; 2014. Chapter 4: Explaining the patterns of breakdown; p. 89–122. 37. Reason J. Managing the risks of organizational accidents. Burlington: Ashgate; 1997. Chapter 9: Engineering a safety culture; p. 191–222. 38. Communication and optimal resolution (CANDOR) | Agency for healthcare research & quality [Internet]. Ahrq.gov. 2019 Available from: https://www.ahrq.gov/professionals/quality-patient-safety/patient-safety-resources/resources/candor/index.html 39. Dekker S. Just culture. 3rd ed. Boca Raton: CRC Press; 2017. Preface; p. ix–xvi. 40. Dekker S. Just culture. 3rd ed. Boca Raton: CRC Press; 2017. Chapter 3: Safety reporting and honest disclosure; p. 61–90. 41. Weick K, Sutcliffe K. Managing the unexpected: assuring high performance in an age of complex-ity. San Francisco: Jossey-Bass; 2001. Chapter 3: A closer look at process and why planning can make things worse; p. 51–84. 42. McLeod RW, Bowie P. Bowtie analysis as a prospec-tive risk assessment technique in primary healthcare. Pol Pract Health Saf. 2018;16(2):177–93. 43. Society to improve diagnosis in medicine [Internet]. Society to improve diagnosis in medicine. 2019 [cited 27 January 2019]. Available from: https://www.improvediagnosis.org/ 44. Improving diagnosis in health care: health and medi-cine division [Internet]. Nationalacademies.org. 2019 [cited 27 January 2019]. Available from: http://www.nationalacademies.org/hmd/Reports/2015/Improving-Diagnosis-in-Healthcare.aspx 45. Hollnagel E. The ETTO principle: efficiency- thoroughness trade-off. London: CRC Press; 2009. Chapter 2: From rationality to ETTOing; p. 23–40. 46. Weick K, Sutcliffe K. Managing the unexpected: sus-tained performance in a complex world. Hoboken: Wiley; 2015. Chapter 6: Commitment to resilience; p. 94–111. 47. Agency for Healthcare Research and Quality: AHRQ.gov 48. Erik Hollnagel website: www.erikhollnagel.com 49. Institute for Healthcare Improvement: IHI.org 50. Society to Improve Diagnosis in Medicine: https://www.improvediagnosis.org 51. Safety Differently: The Movie: https://youtu.be/moh4QN4IAPg 52. Teaching and Assessing Critical Thinking: https://medicine.dal.ca/departments/core-units/cpd/faculty-development/programs/TACT.htmlM. Green and L. E. Budin

143© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 R. K. Shah, S. A. Godambe (eds.), Patient Safety and Quality Improvement in Healthcare, https://doi.org/10.1007/978-3-030-55829-1_7Communication with Disclosure and Its Importance in SafetyKristin Cummins, Katherine A. Feley, Michele Saysana, and Brian WagersK. Cummins Children’s Health, Dallas, TX, USAe-mail: [email protected] K. A. Feley Indiana State Nurses Association, Indianapolis, IN, USA M. Saysana (*) Indiana University Health; Indiana University School of Medicine, Indianapolis, IN, USAe-mail: [email protected] B. Wagers Indiana University School of Medicine, Riley Hospital for Children, Department of Emergency Medicine and Pediatrics, Indianapolis, IN, USAe-mail: [email protected] Objectives• To demonstrate the critical role of com-munication in contributing to and pre-venting medical errors• To demonstrate the structure and pro-cesses needed to support effective communication• To demonstrate how effective bidirec-tional communication drives a culture of safetyVignette 7.1An 8-month-old child required extracorpo-real membrane oxygenation (ECMO) due to progressively worsening respiratory sta-tus. The child was admitted with a diagno-sis of respiratory failure, and the team had increasing difficulty oxygenating the patient. Given the patient’s illness severity and potential for a good resolution of symptoms with proper support, the deci-sion was made to place the patient on ECMO.This patient was located in the pediatric intensive care unit (PICU), and given the instability of the patient, the decision was made to perform cannulation of the blood vessels and initiation of ECMO at the patient’s bedside. Per the surgical team’s routine practice, the ECMO cart was read-ied outside the patient’s room with the materials that are required for placement of the ECMO cannulas into the blood vessels of the patient. It is routine practice in this institution to place both the size of cannula the surgeon estimates will be required for successful oxygenation and filtration of the patient’s blood and the next size down on the cart holding the materials. This estima-tion is made based on the child’s weight,

144Many adverse events involve communication difficulties. Approximately 30% of adverse events in the operating room and 70% of sentinel events, in general, involve a breakdown in effec-tive communication [1]. Studies completed in the relatively controlled setting of the operating room still indicate that interruptions, distractions, and provider stress contribute to communication errors [2]. The case in the vignette was performed at the bedside in an inherently stressful environ-ment where the patient required bedside ECMO cannulation. The chaos surrounding emergent cannulation stands in direct contrast to the con-trolled and planned environment of most proce-dures completed in the operating room.This case highlights several communication- related opportunities for improvement among the team during the event itself. The first opportunity is in pre-procedure/event preparation. A team should always prepare prior to an event. Even in an emergent situation, a team has the luxury of but anatomical differences can cause this estimation to fail, and at times, a catheter size smaller than originally estimated is required. Both of these cannulas are placed on the cart, so they will be immediately available to the surgeon performing the procedure.The surgeon verified the cannula size needed outside the room and then returned to the patient’s bedside to perform the pro-cedure. It is the team’s routine to place the desired size of catheter as well as the next size down on the sterile field so the equip-ment is at hand should vein size dictate a smaller catheter be placed. The scrub nurse placed the 27 French catheter on the porta-ble table next to her while awaiting the pro-cedure, but this table was moved out of the room unbeknownst to her to make room for the portable fluoroscopy machine. The sur-geon took the catheter that remained on the surgical table and proceeded to cannulate the patient. The catheter size was not veri-fied prior to placement into the child’s blood vessel. In the operating room, a time- out is completed prior to surgery verifying the necessary materials and procedure to be completed. Outside the operating room, this is not a routine practice. In this case, no time-out was completed, and the patient was cannulated successfully with what the surgeon thought was a 27 French catheter and placed on the ECMO circuit. The actual catheter size was a 23 French cathe-ter. The surgical team then left the PICU to perform another case in the operating room.It was immediately noticed by the ECMO technician that the catheter placed into the blood vessel was not a 27 French catheter but was the next size down. The decision was made by the cardiac intensiv-ist to attempt to use this catheter to provide adequate therapy. Throughout the night the ECMO technicians had difficulty maintain-ing appropriate blood flow volume through the ECMO circuit and thus struggled to provide the highest level of care to the patient.The next morning, the ECMO team relayed the difficulties in achieving proper blood flow to the cardiac intensivist. The surgeon was also notified, and the decision was made to replace the 23 French cannula with a 27 French cannula. The family was notified of the error in placing the incorrect size catheter and the resultant difficulty with blood flow it caused. They agreed to have the second procedure to place the cor-rect size catheter. The team completed this procedure which required the patient to be taken off the ECMO circuit for approxi-mately 5 minutes, and during this time the patient had a medical arrest which required a short period of cardiopulmonary resusci-tation with return of spontaneous circula-tion within 5 minutes. The child did well once the new catheter was inserted and was able to be weaned off ECMO at a later date with no apparent lasting harm.

145taking a few seconds to brief on the procedure, necessary equipment, possible complications/contingencies, and expected sequence of events. In this case, the team did not have this opportu-nity to “pre-brief” given the unstable nature of the child and the perceived necessity of a quick decision to place the child on extracorporeal membrane oxygenation. Regardless of the situa-tion, there is always time to ask for quiet in a pro-cedural space and for extraneous conversation to be held outside the immediate procedure area. This strategy of asking for “quiet space” can effectively reduce some of the chaos and distrac-tion common in high-risk situations with clini-cally deteriorating patients [3]. Some organizations have employed similar strategies regarding medication practices. For instance, it is common to have a “protected zone” around the automated medication-dispensing machine where nurses can focus solely on selecting the proper medication prescribed and drawing up the correct ordered dosage. Often these areas are marked off by signage or tape on the floor, and other staff members are educated to not interrupt or distract the staff member involved in work requiring critical focus in that space (Key Points Box 7.1).One communication strategy that the team did not use in this case was a pre-procedural time- out. The pre-procedural time-out reduces intra-operative errors [3]. One component of most time-outs is the inclusion of the specific equip-ment a practitioner requires to perform the proce-dure. In this case, had the team used a time-out and specified that the size of ECMO catheter was to be a 27 French, the scrub nurse would have realized that only the 23 French catheter was on the procedure cart and could have ensured the proper size were present (Key Points Box 7.2).A vital aspect of time-outs is the closed-loop communication that they facilitate and require. Had the surgeon asked for a 27 French catheter explicitly and the scrub nurse confirmed they provided a 27 French catheter back to the sur-geon, the inadvertent placement of the incorrect size catheter would have been avoided. One must use techniques to decrease error such as reading back an order that is given to the provider or ask-ing clarifying questions to clearly delineate what is being asked of an individual. In this case, if the surgeon had paused to ask their colleague hand-ing them the catheter to confirm that it was, in fact, the 27 French size they desired, the errone-ous placement would have been avoided.When physicians foster an environment in which they are open to others questioning them, this helps breakdown perceived power hierar-chies between team members. The hierarchy in medicine can contribute to error because team members may not be comfortable speaking up and reporting problems in a timely fashion [4]. In the apparent cause analysis regarding this case, the team members related that there was no inter-play of questions between the involved individu-als, some of which may have been due to the hierarchical nature of the service that performed the procedure.It is well known that hierarchy can be a detri-ment to safety culture [5]. In the past, the power gradient present between more senior leaders and direct reports have led to deadly consequences in industries outside of healthcare with one of the best-known failures was the KLM 4805 flight col-Key Points Box 7.1A pre-procedural time-out or a pre-event brief can help to orient the team to the expected facets of a procedure, the neces-sary equipment, and potential areas of risk to the patient. Many studies link pre- procedural time-outs or pre-event briefs to improved outcomes for patients [3].Key Points Box 7.2 Components of the Time-out 1. Verify the correct patient. 2. Verify the correct procedure. 3. Verify the correct site.

146lision in the Canary Islands in 1977 that killed 583 people in healthcare [6]. In this example, the junior pilot knew that the pilot’s attempt to take off was an error but did not challenge the senior pilot due to cultural norms and deference to seniority. This silence was a contributing factor in this tragedy. Organizations that seek to employ high reliability principles in their safety work must seek to break down hierarchical power structures so that all members of an organization feel empowered to speak up and make patient safety threats known. An organization that encourages all members of the team to make safety threats known immedi-ately can lessen the risk of events occurring and therefore lessen the chance of harm to patients.Once an error occurs such as using the wrong size cannula in our case, it is our professional duty to disclose this to the patient and family. Many professional organizations such as the Joint Commission and the American College of Physicians endorse the practice of disclosure as a professional and ethical duty after an error occurs [7–9]. Patients and families also expect that we will be transparent and honest with them if an error occurs. It is also important to note that dis-closure is a process and usually occurs over time as more information is available to help under-stand why an error occurred.Before we go further in the discussion of dis-closure, it is necessary to define what disclosure is. Full disclosure includes an acknowledgment that an error occurred as well as an explanation of the error and connection between the error and harm to the patient and further treatment to miti-gate the error [10, 11]. Patients and families also want to know how the organization will prevent this from happening again. Of course, disclosure should be done in a way that the patient and fam-ily understand the event and its effects.Studies have shown that barriers are still pres-ent to providing full disclosure to patients and families usually because of fear of malpractice [12]. Hospitals and health systems can offer sup-port to physicians through their patient safety, patient advocate, and/or risk management depart-ments to help guide physicians on how to do dis-closure. Petronio et al. describe a two-step process called the Mistake Disclosure Management Plan (MDMP) for disclosure. The first step is to prepare the physician and the sec-ond step is mistake disclosure strategies. The preparation step considers the emotional impact of the error on the physician and also involves investigating the error to understand how it occurred. The mistake disclosure strategies step considers the timing of the disclosure, the people included in the disclosure, as well as the steps of disclosure including the event and an apology [12]. When an error occurs in a pediatric patient, parents determine whether or not their child should be present for the disclosure [13] (Key Points Box 7.3).Disclosing an error especially if there was harm to the patient is essentially delivering bad news such as a diagnosis of a new illness. As with delivering bad news, physicians need to prepare that patients and families may react emotionally and need time to process the information. Patients and families should be given time to process the information and the opportunity to seek clarifica-tion. This may even happen after the initial dis-closure conversation has occurred. The physician should avoid blaming others or making excuses for the error as this may further erode trust in the hospital. The physician should also avoid specu-lating or jumping to conclusions as to why it hap-pened especially if an investigation is still underway at the time of the initial disclosure con-versation. Patients and families feel strongly that the attending physician involved in the error do the disclosure with the patient and family. This helps maintain trust in the physician and team. Key Points Box 7.3Mistake Disclosure Management Plan is a two-step process for disclosure including a preparation step and a strategy step. The preparation step considers the emotional impact on the physician as well as the error investigation. The strategy step includes the timing of disclosure, people included in the disclosure, and the steps of disclosure.

147Patients and families also want to know that the institution and physician take this event seriously and are committed to improving safety and preventing the error from occurring again. This may mean they want to know about specific improvements put in place to prevent a recur-rence of an event for themselves and other patients [14].Lastly, a key piece of disclosure is an apol-ogy. Patients and families appreciate an authen-tic apology because it is an act of empathy. As of December 2018, 39 states as well as the District of Columbia have apology laws to support med-ical professionals in apologizing to patients and families when something unexpected happens such as a medical error with harm. These laws help prevent saying “I’m sorry” to be used against a physician in a medical malpractice case [15, 16].Some health systems including the University of Michigan Health System have adopted a communication- and-resolution program (CRP) to disclose unexpected patient outcomes either from complications or medical error. CRP includes disclosure and apology to the patient and family quickly while also investigating the outcome. If the investigation reveals a deviation in the standard of care, then the institution offers a financial settlement to the patient and family and makes system improvements to prevent the outcome from transpiring in the future. If the care provided was appropriate, then the institu-tion shares the findings with the patient and family and defends the physician if litigation ensues [16].Key Points Box 7.4.Effective communication plays a crucial role during the investigation and interview process, and also after the root cause(s) has been identi-fied to close the feedback loop. Lack of or insuf-ficient communication can also be a cause of an error. In the apparent cause analysis regarding the case of the wrong size cannula placement, the interviewed team members relayed the lack of communication was part of the root cause of the event that led to a secondary procedure to replace the cannula.Vignette 7.2The members of the team realized several opportunities for improvement in the practice of ECMO cannulation and desired to perform a cause analysis. The team members performed this analysis with all team participants including the surgeon who placed the catheter and the cardiac intensivist. They determined the gaps in practice that allowed the mistake to occur. One key gap identified was the omission of the time- out process. The time-out was not completed because it is not routine to complete a time-out when not in the operating room despite the staff members all being surgical staff. This realization led to an organization- wide decision to require a time-out whenever a procedure is done, regardless of physical location. This decision was disseminated through presentations at the surgical mor-bidity and mortality conference, the ECMO morbidity and mortality confer-ence, and the hospital-wide surgical qual-ity assurance conference.Key Points Box 7.4Response to a patient harm or near miss event should first be to establish patient safety; second to sequester any equipment, devices, or products involved; and to begin an investigation or review of the event [17]. The purpose of an investigation is to gain an understanding of what led to the event’s occurrence and to assist in determining an apparent or root cause(s). By identifying the cause, corrective and preventative actions should be set to proactively prevent the recurrence of the same or similar event.

148Just as closed-loop communication may have prevented the error from occurring in our case, closing the communication loop, known as a feedback loop of a root cause analysis or appar-ent cause analysis is just as important. Communicating outcomes of an incident analysis should occur with those involved, those who reported, those that may be affected in the current state and future, and especially team members and leaders held accountable for implementing recommendations as determined in the analysis [18] (Key Points Box 7.5).Participating in an investigation after being involved in an event can be intimidating, and even frightening. Verbal communication, from the interviewer to interviewee, should explain the purpose of the investigation, not assign blame and clearly communicate that the interview is being conducted to identify system issues or vul-nerabilities [17]. Effective communication tech-niques of the interviewer include active listening, open questioning, and paraphrasing to verify what was heard (Key Points Box 7.6).Information collected during the interview process is assembled as a visual tool – examples include process mapping and cause and effect or fishbone diagram – used to communicate the event flow and contributing factors, and used to highlight gaps or opportunities for improvement. Clear and concise delegation of action items to responsible parties can be considered part of the feedback loop after an event.All patient safety issues may not lead to a full investigation and analysis process. Incidents sub-mitted through electronic reporting or paper methods also require feedback loop communica-tion. For example, if actions resulting from sub-mitted incident reports were shared with the original submitters, they would gain a better understanding of the potential benefits to future patients and the health system of increased and timely event reporting. They would also see that these benefits outweigh the challenges of report entry and the associated risks. The timely report-ing and resolution of problems is integral to jour-ney to high reliability which is discussed in other chapters of this text [19].The importance of communicating medical errors throughout the organization cannot be underestimated. According to the Lucian Leape Institute, established by the National Patient Safety Foundation, transparency is the “most important single attribute of a culture of safety” Key Points Box 7.5 When Communicating Outcomes of an Incident Analysis, Make Sure to Communicate to 1. Individuals who were involved in the event 2. Individuals who reported the event (if they were not involved in the event) 3. Individuals who may be affected by the current and future state 4. Individuals held accountable for imple-menting the recommendations from the analysisKey Points Box 7.6 Interviewer Techniques for Effective Communication Include the Following 1. Active listening. 2. Open questioning 3. Paraphrasing to verify what was heardVignette 7.3The safety event classification team reviewed this case at a regularly scheduled meeting. The ECMO and perioperative team involved presented their report on the event and made recommendations for prac-tice improvements to ensure this type of error would not happen in the future. These recommendations were publicized through-out the organization. Appropriate reports were made to the state as this case involved an improper implant which requires report-ing in this hospital’s state.

149[20]. Healthcare organizations with strong safety cultures are transparent in the sharing of medical errors because they aim to prevent future events. In the absence of transparency, distrust and hos-tility permeate the organization [20]. The land-mark publication To Err Is Human shed light on medical errors by highlighting that the majority of errors do not result from individual negligence but rather are caused by broken systems that inadvertently set caregivers up to make mistakes [21]. Healthcare leaders must encourage and reward frontline team members for reporting near misses to identify possible broken systems. Early reporting helps prevent future errors which may have devastating effects on patients especially if the error reaches the patient and causes irrevers-ible harm [22]. As healthcare leaders have increased their knowledge around medical error causation, organizations have begun encouraging caregivers to share events to identify necessary system and process improvements [21].Healthcare leaders must appreciate that orga-nizations often sustain collective harm in self- esteem and confidence following significant medical errors [23]. Witnessing or hearing about medical errors reminds us of our own fallibility and the delicate nature of the procedures and treatments we routinely perform on patients every day. Caregivers must feel safe in their envi-ronment to openly discuss medical errors, and they must believe they will be treated fairly for disclosing mistakes [24]. Healthcare leaders con-tribute to building this environment and earning caregiver trust by compassionately communicat-ing medical errors and supporting those involved. The transparent communication of medical errors is essential to promote healing and performance improvement throughout the organization [23].When preparing to communicate a medical error throughout the organization, it is vital to consider who will be communicating the mes-sage, what will be communicated, and how it will be communicated. Individuals involved in the medical error communication should know the event well and understand the key learnings. The individuals should communicate the medical error in a sincere, compassionate, factual manner. Individuals must reliably communicate the event to build trust and collegiality among team mem-bers. Individuals should not invoke their personal opinions or judgments into the report. Ideally, the individual or team communicating the medical error should also understand systems failures to prevent delivering a message of blame and shame. The team communicating the medical error should partner with the hospital or depart-ment leadership, quality and safety leadership, as well as risk management and/or the legal team, to provide input into the message. Often, quality and safety leaders communicate medical errors at various councils throughout the organization and are very effective at doing so. However, leaders should not miss the opportunity to allow those involved in the medical error to participate in communicating the event if they wish because doing so keeps them a part of the learning and reduces their feelings of isolation [24]. Team members involved in events often grieve and need to be included in the solution. By involving the affected team members when sharing an event, those involved understand the process bet-ter and often find support from their colleagues. This may be critical to keep them engaged in their profession and prevent them from leaving healthcare (Key Points Box 7.7).The National Patient Safety Foundation advises medical errors should be communicated with the goal of improving care [24]. Therefore, individuals must consider the appropriate places to share the medical error, the purpose of sharing the medical error, and how much detail is needed to effectively communicate. For example, signifi-cant medical errors should be communicated to the healthcare organization’s executive leader-ship team and quality board because they are Key Points Box 7.7 Event CommunicationWhen communicating an event, ensure that those involved with the event – hospital and department leadership, quality and safety leadership, as well as risk management and/or the legal team – provide input into how to communicate the event and follow up.

150responsible for prioritizing transparency, safety, and continuous improvement [24]; in the quality improvement parlance, this is referred to as “spread.” This can be accomplished by showing pictures or taking them to the clinical space where the event occurred to show them how it could happen. This can be a very powerful expe-rience for leaders and board members. In general, communication at this level involves a general overview of the error, contributing factors, and strategies to prevent the medical error in the future. It is also important to assess the clinical knowledge base of those on the executive team or board when you are communicating. When the ECMO case was presented to the quality board, board members immediately questioned why two different sizes of ECMO cannulas were even available in this situation. From their viewpoint, having only one size catheter available would be an easy fix. However, the clinical leader present-ing the case was able to paint a picture to the board of what it looked like to connect a patient to ECMO and how complicated the procedure was. The clinical leader was able to explain how not having both sizes available would be detri-mental to the patient if the team had to run throughout the hospital to find another size and how it is not a rare occurrence to need a different size. The board understood the complexity of the situation following the explanation and gained a better understanding of why time-out procedures are critical and need to be hardwired outside of the operating room (Key Points Box 7.8).Medical errors should also be shared with frontline caregivers to promote vigilance and identify system and process improvements [23]. Hospital quality councils or morbidity and mor-tality conferences often serve as the venues for medical error communication. Some organiza-tions have scheduled quarterly or monthly sharing of events to ensure the learnings are spread. Individuals sharing the case in these venues often provide more detail around the event including a synopsis of the patient’s clinical presentation, happenings leading up to the event, the event itself, the patient outcome, causative factors, sug-gested mitigation strategies, and how the event impacted the clinical team. Caregivers directly involved in the event should be made aware that the case will be discussed in the venue. When medical errors occur, caregivers often lose confi-dence in themselves and still feel accountable even if the case is treated and discussed from a systems perspective [25]. These caregivers have been referred to as second victims and often expe-rience significant emotional turmoil after the event and need support from their colleagues [26, 27]. Therefore, involved caregivers should be included in the communication process, and their concerns should be addressed before releasing the event information. Not doing so will compromise transparency efforts [21]. Moreover, most care-givers involved in medical errors want to contrib-ute to future prevention efforts, so it is worth the extra effort to ensure they are treated with com-passion and respect throughout the communica-tion process [23, 25] (Key Points Box 7.9).Event sharing in these venues often leads to a robust discussion about the event and potential system fixes, so these conversations need to be facilitated by informed, well-prepared individu-als identified in advance. In today’s environment, these discussions are often very supportive and Key Points Box 7.8 Spread the MessageIndividuals must consider the appropriate places to share the medical error, the pur-pose of sharing the medical error, and how much detail is needed to effectively communicate.Key Points Box 7.9 Second VictimsCaregivers who were directly involved in the event often lose confidence in them-selves and still feel accountable even if the case is treated and discussed from a sys-tems perspective [25]. They often experi-ence significant emotional turmoil after the event and need support from their colleagues.

151even therapeutic, but the facilitator must be pre-pared to address and discourage any comments that are hurtful or discouraging to the caregivers involved. The facilitator’s primary focus during these discussions should be to maintain a safe environment for everyone to discuss the medical error; not doing so will quickly erode the trust of the caregivers [23]. Healthcare leaders should keep in mind these discussions are critical to identify future medical error mitigation strategies and to promote a safe culture, but the discussions will only be effective if caregivers feel comfort-able discussing cases. When the ECMO event was discussed in the hospital quality council, the clinical team involved in the event presented the case and were relieved to find themselves sur-rounded with support from their colleagues. The council also agreed to support hardwiring the time-out process throughout the hospital follow-ing the discussion of the event.Quality and safety leaders should also report medical errors to their patient safety organizations (PSOs) if they are involved in one. There is power in reporting significant medical errors to the PSO because the data is compiled and analyzed with other like organizations and trends are often identi-fied that would not have been identified at the indi-vidual facility level. The Patient Safety and Quality Improvement Act of 2005 (Public Law 109-41) has enabled the creation of PSOs and provides federal legal protection to information reported to a PSO for the purpose of improving patient safety [28]. The event investigation information gathered and reported to the PSO is called “patient safety work product (PSWP)” [28]. It is important for quality and safety leaders to understand this protection to address any concerns their organization has with sharing this important information with the PSO; of course, we suggest this be done in collaboration and with engagement from your organization’s risk management team (Key Points Box 7.10).Communication is woven throughout health-care delivery and is critical to ensuring the highest quality and safest care. As illustrated by the case vignette, breakdowns in communication can lead to unintended outcomes and preventable harm. However, communication must be leveraged to help patients, families, and team members recover after the event and be used to help prevent future events. Strong communication only enhances quality improvement and patient safety efforts to make systems safer and more reliable. The com-munication of medical errors is critically impor-tant in promoting a culture of safety. Healthcare leaders should not underestimate the power of transparency in preventing future harm.Key Points Box 7.10Caregivers often appreciate knowing their organization shares safety events with other organizations to prevent harm beyond their walls.Editors’ CommentsThe prior chapter helped the reader under-stand how to respond when an event occurs; this chapter builds on the prior chapter by going deep on the topic of communication and disclosure after an event. A key phrase by the authors in Chap. 7 is “bidirectional communication” – by understanding and communicating with front line and families and subsequently listening to them, trust can be built and lead to a successful resolution.The authors cite the commonly known literature that attributes communication breakdowns to harm. This is perhaps one of the key learnings of the chapter and one that the editors of the textbook have seen time and time again in their institutions and when reviewing events that have transpired in other organizations. As the authors astutely point out, there are myriad forms of communication and breakdowns that can occur. The corollary is that there are many opportunities for communication to help and ameliorate an issue – the chapter highlights several of these (e.g., time-out).A significant portion of this chapter explains the importance of disclosure and how to properly communicate such. It is not as simple as saying “sorry.” The authors explain the role of a Mistake Disclosure

152 Chapter Review Questions 1. Effective communication/interview tech-niques during an event review include: A. Active listening B. Using open-ended questions C. Paraphrasing what was heard D. All of the aboveAnswer: D. Explanation: Verbal communi-cation, from the interviewer to interviewee, should explain the purpose of the investiga-tion, not assign blame and clearly communi-cate that the interview is being conducted to identify system issues or vulnerabilities [17]. Effective communication techniques of the interviewer include active listening, open questioning, and paraphrasing to verify what was heard. 2. Closed-loop communication should be used to: A. Reduce misunderstandings. B. Keep the conversation between two individuals. C. Reduce unnecessary dialogue. D. Convey “you” statements.Answer: A. Explanation: Closed-loop com-munication is used to clearly communicate information and should explain the purpose of an event, reduce misunderstandings, and can occur in a team setting. It should be non- judgmental as well. 3. Full disclosure of an error includes the follow-ing except: A. Acknowledgment that an error occurred B. Explanation of the error and harm it caused C. Blaming the person who committed the error D. Treatment plan if harm occurredAnswer: C. Explanation: Full disclosure includes an acknowledgment that an error occurred as well as an explanation of the error and connection between the error and harm to the patient and further treatment to mitigate the error. Blaming the individual who com-mitted the error is not productive and not part of the full disclosure process. 4. Response to patient harm involves which of the following: A. Establish patient safety. B. Sequester any equipment, devices, or products involved. C. Begin an investigation or review of the event. D. All of the above.Answer: D. Explanation: Response to a patient harm or near miss event should first be to establish patient safety; second to sequester any equipment, devices, or products involved; and to begin an investigation or review of the event. The purpose of an investigation is to gain an understanding of what led to the event’s occurrence and to assist in determin-ing an apparent or root cause(s). 5. Significant medical errors leading to harm should be shared with all of the following: A. Patient involved. B. Frontline staff C. Executive leadership D. Board of directors E. Patient safety organization F. All of the aboveManagement Plan which is an excellent resource that organizations should consider having ready to use when necessary.The chapter also provides a primer on how to communicate through an organiza-tion regarding an event that has occurred. In our institutions, we inform the entire hospital as well as the board when a signifi-cant error or similar instance occurs. The authors nicely take the reader through steps that need to be considered when reporting through an organization and beyond as well as to when report.As we have seen in other chapters, the authors do again discuss the importance of communication with regard to those affected or involved with the error, issue, etc. As organizations continue to advance their culture toward high reliability, we must always be aware of those involved with the issue.

153Answer: F. Explanation: Significant medi-cal errors leading to harm should be commu-nicated with all of the above parties to ensure that the patient receives the appropriate treat-ment in response to the error and to prevent the error from happening again in the institu-tion as well as other institutions.References 1. Lingard LS, Espin S, Whyte G, Regehr G, Baker GR, Reznick R, et al. Communication failures in the operating room: an observational classification of recurrent types and effects. Qual Saf Health Care. 2004;13:330–4. 2. Leonard M, Graham S, Bonacum D. The human fac-tor: the critical importance of effective teamwork and communication in providing safe care. Qual Saf Health Care. 2004;13:85–90. 3. Pelligrini C. Time outs and their role in improving safety and quality in surgery. Bulletin of the American College of Surgeons. June 1, 2017. Online. Available: http://bulletin.facs.org/2017/06/time-outs-and-their-role-in-improving-safety-and-quality-in-surgery/. Accessed 22 Feb 2019. 4. Dingley C, Daughtery K, Derieg M, Persing R. Improving patient safety through provider com-munication strategy enhancements. In: Henrikesen K, Battles JB, Keyes MA, Grady ML, editors. Advances in patient safety: new directions and alternative approaches, vol. 3. Rockville: Agency for Healthcare Research and Quality; 2008. p. 1–18. 5. AHRQ. Patient safety primer: culture of patient safety. Online. Available: https://psnet.ahrq.gov/primers/primer/5/Culture-of-Safety. Accessed 22 Feb 2019. 6. Federal Aviation Administration. Lessons learned from civil aviation accidents. KLM 4085 Collision with Pan AM 1736 at Tenerife. Online. Available: https://lessonslearned.faa.gov/ll_main.cfm?TabID=1&LLID=52&LLTypeID=2. Accessed 22 Feb 2019. 7. The Joint Commission. Comprehensive accredita-tion manual for hospitals: the official handbook. Oak Brook: Joint Commission Resources; 2017. 8. Sulmasy L, Bledsoe TA, ACP Ethics, Professionalism and Human Rights Committee. American college of physicians ethics manual: seventh edition. Ann Intern Med. 2019;170(2_Supplement):S1–S32. 9. AMA Code of Medical Ethics’ opinions on patient safety. Virtual Mentor. 2011;13(9):626–8. 10. Fein SP, Hilborne LH, Spiritus EM, et al. The many faces of error disclosure: a common set of elements and definition. J Gen Intern Med. 2007;22(6):755–61. 11. Agency for Healthcare Research and Quality. Disclosure of error. Available from: https://psnet.ahrq.gov/primers/primer/2/Disclosure-of-Errors. Accessed 25 Jan 2019. 12. Petronio A, Torke A, Bosslet G, Isenberg S, Wocial L, Helft PR. Disclosing medical mistakes: a com-munication management plan for physicians. Perm J. 2013;17(2):73–9. 13. Coffey M, Espin S, Hahmann T, Clairman H, Lo L, Friedman JN, et al. Parent preferences for medical error disclosure: a qualitative study. Hosp Pediatr. 2017;7(1):24–30. 14. Moore J, Bismark M, Mello MM. Patients’ expe-riences with communication-and-resolution pro-grams after medical injury. JAMA Intern Med. 2017;177(11):1595–603. 15. National Conference of State Legislatures. Medical professional apologies statutes. Available from: http://www.ncsl.org/research/financial-services-and-com-merce/medical-professional-apologies-statutes.aspx. Accessed 27 Jan 2019. 16. Mello MM, Boothman RC, McDonald T, Driver J, et al. Communication-and-resolution programs: the challenges and lessons learned from six early adopt-ers. Health Aff. 2014;33(1):20–9. 17. National Patient Safety Foundation. RCA2 improv-ing root cause analyses and action to prevent harm. Boston: National Patient Safety Foundation; 2015. 18. Peerally MF, Carr S, Waring J, Dixon-Woods M. The problem with root cause analysis. Qual Saf Health Care. 2017;26:417–22. 19. Nuckols TK. Incident reporting: More atten-tion to the safety action feedback loop, please. Available from: https://psnet.ahrq.gov/perspectives/perspective/108/Incident-Reporting-More-Attention-to-the-Safety-Action-Feedback-Loop-Please?q=feedback+loop%20-%20ref17. Accessed 25 Jan 2019. 20. Leape L, Berwick D, Clancy C, Conway J, Gluck P, Guest J, et al., Lucian Leape Institute at the National Patient Safety Foundation. Transforming health-care: a safety imperative. Qual Saf Health Care. 2009;18(6):424–8. 21. Kohn LT, Corrigan JM, Donaldson MS, editors. To err is human: building a safer health system. Washington, DC: National Academies Press; 2000. 22. Rothschild JM, Landrigan CP, Cronin JW, Kaushal R, Lockley SW, et al. The critical care safety study: the incidence and nature of adverse events and seri-ous medical errors in intensive care. Crit Care Med. 2005;33(8):1694–700. 23. Denham CR. TRUST: the 5 rights of the second vic-tim. J Patient Saf. 2007;3(2):107–19. 24. National Patient Safety Foundation, Lucian Leape Institute. Shining a light: safer health care through transparency. Boston: National Patient Safety Foundation; 2015. 25. Kachalia A. Improving patient safety through trans-parency. N Engl J Med. 2013;369(18):1677–9. 26. Wu AW. Medical error: the second victim. BMJ. 2000;320(7237):726–7.

154 27. Scott SD, Hirschinger LE, Cox KR, McCoig M, Brandt J, Hall LW. The natural history of recov-ery for the healthcare provider “second victim” after adverse patient events. Qual Saf Health Care. 2009;18(5):325–30. 28. Agency for Healthcare Research and Quality. The patient safety and quality improvement act of 2005. Available from: http://www.ahrq.gov/policymakers/psoact.html. Accessed 25 Jan 2019.

155© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 R. K. Shah, S. A. Godambe (eds.), Patient Safety and Quality Improvement in Healthcare, https://doi.org/10.1007/978-3-030-55829-1_8Using Data to Drive ChangeLisa L. Schroeder Opening Question/ProblemThis chapter is not about specific tactics to reduce the number of patients leaving prior to being seen in the ED, nor is it about teaching statistics; it is L. L. Schroeder (*) Children’s Mercy Kansas City, University of Missouri-Kansas City School of Medicine, Department of Medical Administration, Kansas City, MO, USAe-mail: [email protected] 8.1A Pediatric Emergency Department (ED) and Level 1 Trauma Center with over 67,000 visits annually admits approxi-mately 9% of presenting patients. As one of Chapter Objectives• To describe the importance of data in the design, implementation, and assess-ment of improvement work• To describe key differences in qualita-tive versus quantitative data and the advantages of each in driving improvement• To explain the differences in informa-tion gained from run charts versus con-trol charts• To understand the value of Pareto charts and scatter plots in the improvement processthe hospital’s board metrics, the ED tracks the proportion of patients leaving without being seen (LWBS) by a provider. While the daily LWBS rates are often low (<1%), some days the rate reaches over 10% of patients presenting for care, and averages have been running at approximately 3.3% (over 2000 patients/year). Benchmark hos-pitals consistently record rates of less than 2.5% with low variability. ED leadership has followed this number for years but has had difficulty driving change, partially due to the multiplicity of factors that they believe drive the LWBS rates, including seasonality, acuity, overall numbers of patients daily, hospital inpatient census, staffing, efficiency of providers, etc. Hospital administrators have challenged the ED to reduce these rates, emphasizing that the emergency department is the pri-mary portal of entrance for the majority of admissions and should be the first choice for parents with sick or injured children.

156about learning to utilize data in several different formats to help assess situations, develop action items, and track progress toward incremental improvement goals. Providing frontline caregiv-ers actionable, visible, meaningful data around team-level goals supports improvement initia-tives that drive change.It is a capital mistake to theorize before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes [1]) Data as Foundational to ImprovementHealthcare represents a complex, ever-evolving, and sometimes unpredictable system. Driving change in such a system can feel daunting and perhaps even unachievable at times. In an attempt to provide some direction for the assessment of quality of care, Avedis Donabedian proposed a shift from focusing primarily on the outcome of medical care in defining quality to focusing on a triad of measures, comprised of structure, pro-cess, and outcome measures in his 1966 land-mark article “Evaluating the Quality of Medical Care” [2]. In this foundational model, “structure” refers to the resources available, qualifications of providers, governance, etc., “process” relates to the way the work is done or the components of care, and “outcome” indicates the result achieved. He further points out that focusing on outcomes as measures of quality, while important, must be done with discrimination. The use of outcomes alone to validate the quality of care contains inherent risks. First, not all outcomes are easily measured. While results such as mortality are straightforward, parameters such as satisfaction or quality of life may not be clearly defined. Additionally, the question of relevance comes into play when considering outcomes as mea-sures of quality. For example, survival as a sole criterion of success may not be appropriate in situations that produce severe crippling condi-tions or poor quality of life. Finally, measuring outcomes alone does not provide insight into the factors that may have attributed to the overall out-come. In response to this message, Dr. Donald Berwick suggested that the priority of measuring quality should not hinge solely on end results, but rather “a more subtle interplay among structure, process and outcome” [3].Donabedian goes on to conclude that in order to drive change “more often one needs to ask, ‘What goes on here?’ rather than ‘What is wrong; and how can it be made better?’ This does not mean that the researcher disowns his own values or social objectives. It does mean, however, that the distinction between values and elements of structure, process or outcome is recognized and maintained; and that both are subjected to equally critical study. Partly to achieve this kind of orien-tation, emphasis must be shifted from preoccupa-tion with evaluating quality to concentration on understanding the medical care process itself” [2]. As more emphasis is placed on quality improvement in healthcare and more resources are focused on eliminating harm, it is important to understand the impact of initiatives that are proposed or put into place. This understanding is best achieved through the gathering and interpre-tation of data, in both qualitative and quantitative forms. Merriam Webster defines data as “factual information (such as measurements or statistics) used as a basis for reasoning, discussion, or cal-culation” [4]. Although various models to guide improvement have been described over the years, each includes the critical need for data to under-stand the current state or performance and to learn if interventions result in meaningful improvement.In this era of advancing informatics, data has become much more available in many forms as well as varying levels of reliability. This avail-ability can help in all stages of an improvement project yet must be used responsibly with an understanding of limitations, especially when comparing different systems or institutions [5].Carefully planned data collection is critical to each stage of improvement, and analysis can help assess the scope of a problem, learn where the greatest opportunities for improvement lie, plan interventions based on potential impact, and track ongoing successes or failures. This assess-ment is critical in helping to set priorities for new or ongoing improvement initiatives. Data can also help in determining the need for further sup-

157port and resources such as people, tools, and technology to help drive improvement.Finally, in order to be used to influence prac-tice, data must be perceived as valid. Healthcare workers, as improvement scientists, expect timely, credible data to drive improvement and avoid potential harmful changes stemmed by inaccurate data [6].Improvement teams need data, both qualita-tive and quantitative, to assess and judge their work. Gerald Langley describes five types of data useful in supporting improvement efforts. These include continuous measurements, counts of observations, documentation of individuals’ thoughts or feelings about an issue, ratings, and rankings [7]. Data collection must start with a plan, including what data will be collected, how the data will be collected and by whom, and when and where the data will be collected.The role of data in helping to tackle complex issues can be more clearly understood if one con-siders the five stages of quality improvement cycles, as delineated in Key Point Box 8.1. The first two phases focus on ensuring the right issue is identified and studied. The third phase centers around the implementation of strategies. Phases 4 and 5 focus on outcomes and sustainment of improvements [8]. Each of these phases relies on the evaluation of gathered information to help inform decisions. Defining the ProblemPhase 1 asks “What is the problem?” or “Is there a problem?”. This phase, while seemingly the most basic, asks for problem identification and clarification, along with its impact, and may rely on both qualitative and quantitative data.Vignette 8.2Although the numbers of patients eloping prior to evaluation by a provider have been tracked for years, little had been done with the data other than the acknowledgment that the rate of LWBS patients increased dramatically when the daily ED census increased, especially when it reached about 190 in a 24 hour period. In addition, peaks in LWBS rates were sometimes noted dur-ing times of high inpatient census and when the ED was boarding admitted patients, when the mental health patient volumes increased, or when many time- consuming ED procedures needed comple-tion, such as suturing or procedural sedation. The ED had increased staffing during traditionally “busy” hours but had not noted a significant or sustained decrease in the number of patients leaving without being evaluated by a provider.While data was available to suggest cer-tain contributing factors, much of the infor-mation regarding the reasons for the high rates of patients leaving prior to treatment were based on superficial evaluation of data and assumptions. It was largely assumed that many of the factors which appeared to be associated with high LWBS rates were outside of the control of the ED or the hospital, and this was simply a byproduct of working in a busy ED. The culture was simply to accept it and try to work harder and faster during busier times.Key Point Box 8.1Phases of quality improvement projectsPhase 1 – Defining the problemWhat is the problem?Is there a problem?Phase 2 – Diagnosing the problemWhat are the defects?What can we improve?Phase 3 – Implementing changeHow can we improve?What experiments should we try?Phase 4 – Measuring improvementHave we improved?Phase 5 – Sustaining improvementIs improvement sustained?

158In order to clearly define the problem, multi-ple types of data are often necessary. Historical data can help assess current performance and establish a case for the need for improvement, as well as helping to establish overall targets or aims for the project. It can also aid in prioritization of problems and projects, especially with complex, multifactorial problems. Carefully planned data collection may also add insight and understand-ing to the needs and opinions of key stakeholders. Several different tools can be used to help estab-lish baseline data. Run ChartThe run chart is a simple graph of data over time, typically looking at only one variable, that can help determine how a process is performing and the impact of changes. When variation is observed, run charts can help begin to identify whether it is random or nonrandom.In creating such charts, understanding the types of data being collected is key, specifically whether it is attribute or continuous data. Attribute data is counted (discrete) and easier to obtain. It can be broken down into count (incidences, non-conformities, or defects) or classification data (defectives or nonconforming units). Continuous data, on the other hand, is measured, such as the length of time from initial presentation to being seen by a provider. This type of data often pro-vides more information for planning or evaluat-ing a system. See Key Point Box 8.2 for more information regarding types of data. Chapter 9 discusses the differences between these variables further as well as the use of run and control charts.Common specific types of data used for improvement include count, classification, per-centage, and rate data. Counts (frequencies) are fairly straightforward – the total number of events occurring during a defined time frame, such as the number of patients presenting to the ED per day. Classification data are documented into one of two categories (e.g., defective or not defec-tive). Percentages give information about the rate of occurrence, which may be necessary to under-stand if change is really happening, especially in situations in which the denominator may not be consistent. For instance, when measuring LWBS numbers, raw numbers may not tell the true story regarding the progress of an initiative. A total of five patients leaving without evaluation on a day Vignette 8.3ED leadership understood that patients and families often have a choice in healthcare, yet believed that acutely ill or injured chil-dren are best cared for in a specialty pediat-ric hospital. Further, the hospital administration understood that each family eloping prior to evaluation could have neg-ative health impacts on children who were not seen and a negative financial impact on the hospital overall. They needed to under-stand when and why families were choos-ing to leave their hospital without being seen. It had long been believed that overall volume and wait times to see a provider were the driving factors in families choos-ing to leave, though percentages of LWBS patients were now increasing despite stable to slightly decreased overall ED census.Key Point Box 8.2AttributeContinuousCharac-teristicsCountableMeasurable/continuousTypesCount dataClassification dataExamplesNumber of patients affectedPercentage of patients affectedNumber of days between eventsRates of occurrencesNumber of patients with total length of stay over 3 hoursAverage length of stay

159when only 150 present for care (3.3% of total patients) may be evaluated differently than a day in which 5 patients out of a total 260 leave (1.9%). However, careful analysis of either set of data using various stratification strategies may reveal other trends and conclusions not evident from the initial perusal of the gross data.A rate is similar to a percentage in that it includes a numerator and denominator; however, with rates the two numbers are not alike. For instance, when studying patient falls for hospital-ized patients, one may want to look at the number of falls per 100 patient days as opposed simply to what percentage of patients fall while hospital-ized. In this scenario, the numerator is the num-ber of falls, but the denominator is the number of inpatient days for the month, creating a rate. Measures which include rates should include the word “per” in the title (e.g., falls per 1000 patient days) [9].Run charts are fairly quick and easy to put together and are somewhat intuitive to under-stand. A run chart is developed by first creating an x-axis (horizontal line) which represents a unit of time or a sequence then the y-axis (vertical line) to represent the range of data which are col-lected. A minimum of 10 data points is suggested for meaningful interpretation. Each point is then connected by a single line, and the median value for the data is determined, adding that line to the chart. The use of the median is preferred over the mean in run charts as it gives a good estimate of central tendency, without putting undue impor-tance on extreme values that will affect the mean. The median remains constant regardless of whether the data display a normal or skewed dis-tribution [10].Figure 8.1 depicts a run chart demonstrating the percentage of patients leaving the ED prior to being seen by a provider daily through the month of April, with a median of 2.6%.Now that the run chart is created, it can be used to begin to analyze process variation. Random variation is natural variation inherent to any pro-cess and is due to regular or ordinary causes. The presence of only random variation indicates sta-bility in the process. Nonrandom variation, how-ever, suggests irregular or unnatural causes and April LWBSDate1098765432101/1/20194/3/20194/5/20194/7/20194/9/20194/11/20194/13/20194/15/20194/17/20194/19/20194/21/20194/23/20194/25/20194/27/20194/29/2019PercentageMedianTarget% LWBSFig. 8.1 Run chart demonstrating left without being seen rates

160creates unpredictability in a process. While all processes will display some degree of random variation, nonrandom sources of variation should be identified and eliminated specifically prior to attempts to shift the median by changing an entire process. Nonrandom variation can also be desir-able, especially when it displays improvement in your processes or outcomes due to a planned intervention. Chaper 9 (Quality Methodology) discusses rules for determining random vs. non-random variation, focusing on shifts, trends, num-ber of runs, and any astronomical points. Assessing variation from a run chart begins with the identification of runs. A run is defined as one or more consecutive data points above or below the median. Points falling on the median are not counted. Once runs are identified, analysis can begin. Although a variety of rules and options have been created for the analysis of run charts, the Institute for Healthcare Improvement (IHI) has agreed on four simple rules which they feel are the most relevant to healthcare [9, 10]. Rule 1: ShiftsThe first rule in assessing a run chart is to identify a shift in the data. This rule helps deter one from becoming too excited or disheartened by a small number of data points moving the same direction. A shift occurs when six or more consecutive points fall on the same side of the median. In our example in Fig. 8.1, the run beginning April 10 contains seven consecutive points below the median, indicating a shift. This shift should lead one to investigate further to learn potential rea-sons for this nonrandom pattern.Further evaluation revealed that during those days, the hospital inpatient census was lower than usual with fewer scheduled surgeries than usual, in addiiton to ED volumes being lower. As a result, admitted patients were transferred to the floor more quickly and there were no ED rooms occupied by “boarded” inpatients. Additionally, staffing was very good that week due to very few “call-ins”. This combination of more available exam rooms and excellent staffing led to a decreased wait time, and fewer patients leaving prior to evaluation. Rule 2: TrendsThe second rule is applied to define trends in data. The IHI defines a trend as five or more data points constantly going the same direction, with-out regard to whether they cross the median [10]. Points that repeat a previous value are not included when determining trends. In our exam-ple, no trends are identified. Rule 3: Too Many or Too Few RunsThe third rule simply identifies the number of runs present in the chart vs. the number expected based on the number of useful observations. The number of useful observations is first calculated by subtracting the number of data points that fall on the median line from the total number of data points. This number is then identified on a stan-dard table that identifies the number of runs that should be observed based on the number of use-ful observations. The table most commonly used was developed by Swed and Eisenhart in 1943, an excerpt of which is included in Table 8.1 [11]. If too few or too many runs exist, compared to the expected number range, this signals nonrandom variation. In our example above, there are 30 use-ful data points (none fall directly on the median line) and 11 runs. According to the table, the lower number of expected runs is 11, and the upper number is 21, indicating that variation is random. Rule 4: Astronomical Data PointsRule 4 is an observational rule to quickly identify a data point that may warrant further investiga-tion. While every chart has high and low points, here we are looking for points that drastically vary from the rest of the data. On our run chart, discreet spikes are noted on April 8 and 25, significantly higher than other dates during the month, suggesting astronomical points. These data points should lead to further investigation, including double-checking the data to make sure the same rules were applied in collecting it for

161that occurrence, then taking a deeper look into potential causes of this variation.Run charts overall are great for understanding current performance of a process. They are useful in establishing baseline data and following as interventions are implemented. Annotating run charts to indicate points of intervention or PDSA cycles can help demonstrate the effects of the interventions. (See Chap. 9, Quality Methodology). Statistical Process Control Chart (Shewhart Chart)The statistical process control chart (a.k.a., Shewhart chart or control chart) also displays data in time order but includes the mean with upper and lower control limits, statistically defin-ing the predicted boundaries within which data are expected to fall. Thus, it highlights situations where data fall out of normal variation, or “out of control.” The control limits are set at three stan-dard deviations above and below the mean, indi-cating a 99.73% probability that a data point will fall within the control limits. While run charts can only indicate whether variation is random or nonrandom, control charts can be used to help distinguish between common cause and special cause variation. Common cause variation refers to expected or natural variation inherent to the process, whereas special cause refers to variation related to specific circumstances. Data points falling outside of the upper or lower control lim-its indicate special cause variation. There are four other special cause rules which can be applied and are discussed in Chap. 9 (Quality Methodology). In addition, control charts allow you to determine the capability of the process and help predict the future behavior of the process. The three types of control charts most commonly used include C-charts, which count the number of occurrences, P-charts which depict percent-ages of occurrences, and U-charts depicting rates of occurrences.Figure 8.2 represents a P-chart, illustrating the daily percentage of patients leaving prior to eval-uation by a provider in the month of April. Although variation exists, most data points remain within the control limits, indicating com-mon cause variation. On two dates however, (April 8 and 23), the LWBS rate exceeded the upper control limit, indicating special cause vari-ation. This should serve as a signal for further evaluation. When the team investigated these Table 8.1 Table to determine whether too few or too many runs exist on a run chartNumber of useful observationsLower number of expected runsUpper number of expected runs103911210122111341114412155121651317313186141961520616217162271723717248182581826919271019281020291020301121311122321123331223341224351224361325371325381426391426401527411527421628431628441729451730Used with permission of The Annals of Mathematical Statistics, from Tables for Testing Randomness of Grouping in a Sequence of Alternatives, Swed and Eisenhart [11], permission conveyed through Copyright Clearance Center, Inc.

162dates, they learned that the ED had significant issues with inpatient boarding both days, with inpatients occupying 40% of their rooms for over 12 hours on the 8th and 25% of the rooms on the 25th, a day which also included a high volume of patients requiring 1:1 observation in the ED. This demonstration of the effect of boarded patients on LWBS rates was shared with administration and inpatient units to help create a better under-standing of how issues in other areas of the hos-pital, such as lack of an overflow unit, can affect the LWBS rate. Qualitative MeasuresQualitative data collection at the beginning of a project can help gain deeper insight into issues, capturing feelings, opinions, and ideas from both staff and the customers (patients and families). Surveys, especially those with free text questions may lend insight into the process from varying points of view. Surveys carry the advantage of being quick, relatively easy to create and distrib-ute to large numbers of respondents. However, response rates can be extremely variable, and since the questions are set, there is limited to no ability for follow-up or clarifying questions.Interviews can be performed individually or in groups and can be particularly helpful to appreci-ate deeper meaning underlying individual or group perspectives or feelings. Interviews by nature are time-intensive, requiring skilled facili-tators and notetakers, as well as the skills neces-sary to reliably analyze large amounts of narrative. Qualitative data are particularly helpful in understanding the story or deeper context asso-ciated with a process. Qualitative data are gener-ally depicted by sharing of a story, for example, describing a scenario associated with a patient that left before being seen. Additionally, common themes elicited from qualitative data from inter-views or surveys can be used to help identify per-ceived barriers to implementing changes, to identify competing priorities among stakehold-ers, to uncover biases that may interfere with progress, etc. This may provide information cru-cial to the improvement planning process.01234% LWBS56789104/1/20194/2/20194/3/20194/4/20194/5/20194/6/20194/7/20194/8/20194/9/20194/10/20194/11/20194/12/20194/13/20194/14/20194/15/20194/16/20194/17/20194/18/20194/19/20194/20/20194/21/20194/22/20194/23/20194/24/20194/25/20194/26/20194/27/20194/28/20194/29/20194/30/2019DateApril LWBS% LWBSTargetAverage (CL)UCLLCLFig. 8.2 P-chart (control chart) representing left without being seen rates for April

163 Diagnosing the ProblemPhase 2 of a quality project is considered the diagnostic phase, asking the questions “Where are the defects in the process?” and “What can we improve?”. In this phase, the team takes a deeper dive into the problem, focusing on struc-tures and processes involved. The goal of this phase is to identify areas of greatest impact, pri-oritize opportunities for improvement, and iden-tify barriers that may impede the project’s success.The data tools already mentioned may again help support this phase, though a deeper under-standing is crucial here. The initial phase high-lighted that a problem exists or that the current state is not the desired state and may have led to some hypotheses as to causes, but in this phase the focus is on identifying the roots of the problem. Scatter PlotA scatter plot creates a geographic representation of the relationship between two variables. One variable is placed on the x-axis and the other on the y-axis. For a scatter plot to be useful, it should include at least 20–30 data points. If the two vari-ables are related, the points will follow a diagonal line or a curve. Scatter plots are very useful in demonstrating correlation, but one must keep in mind that correlation does not necessarily indicate causation. More closely clustered dots on the graph indicates a stronger relationship between the two variables than dots which are spread fur-ther apart. When the pattern of dots rises from the lower left to the upper right of the chart, a positive correlation is suggested, that is, when one vari-able increases, the second increases as well. On the other hand, an inverse relationship is sug-gested when the pattern of dots declines from the top left to the bottom right of the graph.Figure 8.3 demonstrates a positive relation-ship between the number of patients presenting to the emergency department per day and the LWBS rate. This graph shows that when the cen-sus remains under about 180 patients/day, LWBS rates remain consistently under 3%. However, when daily census surpasses 190, there is a sharp increase in the rate of patient eloping prior to evaluation, with rates rarely falling under 4%.1614121086420150160170180190Daily censusPercentage LWBSEffect of census on lwbs rates200210220230240Fig. 8.3 Scatter plot demonstrating the relationship between total census and LWBS rates

164Scatterplots also allow us to see unusual pat-terns, such as data affected by special cause vari-ation or clustering of data points that may suggest the need for further investigation. In the example above, the plot demonstrates correlation, but does not account for all causes, as there are several outliers on the plot. While the trend indicates a strong relationship, clearly there were certain days in which the census was quite high, but LWBS did not spike, as well as a few outliers in days with lower census. These easily visualized outliers may suggest special cause variation for particular points. Additionally, trends may be suggested by scatter plots as well, which may lead to the need for further investigation of causes for the patterns. Importantly, one must remember that regardless of the strength of the correlation, a scatter plot does not identify the reason for the correlation, only the relationship. Key Point Box 8.3 highlights key concepts for scatterplots.During this diagnostic phase, more emphasis should be placed on control charts as well. In the first phase, a baseline is established, and variation may have been identified, but in this phase, iden-tified variation should be studied. Identification of the type of variation the process is experienc-ing will help to determine the next steps in improving the methods. When a process only dis-plays common cause variation, one can reliably predict how a process will perform, within statis-tical limits.Common cause variation, though expected, should not necessarily be accepted as unimport-ant to address. It does not tell you that a process is good, only that it is stable and therefore pre-dictable. For example, in our case, the LWBS rates are mostly in control, but still undesirable compared to the goal of 2.5%. Changes to pro-cesses should be considered when the process demonstrates control but unacceptable results. The use of subject matter experts can be helpful to determine when stable systems (i.e., showing common cause variation) need improvement.Special cause variation should be studied carefully in this phase as well. This type of varia-tion results from unnatural or irregular causes that are not inherent to the process as a whole. They indicate instability in a process or chaos. When a process demonstrates instability and unpredictability, changes to the process as a whole may lead to wasted efforts or further insta-bility. Because they indicate a lack of control in a process, special causes should be investigated to determine the underlying reason for the variation, so steps can be taken to eliminate or minimize the cause of the variation when that variation has a negative effect. It is important to remember that positive variation can create special cause as well, and these instances should also be investi-gated to learn what went well for a particular event and if those factors could be replicated for future improvements. While some instances of special cause variation are truly unique, some-times patterns can be identified when a deeper look is taken. One helpful tool in assessing varia-tion is the abnormality tracker. Abnormality TrackerThe abnormality tracker (a.k.a., histogram) can be used to assess factors associated with variations or outliers. By tracking individual fac-tors that may contribute to certain outcomes, it may be possible to identify areas for improve-ment. This tool can be created very simply by creating a visual system in which a simple tally mark is placed each time there is variation, with an assessment of the perceived causes at the time. This will basically create a visual histogram showing frequencies of events – how many times a particular event occurs related to the identified problem. An abnormality tracker is simple to use and requires no specialized statistical program but provides a good visual depiction of the rea-sons for process failures. It can also help provide Key Point Box 8.3Scatter plotsNeed at least 20–30 points to be usefulShow correlation, not causationGreat for pattern recognitionOutlying points indicate variation

165a subgrouping strategy for your ensuing data analysis.For instance, in the case of ED LWBS, the overall problem was thought to be related to the volume of patients. While this was partially true, a deeper look revealed that certain additional fac-tors contributed significantly to the rate. The team tracked the number of patients boarding in the ED awaiting inpatient beds, the number of patients requiring mental health evaluations, the number of patients with extended time from admission order to inpatient bed placement, and the overall wait time to see a provider.Figure 8.4 demonstrates the abnormality tracker used. The staff added to the tracker any day in which the LWBS rate was over 2.5% (the stated goal), with a tally for each of the identified contributing factors. From this, they were able to determine that the likelihood of patients leaving prior to evaluation increased when wait times surpassed 2 hours from the time of arrival, which led the team to brainstorm ways to decrease the time to initial contact with a provider. In addition, the impact of patients requiring 1:1 monitoring is demonstrated as an important contributing factor. Interestingly, some potential causes of staff though would be major contributors to the prob-lem, such as multiple sedations in a day did not appear to affect the LWBS rate significantly. Pareto ChartThe Pareto chart has been widely used in quality improvement work. It was developed by Vilfredo Pareto, an Italian economist and philosopher, who noted that 80% of the land in Italy in the 1800s was owned by 20% of the population. This concept, referred to as the Pareto principle, states that for many events or problems, a small number of factors will account for the majority of the rea-sons that the problem occurred [9]. Key Point Box 8.4 highlights key features of Pareto charts.The Pareto chart is a type of histogram that includes both bars and a cumulative line. The bars, which represent frequencies, are arranged in descending order, and the cumulative total is represented by the line. The right-sided vertical axis measures the percentage of the total contri-butions of each factor. This is particularly useful when analyzing complex problems that may have many underlying causes to help identify the most common sources of defects, which can help with prioritization of tests of change. Figure 8.5 illus-trates the ice cream flavor preference for ED staff members, which was used to help determine which flavors to buy for the celebration when the team reached the goal of a 0.5% reduction in overall LWBS rates. Since the budget only allowed the purchase of three flavors, a quick sur-vey was taken to ensure that the majority of pref-Fig. 8.4 Abnormality tracker depicting factors contribut-ing to LWBS ratesKey Point Box 8.4Pareto chartsBar and line graphBars arranged in descending orderUseful to analyze frequencies of eventsHelpful to determine predominant factors

166erences would be covered. From this, we learned that almost 90% of the staff preferred either vanilla, chocolate, or strawberry ice cream. Qualitative DataIn the diagnostic phase of the project, qualitative data may be equally important as quantitative in helping to identify potential interventions and possible barriers. As in the first phase, interviews can help gain valuable insight and help discover potential barriers to improvement strategies. Simple discussions with staff may uncover cul-tural barriers to change and may surface innova-tive ideas. Interviews with families (customers) may lead to a deeper understanding of the prob-lem and factors which may make them more likely to behave in one way or the other.Another qualitative method to evaluate a pro-cess is through observation. Observations repre-sent a powerful method of gathering information to truly understand how a system is functioning from another perspective. As discussed in other chapters, direct observations done on the “shop floor” may be critical to understanding the prob-lem better. Gemba walks, as discussed in Chap. 5, present a great opportunity for observation. By using direct observation, one can see behaviors and the impact of human factors in a real-world setting. They may also include observing each step in a process, including time stamping of each individual step. Observations can be time- intensive and can be subject to the “Hawthorne” effect, in which individuals alter behaviors due to their awareness of being observed. Therefore, there is value in performing multiple observa-tions using several trained observers.VanillaChocolateStrawberryPecanIce cream flavorIce cream preference05Number of staff1015202530354045500%10%20%30%40%50%60%70%80%90%100%CoffeeRainbow sherbetPistachioFig. 8.5 Pareto chartVignette 8.4The improvement team had noted that the rates of patients choosing to leave prior to evaluation tended to increase when wait times were long. However, they noted on some days with extended wait times very few families left. The team spent several hours in the waiting room observing behav-iors, noting that certain behaviors of the nurses at the front desk impacted the likeli-hood of a family choosing to stay or go. For

167Although the collection of this type of qualita-tive data may be time-consuming, it provides insight that numbers alone simply cannot, espe-cially when formulating plans for interventions. Implementing ChangeThe third phase of improvement projects is often referred to as the intervention phase. In this phase, potential interventions are identified, per-formance measures are defined, improvement experiments are implemented, and progress is continually monitored. Reliable data collected in the diagnostic phase will help guide the team in creating interventions and implementation strate-gies. This phase is often referred to as the PDSA or plan-do-study-act cycle and is discussed in detail in Chap. 9 (Quality Methodology). This phase, often referred to as the experimentation phase, includes small-scale trials with continuous assessment of the impact to guide further refine-ments prior to disseminating the strategies more widely. Tests of change are initiated, studied, adapted, and studied again. Data management and monitoring in this phase must be robust, timely, and transparent in order to plan next steps and to maintain engagement. Staff may have a difficult time adapting to multiple changes if they cannot appreciate the effects of the changes. Further, the knowledge that each change will be studied for positive effects as well as unintended consequences, and a more permanent change will not be initiated until the process is fully vetted helps improve staff acceptance of the process.The importance of knowledge gained from each experiment, regardless of success is evident throughout history, perhaps most famously by Thomas Edison. In his biography, “Edison: His Life and Inventions” an anecdote is shared by his longtime associate Walter Mallory, highlighting one of Edison’s most famous quotes. When asked “Isn’t it a shame that with the tremendous amount of work you have done you haven’t been able to get any results?”, Edison reportedly smiled reply-ing “Results! Why, man, I have gotten a lot of results! I know several thousand things that won’t work” [12].Several of the tools already described can be helpful in this phase, particularly control charts and abnormality trackers. As each new experiment is initiated, it is critical to study the effects, both positive and negative, of that implementation. It is important to consider any “balancing measures” or other areas that may be affected by the implemen-tation of the intervention. Balancing measures help the team ensure that possible unintended con-sequences of a new process are recognized and addressed. For example, a plan to decrease ED inpatient boarding of patients by emphasizing rewarding physicians for early discharge of inpa-tients could lead to some patients being discharged before they were truly ready, leading to return vis-its. This information is then used to adapt the experiment and plan the next steps.Additional details about the PDSA cycle, especially its merits and potential pitfalls when misused, can be found in multiple other texts, including The Improvement Guide [7] and Quality Health Care [9].instance, one nurse greeted every family on arrival and made a point to let them know that the wait was longer than usual that day but assured the families that their child was important and would be seen as soon as possible. She further asked them to let her know if anything changed or they would like for her to reassess their child. She also frequently scanned the waiting room for patients who had been there longer than an hour and touched base with the families. In contrast, the team observed another nurse who seemed overwhelmed with each patient signing in. She also let them know that there was a long wait, so they should take a seat and the team would get to them when they could. After the initial assess-ment, she did not offer comfort measures or reassessment while waiting. The team quickly learned that families were much less likely to leave if they felt that they were attended to.

168 Visual Management BoardsVisual management boards may be particularly helpful in this stage of rapid process change. A well-designed visual management board pro-vides information about performance at a glance, typically including unit specific measures chosen to align with the overall strategy of the system. This is the ideal platform to display metrics reflecting current improvement efforts and proj-ects and engage an entire team in the improve-ment initiative. A dynamic visual tracking board can also help to identify problems or defects ear-lier than traditional methods. The idea of creating such a board may seem daunting to some, but it does not have to be complicated. By nature, the board needs to be easily changed and updated to reflect the current state.A few tips can help to ensure the success of a management board: 1. Keep it simple. It should be easily understood by all the staff in the department and ideally by visitors from other areas of the organiza-tion. When the Chief Nursing Officer or CEO stops by, he or she should be able to under-stand what is being measured and progress being made. Similarly, if displayed in a public access area, visitors or families should be able to recognize progress on the board as well. 2. Choose process measures that are achievable and frequent enough occurrences to provide meaning with regularity. Displaying varia-tions that occur only once or twice a year may not provide meaning on a day-to-day basis for many staff members. For instance, an ED may want to track the frequency of mislabeled specimens occurring weekly as each mislabel may lead to recollection of specimens, creat-ing discomfort for patients, prolonging the overall length of stay for the patient, and decreasing capacity to see the next waiting patient. However, regular tracking of patients who have a splint applied to the wrong extrem-ity, an occurrence which certainly prolongs the length of stay due to the need for a repeat procedure, but occurs less than once a year is unlikely to provide meaning toward the over-all goals of the department. 3. It is called a visual management board – make sure it lives up to its title! The board should be populated with tools such as graphs, charts, and other visual tools such as red/green indica-tors so staff can understand the current state at a glance. Consider presenting data over time. 4. Choose its location wisely. As in real estate, location matters. The visual board should live in an area accessible to all staff who may be affected or influenced by it. Some even Vignette 8.5Since the team did not feel that they could significantly affect the variable rates of patient arrival or overall daily census, and reducing overall time in department was important but would take longer to achieve, relying on many areas of the hospital, they decided to start with interventions that would not require a significant increase in resources. The first intervention included a script for the greeting nurse and regular rounding on waiting patients to reassess and remind them that we know they are waiting and want to address their urgent and emergent healthcare needs. Many of the staff began offering simple comfort measures to families when they checked on them, such as water or blankets. Within the first week, there was a small decrease in the LWBS rate, despite no decrease in overall census or door to doctor time. With this, the staff also noticed that not only were fami-lies less likely to leave, they were less upset with the wait times. Soon, the staff began to provide suggestions on other things that could be done for patients while they were waiting, such as educational videos about seasonal conditions, initiation of oral chal-lenges for children with suspected gastro-enteritis, standing orders for certain situations, etc. Within a month, LWBS rates decreased by 0.7%, and a culture shift was palpable.

169 advocate for more public placement, visible to visitors and patients as well, though not all institutions are comfortable with this level of transparency.Choose the type of display. Visual manage-ment boards can be as basic or as technical as you like. They may be as simple as paper or poster board, as adaptable as whiteboard, or as technical as electronic displays. Each has its merits. Paper versions are simple to create and require less technical expertise and very little money. However, changes can get very messy or require reworking of the board frequently. Size con-straints may come into play as well. Whiteboards are popular in healthcare for a variety of reasons. They are relatively inexpensive, they appear a little more “permanent” than paperboards, and they are easily changed and available in a large variety of sizes. Electronic boards carry the advantage of the ability to capture and record changes automatically, as well as being able to update certain data points close to real time. The ease of use and adaptability depends largely on the skills of those responsible for the creation and changes. Electronic boards also come with a heft-ier price tag and the risk of inaccessibility when technology fails. Thus, the most important fac-tors in choosing which type of board to use are the needs, preferences, and budgets of those who will be using them daily (Figs. 8.6 and 8.7).Fig. 8.6 Visual management board – whiteboard versionVignette 8.6The ED visual management board had tra-ditionally included historical data (the prior day’s numbers) and longitudinal metrics. With the LWBS initiative, a switch was made to more real-time results to inform

170 Measuring ImprovementThe fourth phase of the quality improvement cycle focuses on assessing the impact of inter-ventions, asking the question, “Are we improv-ing?”. In this phase, the impact of experiments is October 27, 2019- 1600CensusLength of stay (minutes)Door to provider (minutes)Patients in waiting room1:1 ObservationsPts. Waiting for inpatient Bed> 1 hourCurrent state411423271101120107224 Hour avg.StaffingRNMD/DOANPRTTechRegistrationChild lifePharmacySocial workUnlabeled/mislabeled labspecimens (by week)Monthly LWBS percentageProcess metrics010.51.522.533.5JanuaryFebruaryMarchAprilMayJuneJulyAugustSeptemberOctober0Aug 1-7Aug 8-14Aug 15-21Aug 22-29Aug 30-Sept 5Sept 6-13Sept 14-20Sept 21-27Sept 28- Oct…Oct 5-11Oct 12-182468Fig. 8.7 Visual management board – electronic versionstaff. Current census numbers were dis-played on the board, in addition to numbers of patients waiting over 1 hour to be seen and those with a total length of stay over 3 hours. In addition, the board contained quick visual clues representing potential barriers that may affect flow for the day, such as faulty equipment, short staffing, high inpatient census, etc. With this, staff could see the current state of the depart-ment at a glance, and certain triggers were put into place that required action by the team working at the time. For instance, in any shift during which there were more than four patients who had been in an exam room for over 60 minutes without being seen, the team was expected to huddle and create a plan to brainstorm ways to help the families get what they needed. This may include shifting of resources “quick round-ing” by providers to get evaluations initi-ated, “check-ins” from nursing staff, calls to the inpatient units asking for help mov-ing admissions out of the ED to the floors, etc. The point is that each day may present different obstacles leading to the extended wait, but with the availability of carefully chosen real-time data, the teams had the information they needed to assess the cur-rent obstacles and create plans for improve-ment. Staff felt more empowered to help make changes.

171measured for both positive and negative effects. Processes are closely monitored for stability, with attention to data that may suggest a trend, either favorable or unfavorable, that may indicate the need for further assessment or adaptation of the process. Special causes should, of course, be evaluated as they present as well. Sustaining ImprovementFinally, once improvement has been achieved and the process has stabilized, it is essential to monitor sustainment. Without ongoing monitor-ing, even projects that have demonstrated dra-matic process and outcome improvements may experience slippage. As improvement projects wrap up, resources may be diverted to other pri-orities, creating the possibility of losing some of the benefits of the new process. Thus, when a project enters this phase, it is critical to consider the overall sustainability of the process and to choose metrics to continue to monitor the sys-tem’s performance. The amount of rigor and detail necessary for this monitoring phase will vary depending on the level of complexity and the anticipated risk of the process reverting to its original state. Often, after a period of demon-strated sustainment, control charts can be used to follow progress, with particular attention given to variations that appear. SummaryUnderstanding the need for reliable data to drive change is an integral part of improvement. The challenge lies in determining which data to col-lect and which of the many tools may work best for the particular project at hand. The tools cho-sen will depend on the complexity of the prob-lem, the phase of the project, and the resources available, but do not have to be elaborate or require a statistician’s expertise. When consider-ing your data, plan carefully, experiment will-ingly, learn from successes as well as missteps, and consider the impact on the team and work-flow. Improvement will follow. Chapter Review Questions 1. True or False – Scatter plots are useful in determining the cause of variation.Answer: False. Scatter plots are useful for identifying patterns and correlation, but do not provide information regarding causation. 2. Which phases of improvement benefit from data analysis? A. Planning phase B. Implementation phase C. Sustainment Phase D. All of the aboveAnswer: D. All phases of improvement, from defining the problem through sus-taining the improvement, can benefit from carefully planned and evaluated data collection. 3. True or False – Run charts can be used to fol-low basic trends but cannot distinguish com-mon cause variation from special cause variation.Answer: True. Run charts may provide an indication of nonrandom variation, but cannot distinguish a special cause. Control charts are useful for determining special cause variation. 4. True or False – The Pareto principle indicates that the majority of problems or defects in a process are caused by a relatively small num-ber of factors.Answer: True. The Pareto principle sug-gests that 80% of the problem can be attrib-uted to 20% of the causes. 5. True or False – Qualitative data is too subjec-tive to be useful in most quality improvement initiatives.Answer: False. Qualitative data plays an important role in improvement projects, par-ticularly in understanding what people believe and how they feel about a problem.References 1. Doyle AC. A scandal in Bohemia. In: The adventures of Sherlock Holmes. London: G. Newnes; 1892.

172 2. Donabedian A. Evaluating the quality of medical care. Milbank Q. 2005;83(4):691–729. https://doi.org/10.1111/j.1468-0009.2005.00397.x. 3. Berwick D, Fox DM. “Evaluating the quality of medical care”: Donabedian’s classic article 50 years later. Milbank Q. 2016;94(2):237–41. https://doi.org/10.1111/1468-0009.12189. 4. Data. Merriam-Webster.com. https://www.merriam-webster.com (18 October 2019). 5. Powell AE, Davies HTO, Thomson RG. Using rou-tine comparative data to assess the quality of health care: understanding and avoiding common pitfalls. Qual Saf Health Care. 2003;12(2):122–8. https://doi.org/10.1136/qhc.12.2.122. 6. Bradley EH, Holmboe ES, Mattera JA, Roumanis SA, Radford MJ, Krumholz HM. Data feedback efforts in quality improvement: lessons learned from US hos-pitals. Qual Saf Health Care. 2004;13:26–31. https://doi.org/10.1136/qshc.2002.4408. 7. Langley GJ. The improvement guide: a practical approach to enhancing organizational performance. 2nd ed. San Francisco: Jossey-Bass; 2009. 8. The Victorian Quality Council 2008. A guide to using data for health care quality improvement. Department of Human Services www.healthvicgovau/quality-council. Accessed 21 Apr 2019. 9. Lloyd RC. Quality health care: a guide to develop-ing and using indicators. 2nd ed. Burlington: Jones & Bartlett Learning; 2019. 10. Institute for Healthcare Improvement. Boston, MA, 2019. http://www.ihi.org/. Accessed 12 Apr 2019. 11. Swed FS, Eisenhart C. Tables for testing randomness of grouping in a sequence of alternatives. Ann Math Stat. 1943.;XIV:66–87. 12. Dyer FL, Edison MTC. His life and inventions, vol. 1. London: Harper & Brothers; 1910.

173© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 R. K. Shah, S. A. Godambe (eds.), Patient Safety and Quality Improvement in Healthcare, https://doi.org/10.1007/978-3-030-55829-1_9Quality MethodologyMichael T. Bigham, Michael W. Bird, and Jodi L. SimonAbbreviationsKDD Key driver diagramPDSA Plan-Do-Study-ActM. T. Bigham · M. W. Bird (*) · J. L. Simon Akron Children’s Hospital, Department of Quality Services, Akron, OH, USAe-mail: [email protected]; [email protected]; [email protected] 9Vignette 9.1According to the 2017 CDC National Center for Health Statistics, nearly 6.2 million chil-dren under the age of 18 have asthma [1]. Childhood asthma leads to increased emer-gency department visits, hospitalizations to acute care and intensive care settings, missed days of school for children, and missed days of work for parents resulting in a significant financial and social burden for patients and families. Evidence demonstrates that early identification and management of asthma, avoidance of asthma triggers, and strict com-pliance with daily medication regimes create the best possible outcomes for children with this chronic condition. A tertiary, free-stand-ing children’s hospital has seen a fairly flat 12-month rolling average in emergency department visits as well as hospital admis-sions despite a multitude of disconnected teams working to solve the problem within • Understand the steps for successfully implementing change• Learn strategies to spread and sustain improvementsChapter Objectives• Utilize the Roadmap for Quality allow-ing for enhanced communication, col-laboration, and coordination among team members and among separate teams that may be working toward an overall goal• Leverage quality tools to identify a goal and create a high-level plan and indi-vidual team plans which will support the overall goal of the improvement project• Execute effective PDSA cycles with predictions, measurement, and deci-sions to determine the next steps• Understand nonrandom variation, com-mon cause variation, and special cause variation

174 Identifying Improvement OpportunitiesIdentifying improvement needs in any healthcare organization can happen through a variety of means; examples include Community Health Needs Assessment (Affordable Care Act requirement), failures or risks (regulatory reviews or audits, seri-ous and near miss events of harm, safety event reports, or external performance benchmarks), patient feedback, high-volume care, and organiza-tional priorities (i.e., strategic plan) [2]. Understanding the Process for ImprovementOnce improvement opportunities are identified, quality improvement methodology is imperative to achieving successful improvement. People with formal and informal roles in quality improvement will be more effective in leading change through the use of quality tools. The Model for Improvement from the Institute for Healthcare Improvement (IHI) allows teams to identify the scope of work and metrics and con-tinually address the work within the framework of Plan-Do-Study-Act (PDSA) cycles [3]. Below are the key Model for Improvement concepts (see Fig. 4.2, Chap. 4): 1. What are we trying to accomplish? (Aim) 2. How will we know that a change is an improvement? (Measures) 3. What changes can we make that will result in improvement? (Interventions) LeadershipSystem-level quality improvement requires teams to interface to solve strategic problems but requires strong leadership. Leaders must identify a strategy to improve that is likely different than small-scale and local quality improvement efforts. In the case vignette described, asthma was deemed a “Transformation Project” (descrip-tors for high-visibility, crosscutting, major initia-tives) and endorsed by the executive team. A transformation project carries several benefits: (1) focus on the “system approach”, (2) commen-surate resource allocation, (3) executive sponsor-ship, (4) engaged stakeholder and steering groups, (5) environment for local team leaders to collaborate, and (6) regular executive and gover-nance review of metrics.To embark on the asthma transformation jour-ney, a multidisciplinary team assembled to ana-lyze baseline data and created a plan to improve the outcome. Executive leaders deployed quality, analytics, project management, informatics, and operational resources for the team in order to drive system-level improvement through enhanced coordination, collaboration, and com-munication, all missing elements in the prior overall state. Local teams were identified as inpatient, emergency department, primary care, home healthcare, school health, pulmonary, allergy, a newly formed asthma care manage-ment team, and a pediatric intensive care team-a collection of nine teams all working toward the same goal. The newly coordinated asthma struc-ture was designed to extend horizontally across their given areas with their current resources. To date, no significant improvements have been demonstrated. What needs to change to create better outcomes for patients with asthma seeking care in this hospital?Due to the disconnection of asthma teams, quality leaders observed unsuccess-ful tests of change being inappropriately duplicated, disparate improvement goals, inconsistent application of evidence, and limited quality improvement methodology being used. A newly developed asthma improvement team structure will seek to coordinate, allow for collaboration, and enhance communication of these nine teams. Quality team members are newly assigned to the project to redefine the struc-ture and methodology necessary to drive the success of the teams.

175teams and vertically, between frontline and lead-ership. A new culture was created where indi-viduals recognized their role in reducing the burden of asthma.With local teams aligned around asthma out-come goals, a clear aim statement was developed. The local teams were asked to standardize the data definitions on the measures for “asthma admission” and “emergency department visit.” This early precision around inclusion and exclu-sion criteria allowed for consistent tracking of data. This organization had an existing asthma registry containing years’ worth of patient-level data, allowing for prospective and retrospective data analysis.The teams started with the Roadmap for Quality (Fig. 9.1) which outlines each step to guide the teams’ progression through the improvement journey. Each stage has associated tools to complete tasks and provide learning as the improvement process continues. Aim and MeasuresStep one in the Model for Improvement requires clarity on “What are we trying to accomplish?” The Roadmap for Quality thus begins with the identification of an aim statement and associated measures. When teams came together and success-fully identified an aim inclusive of detailed mea-surement, the improvement journey commenced.Global aims development precedes specific aims and identifies the direction and intent of the work. A global aim is broad with no measures or timelines included. In this vignette, the Global Aim was:We aim to substantially reduce the burden of asthma for our patients, their families, and our community.Global aims are not specific enough to provide focused improvement targets, so a SMART (Specific, Measurable, Attainable, Relevant, and Time-bound) aim is established to provide focus. The asthma team arrived at the following SMART aim for inpatient admissions and emergency department visits:We will reduce Inpatient Hospitalization rate from 2.7% to < 2.0% (approx. 26% reduction), and ED visit rate from 5.8% to <5.0% (approx. 14% reduc-tion), for all Asthma registry patients by December 31, 2017.In this vignette and in many successful improvement projects, the use of data analysts to enhance the development of data definitions is beneficial. Clearly defined metrics ensure teams are able to maintain data to follow over time, lessening the risk of mid-project modifica-tions. If done well, data element definitions help the team members understand the common goal and answer the question of “What Are We Trying to Accomplish?” A tool to ensure consis-tent and repeatable data definitions is helpful (Fig. 9.2).Roadmap for QualityDevelop Aim andMeasuresDevelop LearningStructure with KeyDriversUnderstandProcess forImprovementMake DecisionsBased onLearningMore PDSACyclesSelect Change forScale-upPlanning forImplementation/SustainabilityPlan for SpreadIdentify TestableIdeasDesign PDSACyclesExecute PDSACylcesFig. 9.1 Roadmap for quality

176 Develop Learning Structure with Key Driver DiagramsKey driver diagrams (KDDs) connect the aim/outcome, with key drivers and interventions (testable ideas) to create a “Learning Structure,” and address the three questions that are part of the Model for Improvement (the Aim, Measures, and Interventions) (Fig. 9.3). Key drivers are big picture items (“what needs to happen”) that allow the teams to reach their aim. To identify drivers, the teams use evidence, data, inter-views, observations, and discussions. Drivers such as technology, engagement, and education are commonly identified on a variety of clinical improvement projects. Other drivers in this case vignette included asthma care coordina-tion, community engagement, and medication management.A key driver diagram is a fluid document and should be reviewed continuously and updated as needed. As teams progress through the improve-ment work, they may discover additional drivers that were not identified initially. Good version control is critical to ensure all improvement teams have the most recent version. The KDD may include color coding to identify work com-pleted, work in progress, or work on hold. Drivers can also be labeled with team identifiers to high-light the areas where teams have the opportunity to collaborate.In the nine-team example from the vignette, it was crucial to have each team develop local team-level KDD to identify contributions toward the system-level aim. As a result, each of the nine teams completed their own team-level KDD (inpatient team-level KDD, Fig. 9.4). A team- Fig. 9.2 Data element definitionVignette 9.2Nine separate asthma teams worked together developed a system-level key driver diagram (aka Learning Theory). Though committed individuals were doing good work in isolation, the need was clear to develop a unified plan with ideas for common drivers and interventions (testable ideas) that could serve in a crosscutting manner allowing more than one team to benefit from PDSA cycles for overlapping interventions. Some teams had garnered past success with identified interventions which were included in the system-level key driver diagram (Fig. 9.3).

177level KDD helps each team focus on work within their control. Identify Testable IdeasWith a maturing KDD, the intervention column is truly the opportunity to identify “testable” ideas. Said differently, interventions are the “how to” for each key driver. Sources of intervention might be better practice learned from within or outside of an organization, an evidence-based interven-tion, learnings from PDSA cycles, or even a best guess theory. A tenet of quality improvement is recognizing that many valuable “testable” ideas are generated by frontline providers and caregiv-ers. Frontline staff live in the current process, thus are often the people who encounter and experience the problems and have spent time thinking of possible solutions. Their input must be solicited. In the case vignette regarding inpa-tient asthma care, frontline staff revealed that they didn’t have a good way of remembering and tracking everything they were supposed to accomplish for an asthma patient prior to his/her discharge. Data analysis informed by the front-line observation of the process demonstrated that one important discharge element, the Asthma Treatment Plan, was only updated <10% of the time upon discharge.Typically, multiple interventions are consid-ered and depicted on a KDD. The interventions are commonly prioritized by the ease of testing, the expected impact on change, strategic align-ment with other improvement efforts, or any combination of these reasons. Once interventions are considered, they are tested through Plan-Do- Study-Act (PDSA) cycles. Interventions may be connected to one or several drivers. The arrows in Fig. 9.4 indicate where interventions connect to Asthma Care CoordinationGuideline-Based Care(Standardization)Identification of Asthma Patients(Risk Stratification)Medication Management& CompliancePatient & Family EngagementTechnologyEducation (Consistent acrossContinuum)Informatics(Metrics, Asthma Registry)BPAs*, Documentation, Define*/Identify*/Pareto HighRisk patients, Registry, AnalyticsStandardized curriculum for IP & OP; mechanism tomonitor; home env’t School Health, trigger avoidanceEpic, MyChart, Reminders, Tele-Health, Smartphones/Apps,e.g.pulmonary effort, Interactive Patient TV, Social MediaEvaluate Home Environment, School Health, 24/7 HotLines; use of Spirometry; Literacy; Behavioral Health issuesEZB, Registry, School, Home Care, High Risk*,Co-MorbidityEvidence-Based; Asthma Pathway, Asthma Treatment Plan,Asthma Control Test, EZB, Flu Vaccine, Use of spirometryE.g. Asthma care Management Team, Office/StaffEngagement; EZB, routine SW assessments for all highrisk patientsOrdering, Filling, Usage; Increase correct/decrease incorrectmedications, Medication demonstration devices forpractices/unitsCommunity EngagementAccess to Care at Appropriate LevelsPopulation Health, ACT Now,ED, Pulmonary, IP, Alternatives, Missed appointmentsResources, Technology, Pt & Family Engagement, Hot-Lines, Support Groups, PhoneINTERVENTIONSHow (specific) we accomplish the DriversKEY DRIVERSWhat (big picture) needs to be done to accomplish the AimSMART AIM9/16/2016 V4.2(PC, SH, AI, HC)(IP, ED/UC)(Pulm, PC, AI)(IP, SH, HC)(IP, SH, HC)(IP, SH, HC)(IP, PC)(SH, HC)(PC)Date:Project Name:Physican Co-ChampionsClinical Transformation Priority: Asthma2015–2017 Asthma Key Driver Diagram (KDD) – System LevelIncrease % of Practicesachieving >20% of OptimalCare (ACT & ATP by 12/31/17)& Flu Vaccine by 06/30/17from <10% to 80%Increase eligible Care Sourcemember’s dispensed asthmacontroller medication, closingthe gap between our 2015performance (42%) and thenational NCQA 90%ile (42.8%),by 20% (42.2%), by 12/31/2016(HEDIS).HEDIS Proxy measure, TBDIn the next 3 years, we aim tosubstantially reduce the burdenof asthma for our patients, theirfamilies, and our community.GLOBAL AIMSecondary Measures:Reduce Hospitalizationrate from 2.70%* to <2%(approx. 26% reduction),and ED visit rate from5.84%* to <5% (approx.14% reduction), byDecember 31, 2017.* Dec 2015, 12-month rollingaverage.Key:Green: In Progress* is Complete(Blue): Teams TestingRed: Protential BarrierFig. 9.3 Key driver program

178drivers. Teams should test interventions regularly using PDSA cycles. Design PDSA CyclesThe use of a standard PDSA cycle template (Fig. 9.5) is highly beneficial as it serves as the historical documentation of the many PDSA cycles the teams will complete over the course of an improvement project [3]. Projects may take years to accomplish during which time the composition of the improvement team may change. The PDSA documents memorialize all tests of change. Often, the PDSA templates inform the sequence of PDSAs, allowing one PDSA to inform the next PDSA.Those involved in the tests of change should participate in the completion of the PDSA form. The team should agree on a date and a location and identify the people involved in the test. A poorly designed PDSA cycle increases the risk of drawing incorrect conclusions from the results leading to misinformed decisions regarding adoption, adaption, or abandonment of a particu-lar intervention. Measurement of a PDSA impact can be enhanced by an assigned observer who is not directly involved in the test. The team should determine what information would be needed to answer the question the PDSA is designed to answer. Teams should also predict the impact of an intervention, particularly because those inter-ventions with the highest likelihood of achieving improvement may be prioritized for early testing. Much can be learned from a simple test which does not require large numbers of patients, many days of testing, or multiple team members. The majority of the PDSA cycle should be spent in planning.A well-planned PDSA cycle yields informa-tion that informs the team about the subsequent PDSA cycles [4]. PDSAs should start small. Medication Management andCompliancePatient and FamilyEngagement• Revise bedside communication form and add discharge checklist• Revise provider checklist by creating a SmartPhrase for handoff within H&P Template• Develop reference document for providers• Education to Providers• Filling controller prescriptions prior to discharge• Involving the ACM team in Family Centered rounds• Flu Vaccine will be stocked on the floors this season.• Protocal for nurses for all patients and BPA for Physicians• Increased marketing to staff and move unified message to familiesATP Revisions in ProgressRT posting eduction documentation percentages onCommunication BoardRRT/RN education and feedback about educationdocumentation focusing on family teach back ofdelivery device, roles of medication use and ATP• Share run chart data• Asthma IT subcommittee: redefining how reports capture an “asthma” admission and ED visit by using hospital problem listProvider education about asthma pathway and bundleelementsIdentify patients/families willing to provide feedbackabout inpatient process/strugglesAsthma patients now only being placed on 2 Residentteams. Residents are being educated about caring forasthma patients at beginning of block.Asthma Care Mangement (ACM) Team to addressbarriers:Flu Vaccine Campaign:Staff buy-in of importance ofbundleEpic Technology/informaticsIn the next 2 years we aim toreduce the burden ofAsthma for our Patients,Families and Community*Inpatient Optimal Care BundleDefinitation1. Asthma Treatment Plans (ATP)2. Asthma Education3. Influenza vaccine, seasonallyIncreasing compliance with theinpatient optimal care bundle*to 75% by 12/31/17in all inpatient unitsPhysicians/APPs/RRTs/RNs withknowledge of asthma managementguidelines and bundle elementsAsthma Care Coordination throughstandardized, reliable processes withclearly defined responsibilities forcompleting bundleSecondary goals:School ATPs, home care referrals,follow up visits, prescriptionsGray shaded box = completed interventionGreen shaded box = what we’re working on right nowKeyGlobal AIMAimKey DriversTeam: Inpatient AsthmaLead:Date: I08/302017Version 9Design Changes/InterventionsInpatient Asthma Key Driver DiagramFig. 9.4 Inpatient key driver program

179For example, one PDSA cycle for the inpatient asthma team involved adding a bedside paper checklist reminder to review the patient’s Asthma Treatment Plan, confirm asthma edu-cation completion, and confirm influenza vac-cination status. This checklist was the first test of a reminder system that may ultimately become part of the electronic medical record decision support functionality. The data assess-ing the PDSA “value” would include how often the checklist was completed and how many patients had all three of these bundle elements completed. Execute PDSA CycleThis is the “Do” part of the PDSA. Any person involved in the testing, especially frontline staff, should be made aware of the test and clearly understand that the intervention is not a perma-nent change. The asthma checklist was placed at the bedside of five asthma patients for a 24-hour period. The team was delighted to find that all five patients with asthma had the bundle com-pleted. The team was surprised that parents asked about the checklist and why the bundle items were important for their child. They had not PLANBriefly describe the test:How will you know that the change is animprovement?What driver does the change impact?Test the changes:Was the cycle carried out as planned: Y / NRecord data and observations:What did you observe that was not part ofyour plan?Did the results match your predications? Y / NCompare the result of your test to yourprevious performance:What did you learn?Decide to Adopt, Adapt, or AbandonAdapt: Improve the change and continuetesting the plan. Plan/changes for next test.Adopt: Select changes to implement on alarger scale. Develop an implementationplan and plan for sustainability.Abandon: Discard this change and try adifferent one.What do you predict will happen?DOSTUDYACTPlan for collection of data:Task Plan:List the tasksnecessary tocomplete this test(what)?Personresponsible(who)?When?Where? nameOverall team/project aim:What 90 day goal does the change impact?What is the objective of the test?Date of test:PDSA WORKSHEETTest completion date:Fig. 9.5 PDSA template form

180anticipated this reaction from parents, but inspired a parental checklist to inform parents about what should be completed before their child is discharged. This is an example of how a well-planned PDSA could yield data on the desired impact but also facilitate additional learn-ing. The parent and staff partnership creates a shared accountability for bundle completion. This unexpected event during the PDSA cycle would prove to become a future intervention to be added to the KDD. Make Decisions Based on LearningStudying the data (“S” of PDSA) qualitatively and quantitatively will assist the team in deciding if (1) the intervention worked as predicted, (2) the intervention enhanced performance, and (3) there were additional learnings. During early PDSA cycles, teams may find the data recorded does not definitively determine the success or failure of the intervention. This revelation affords the team the opportunity to identify additional data needs. Again, small tests of change provide valuable information even if the test resulted in a failure. Teams can gain valuable insight when failures occur and with little financial or human capital expended due to the small-scale testing. Teams can execute many PDSA cycles and learn about the system rather quickly.The final PDSA step is to “Act.” As PDSA cycles are completed, teams determine if the intervention should be adopted, adapted, or aban-doned. Adoption simply means the PDSA was successful and the team views the intervention as useful in support of the corresponding key driver. This intervention could be eventually imple-mented as a permanent change and/or tested more broadly, with more patients, or in different settings. Adapting an intervention indicates the team needs to improve the intervention and retest. Abandon is the decision to drop the intervention because the data reflected no appreciable change, the intervention was too burdensome, or the intervention proved unreliable. This adopt, adapt, abandon conclusion should be noted on PDSA tracking forms and if adopted or abandoned, then noted on the key driver diagram. More PDSA CyclesWhen interventions require adaptions, additional PDSA cycles should be planned and completed as methodically as prior PDSA cycles. Concurrent PDSA cycles can be completed by leveraging multiple teams in different areas, though careful planning must be made not to deploy too many PDSAs simultaneously to affect the same key driver. If there is improvement noted while there are multiple simultaneous PDSAs, it may be unclear which intervention yielded the improve-ment. The inpatient asthma team conducted the “Checklist” PDSA cycle in one acute care unit and tested “filling controller medications before discharge” in another acute care unit. Select Change for Scale-UpWhen interventions are successful in the small- scale test of change, the team should plan to expand the intervention testing. These tests can be documented and memorialized using a PDSA ramp summary (Fig. 9.6). If the intervention is not successful under altered conditions, be sure to evaluate an adaption to the intervention and conduct another PDSA. Consider if the adaption may negatively affect the areas where the testing was initially completed. If the intervention proves to be successful when scaled up, the intervention is ready for broad adoption. Plan for Implementation and SustainabilityIf interventions are to be implemented, corre-sponding changes need to happen to successfully implement the intervention. Policies, guidelines, education, and Standard Work Instructions may need to be updated to ensure the intervention is made permanent. Testing for a day or a week could be well tolerated, even a temporary ramp-

181 up of tests could be absorbed and fulfilled reli-ably. However, large-scale and more permanent implementation and sustainably can only exist when the intervention becomes the new normal. The asthma bundle checklist education was deliv-ered to the inpatient staff on all acute care units, and an electronic health record bundle checklist was built. Policies were updated to include infor-mation defining the standard work for the bundle checklist completion.Sustaining the observed gains is an essential component of improvement work. Using tactics like hardwiring interventions helps guarantee the successful implementation of the interventions even when the focus is turned to other improve-ment projects [5]. Collecting and reviewing data regularly helps to detect when an intervention begins to fail. If the data identify the intervention as being used regularly, the data collection fre-quency and sample size may be reduced [5]. Assigning a group or person to monitor the data over time will ensure early recognition of a change in the data. SpreadSpreading the work to other relevant areas will include the same considerations as when imple-menting the work in a single area. Policies, guideline, tools, and Standard Work Instructions may need to be altered to support the new inter-ventions. Determine what other areas would ben-efit from the intervention(s), engage the leaders from those areas to share the success of the inter-ventions, and make the case for the change [6]. There is value in sharing the data and a story of how the intervention has led to improvement, as this may accelerate change adoption in new areas. It is important to identify other organizational activity that may conflict with the spread of the team’s work and plan the pace and direction of the spread accordingly. DataLearning from data during the course of an improvement project is essential. Data are col-lected, analyzed, and acted upon prior to the kickoff of the project, for the duration of active improvement (PDSA) and during the sustain phase where intermittent monitoring ensures continued success of the improvement.Data collected prior to the commencement of improvement work identifies the prevalence and significance of the problem and can be used to breakdown a problem categorically (Pareto chart) [7]. A Pareto chart focuses the direction and pri-orities of the improvement work. Pre- improvement data will create the baseline measure of the outcome or process being targeted TestDescription:Cycle 1Cycle 2Cycle 3Cycle 4TestPopulation:Location oftest:Date/Time(From – To):Prediction:Action taken:Test Results:Action –Adapt, Adopt,AbandonDocumentAFTER testDocument BEFORE testAPSDAPSDAPSDAPSDFig. 9.6 PDSA ramp summary

182for improvement. By presenting pre- improvement data in the context of organizational strategy, the need for specific improvement work can be made more compelling.Data collected during the improvement work should clearly align with the desired outcome of the work. Occasionally, direct measurement may not be possible, requiring a proxy measure to be carefully selected. A proxy measure will allow for change to be detected rapidly, but with a strong correlation between the interventions and the outcome. Data collected during active improvement will assist teams with decision- making relevant to success or failure of interven-tions, determine next steps, and may identify problems that were not initially apparent. Data should first be collected and displayed using a simple run chart and progress to more sophisti-cated means using appropriate control charts (often referred to as a Statistical Process Control Chart or Shewhart chart).Post-implementation data should be moni-tored to track the effectiveness of the interven-tions on the desired outcome/process metric. These data can be measured with a sustain phase plan. The burden of data monitoring should be reduced during the sustainability phase of any project and should be measured using a control chart (Shewhart chart) so that processes moving out of “control” can be easily identified. Types of Quality Improvement MeasuresThere are a number of strategies to categorize quality improvement measures. Avedis Donabedian succinctly categorized measures in three ways: structure, process, and outcome [8]. Others have considered a fourth category for quality improvement measures, a balancing measure.Structural measures include those measures that represent the physical space and equipment used to deliver care or manage a process. Some think of these as measures of the environment – and are clearly distinct from process, outcome, and balancing measures. This is occasionally binary, meaning it either exists or it does not, and for that reason is often easier to measure. It is believed that structural measures are often foun-dational to the ability for subsequent process or outcome measures to be achieved. For example, a structural measure in our asthma vignette may be the availability of a care manager position for asthma patients. This role either exists or doesn’t and doesn’t address how care management occurs, how the care manager is contacted, or even how frequently/infrequently patients use the emergency department or are compliant with home medications when contacted by the care manager.Process measures generally represent one or more specific steps of a process that are thought to possibly lead to a particular desired out-come. Most outcomes are derived from a struc-ture that supports success and the multiple processes that each contribute collectively to an outcome. Some have used the analogy of a ladder to Donabedian’s structure, process, and outcome measures. The ground that the ladder is seated on is the structural measure, and the rungs of the ladder serve as individual process measures, each contributing to the journey to the top of the ladder, which is the desired out-come. The inpatient asthma team studied and tested the administration of the influenza vac-cine to patients admitted to the hospital as a process measure – asking if they were able to reliably administer the vaccine.Thirdly, outcome measures represent the state of the patient or population of patients and what is important. It may demonstrate overall system- level performance for patients or the financial picture associated with the improvement. Said differently, an outcome measure is the actual thing that we want to change or improve in the end. In our asthma examples, the outcome mea-sures are emergency department (ED) visits and inpatient hospital admissions. The idea of avoid-ing either situation is important to the patient (and his/her family).Lastly, balancing measures are considered to ensure an improvement in one area is not nega-tively impacting another area. It may be difficult to identify balancing measures, and the impact

183may be realized in a clinical metric or an admin-istrative/financial metric. While trying to reduce inpatient admissions, the emergency department may keep asthma patients in the emergency department longer or send more of them home when an admission would possibly have been wiser. Therefore, admissions decrease, but ED length-of-stay (LOS), ED revisits, and patient satisfaction could each be negatively impacted. It is difficult, often times, to capture all potential balancing measures, but great thought should be given to try and ensure the breadth of balancing measures are captured. Features of a Good MeasureData considerations and metric determinations start with the aim statement. Useful metrics ensure buy-in for the improvement project [9]. When selecting the aim or goal, consider the fol-lowing features of a useful metric: 1. Understandable – The metric is defined in such a way that it conveys, at a glance, what it is measuring and how it is derived. When cre-ating a data definition for the metric, keep the metric clean and simple to understand without multiple exceptions that potentially add unnecessary complexity to the metric and the data collection process. Test out a metric by trying to explain it to someone outside the improvement team. 2. Credible – Credible is offering reasonable grounds for being believed. The credibility of a metric can be increased by staying consis-tent, using the best evidence, citing definitions from outside organizations, or simply having an understandable metric. 3. Comparable – Being able to compare a metric across time periods, groups of users, national benchmarks, or competitors allows the improvement team or sponsor to understand the metric performance. 4. Actionable – This is by far the most important criterion for a metric requiring consideration of what will be done differently based on changes in the number. If the improvement leader or team has little potential to influence change, then the improvement project should be either abandoned, turned over to a team that has the ability to influence, or enlist a sponsor with span of control to champion the change. 5. Aligned – To assess if a metric is aligned, it would be useful to ask, “How does the metric relate to other metrics in the hospital and the hospital’s overall objectives?” Improvement should be tied to strategy, whether organiza-tional, departmental, or local. 6. Accessible – To assess if a metric is accessi-ble, consider the following:• Where are the data available?• Are the data collected manually?• How much manipulation do the data need in order to be in the desired format?• How many calculations does the metric involve?• How many people have to touch the metric? VariationTo understand changes in the data and what those changes mean, improvement teams need to understand what random and nonrandom varia-tion, as well as common cause and special cause variation [7]. When run charts are used to display data, random and nonrandom variation differenti-ates change that occurs randomly or change that is distinct. For control charts, variation is consid-ered common cause if it represents the ebb and flow of a process that is unchanged, whereas spe-cial cause variation indicates the improvement work has either positively or even negatively impacted the measure. Recognizing variation characteristics that signify change in the data alerts the team to explore the reason for the change.There are two main types of variation in any systems: intended and unintended variation. Intended variation is an important part of effec-tive, patient-centered healthcare. It is similar to the concept of variety – one size does not fit all. It is often called purposeful, planned, guided, or

184considered. It is acceptable to both the healthcare consumer and those who work within the deliv-ery system. Unintended variation is due to changes introduced into healthcare structure or process that are not purposeful, planned, or guided. This type of variation creates inefficien-cies, waste, rework, ineffective care, errors, and injuries. Most healthcare improvement projects focus on reduction of these unwanted variations as they are unwelcomed by the consumer and those within the delivery system.A basic premise of improvement work is the idea that variation is a measure of quality and variation has one of two causes: common cause or special cause. Knowing the source of variation and identifying the nature of variation is a critical quality improvement skill. Common cause varia-tions are those causes inherent in the process over time that affect everyone working in the process and affect all outcomes of the process. Conversely, special causes are not part of the process all the time or do not affect everyone but arise because of specific circumstances or interventions.This premise and understanding of the causes of variation become important as data are collected, and the team determines the next steps in the work. Leaders and teams must be vigilant in not unduly reacting to common cause variations but certainly need to be poised to react to special cause variation. Teams should be able to clearly understand what their data are telling them, so they are able to con-vey the improvement story and make decisions about the next steps in the work. Some basic descriptive statistical analysis review may be nec-essary. Understanding the type and distribution of the data will allow the team to determine how to best summarize, present, and analyze data during all phases of the improvement project. More Considerations for Data and MeasurementIdentifying the process or outcome to be measured will help frame the baseline data so that it is in alignment with the aim statement. This numerical baseline data can be easily represented in a histo-gram (a chart that relates the frequency of one variable on the Y-axis, over time, on the X-axis). (See Key Point Box 9.1) The chart below (Fig. 9.7) reflects the rising frequency of ED asthma visits as a count per year and is not adjusted as a compari-son to all patients with asthma or all emergency department visit reasons.Numerical data can also be divided into cate-gories related to the frequency of problem types for process failure and is best visualized in a Pareto chart. A Pareto chart contains discrete X-axis bars representing the frequency of each problem and the Y-axis depicts the % of the fail-ures attributed to each problem. Generally, a line 201602550751001251501752002017ED Asthma Visits/YearED Asthma Visits/Year2018Fig. 9.7 Asthma ED visits/yearKey Point Box 9.1A histogram is a graphical representation using bars to depict the frequency of con-tinuous variables as they fall into a given range. The height of each bar indicates how many fall into each category range.

185is generated depicting a cumulative frequency from left to right on the chart. A Pareto chart helps identify problems may be causing more failures and may guide intervention testing (PDSAs). In Fig. 9.8, the reasons patients with asthma are not compliant with their prescribed medications are identified by category and fre-quency. This information was collected through a patient questionnaire, and the data were entered to generate the Pareto chart. The information was shared among the teams with each team being charged to identify potential interventions they wanted to test using the Plan-Do-Study-Act cycle. These interventions were included in the key driver diagram. Data DisplayQuality improvement data (both pre- improvement data and PDSA data) can be plot-ted using a run chart [7]. A run chart can plot a count or rate (Numerator over Denominator) across time using points connected by lines. This type of visual chart reflects a basic under-standing of the data in comparison to a standard grouping (histogram), is a quick and easy way to begin the tracking of data, and allows for a clear picture of the performance of the process or outcome. Interventions tested through the PDSA cycle may impact the process/outcome and change to the performance of the system, which can be seen in the asthma ED visit rate run chart. In the case of the asthma patients, data are plotted as the number of ED asthma vis-its as a numerator over the number of patients in the hospital’s asthma registry (denominator) per month. The “n” represents the number of patients in the asthma registry for that month. The time series of the run chart should be dis-played on the X-axis and the rate on the Y-axis.A run chart should optimally have the follow-ing elements:• Labels along both axes with a clear descrip-tion of the measurement (%, days, weeks, minutes)• Equal X- and Y-axis tick marks• Title which clearly and simplistically describes what is being plottedCumulative PercentagesCost of Medication0%10%20%% of Compliance Failures30%40%50%60%70%80%90%100%Multiple HomesUnauthorized RxNon-Compliance FactorsController UseForgot Home MedsUse of SpacerReasons for Asthma Medicatiion Non-Compliance24.8%44.0%58.9%73.8%87.9%100.0%172021212735Individual Quantities & PercentagesFig. 9.8 Reasons for asthma medication noncompliance

186• Arrow showing the desired direction of change that is an improvement• Appropriate scale• Identified and labeled goal• Annotations• Line drawn to show the median of the dataThe scale should be appropriate for both the current range of data being displayed and what future data may need to be plotted. When the scale is small compared to the range of the data set, improvement can be difficult to detect; con-versely, when the scale is too large, the data may appear to have significant gains or losses, when in fact, this is simply a product of an inappropri-ate scale. Most of the current data should fall in the middle of the chart, so variation in either direction becomes apparent.Any observer of the chart should be able to quickly understand what is being measured and if the performance is improving or declining. The chart should also display the goal of the project so teams can demonstrate what the final target is and how close the team is to accomplishing the goal set out in the Aim statement (Fig. 9.9). Teams should become accustomed to regularly annotating the chart to reflect PDSA cycles, unusual situations (i.e., abnormally low or high denominators), or any other notes that may not be remembered as the work progresses. Use the first 10–12 data points to calculate the median (the number in the middle of the data set) and plot the median on the run chart parallel to the X-axis. The median can be extended across the chart to reflect the original baseline or may shift when a system change is identified according to run chart rules, as improvement happens. The centerline in a run chart is the median value. If baseline data are not available, use the first 10–12 data points available once the data collection is possible as early improvement planning will not likely affect the early data measurement points. Run chart rules related to the median allow for the detection of changes resulting from the testing of interven-tions [10].There are several run chart rules for nonran-dom variation; the descriptions below are intended to give the reader a primer on nonran-dom variation and are not exhaustive or all- inclusive of the rules [7]. These probability rules indicate a nonrandom change in the system and alert the team that the process or outcome has changed – sometimes for the better or potentially for the worse. These rules are based on succes-sive data points and their relationship to the median. These rules do not indicate if the process is stable or “in control,” only that some intervention(s) has probably caused the observ-able change and the change is not random. Instances of nonrandom variation in data should 01/01/15(n=25603)02/01/15(n=25617)03/01/15(n=25612)04/01/15(n=25615)05/01/15(n=25668)06/01/15(n=25721)07/01/15(n=25722)08/01/15(n=25748)09/01/15(n=25901)10/01/15(n=25935)11/01/15(n=26038)12/01/15(n=25944)01/01/16(n=26096)02/01/16(n=26172)03/01/16(n=26305)04/01/16(n=26410)05/01/16(n=26538)0ED visits per 100 asthma days123456789ED VisitsED CenterlineGoal (5)Asthma Registry ED Visit RatesFig. 9.9 Asthma registry ED visit rates

187be investigated, prompting the formulation of a theory as to why the change occurred and anno-tated on the run chart. Run chart medians can be shifted or recalculated based on data changes identified within the grouping of nonrandom variation in data as long as none of the data points were also used to establish the baseline median.Run chart rules depicting nonrandom varia-tion [10]: 1. Shift – Six or more consecutive points all above or all below the median (Fig. 9.10) 2. Trend – Five or more consecutive points all going up or all going down (Fig. 9.11) 3. Runs – Alternating points in a “zigzag” pat-tern (Fig. 9.12) 4. Astronomical point – A point that is obviously and blatantly significantly different from all of the other data pointsWhen run chart data do not meet these non-random variation rules, any data changes can be considered normal or random variation. When data has been collected past the baseline stage and results in 10–12 additional data points, teams may consider abandoning a run chart in favor of one of the many types of control charts (Shewhart charts) in order to create a clearer picture of the nature of the data. Control Charts (Shewhart Charts)Control charts are similar to run charts in that they too plot data over time. Control charts differ from run charts in that they can identify the pro-cess as being “in or out of control” or the stability of the process and the ability of the process to function predictably. Control charts accomplish this predictability through the use of statistical process control to determine the stability of the system. Predictable processes follow a known pattern, and predictions are made based on that pattern, depicting clearly common cause or spe-cial cause variation. In statistics, that pattern is known as distribution. Knowing the distribution of the data allows for an expected outcome. For a stable, predictable process, 99.7% of the data points will fall within three standard deviations (+ or −) from the mean of the data. Each standard deviation away from the mean is used to identify variation from the average performance (mean). Once a data point falls outside of the third devia-tion, the process is no longer predictable and that point is identified as an outlier (a type of special cause variation). Because data falls outside the limits 0.3% of the time, it would be highly unlikely that a data point outside the control lim-its would be attributed to common cause variation.Fig. 9.10 Six points above/below the mean

188Data within the control limits could still be identified as special cause if they meet the definition for data that differs from the normal distribution. If the goal of the improvement work is to raise or lower an average, data pat-terns will exist inside the expected distribution that signals a change which indicates the inter-ventions are moving the data toward the goal (new average). These rules are discussed after some basic understanding of control charts is established.The anatomy of a control chart is as follows (Fig. 9.13):• There are an upper control limit (UCL) and a lower control limit (LCL).• Typically, the upper (UCL) and lower (LCL) control limits are ±3 standard deviations from the mean.• The centerline is the actual process mean (average).Figure 9.14 represents a normal distribution of data. If this familiar distribution is rotated to the side, a control chart becomes more understand-able. There are many available templates that allow a user to enter a time series, title, and Fig. 9.11 Trend five or more pointsFig. 9.12 Alternating points zigzag pattern

189accompanying data to ultimately generate a con-trol chart. Creating control charts through manual computation is not overly challenging but will require the use of some additional resources.The type of control chart used to display data depends on the type of data being used. The decision tree in Fig. 9.15 is helpful when choosing a control chart. Continuous vs. attri-bute data is the initial bifurcation in the deci-sion tree. Continuous data are data that have a broad range of values that could be anywhere within a range of data. Common examples of continuous data are body weight or time in sec-onds. Attribute data, on the other hand, is more discrete and usually can only take a limited set of values. For example, attribute data may be either in range or out of range. Identifying Special Cause VariationPoints that fall outside the control limits are indicative of special cause variation and require investigation. Other special cause rules also indi-cate a change to the process with the data points remaining within the control limits. The system may be performing within control, but not where the team has set the goal. The following special cause rules help to identify such changes as the interventions are tested through the PDSA cycle and are used to answer the improvement question related to “Is the change an improvement?” Most importantly, these rules allow teams to determine the need to react to the data and make changes or if the data is demonstrating normal variation and do not require mediation.12Lower Control Limit (LCL)Upper Control Limit (UCL)Centerline (CL)345678910111213..kStatisticCL = Average of StatisticUCL = CL + 3σsLCL = CL – 3σsFig. 9.13 Anatomy of control chart99.73%95.44%68.26%-1σ+1σ-2σ+2σ-3σ+3σ99.73%95.44%68.26%-1σ+1σ-2σ+2σ-3σ+3σUpperControlLimitLowerControlLimitAverageFig. 9.14 Control chart bell curve

190The rules governing special cause variation for control charts are as follows [7]: 1. Eight or more consecutive points above or below the centerline (Fig. 9.16) 2. Six or more points increasing or decreasing (Fig. 9.17) 3. Two out of three consecutive points near an upper or lower control limit (Fig. 9.18) 4. Fifteen consecutive points near the centerline (Fig. 9.19) 5. A single data point outside of the control lim-its (Fig. 9.20) Don’t Get Lost in DataQuality improvement in healthcare is a moral imperative. Each and every patient deserves high- quality care. As such, improvement teams must never allow the measurement and nuances of healthcare data to negate the fact that the data often represent real people or processes that Attribute Data(Count or Classification)Variables Data(Continuous)Type of DataCount(Incidences ornonconformities)Classification(Defective ornonconformities units)SubgroupSize of 1FixedSubgroupSizeFixedSubgroupSizeVariableSubgroupSizeVariableSubgroupSizeVariableOpportunities/SampleSizeFixedOpportunities/SampleSizeC ChartU ChartNP ChartP ChartPercentDefectiveAverage &RangeAverage &StandardDeviationIndividualMeasures &MovingRangeNumber ofDefectiveNumber ofIncidences(Count)IncidencesPer UnitX-bar & SChartsX-bar & RChartsX& MRChartsWhich Control Chart Should I Use?Fig. 9.15 Control chart decision treeUCLCLLCLFig. 9.16 Eight or more consecutive points above or below the centerlineUCLCLLCLFig. 9.17 Six or more points increasing or decreasing

191affect people. Goals and ongoing measurements often contain language about percent reductions or increases, dollar costs, failures, and special causes. Inspiring improvement requires leaders to be sure to equate numerical measures back to the people impacted by the care delivered every day. A better practice is to consider phrasing goals in terms of the number of patients, so this notion is not forgotten as teams move forward in their work. This understanding will serve as a motivator for continuous improvement.UCLCLLCLOuterone-thirdof chartFig. 9.18 Two out of three consecutive points near outer third of control limitUCLCLLCLInnerone-thirdof chartFig. 9.19 15 consecutive points close to centerlineEditors’ CommentsEach and every chapter in this textbook is important; each and every chapter is value- add for the novice as well as experienced improvement scientists. This chapter serves as a primer for the novice or casual quality improvement scientists and forward think-ing and directional for those that are more advanced in their improvement journey. Using the case vignettes, the authors mas-terfully navigate the quality improvement process using methodologies as their framework. We sincerely appreciate the authors demonstrating specific strategies that they have employed in their organiza-tion (e.g., “data element definition form”). These concrete examples are invaluable for organizations that want to use this chapter as a foundation to build upon or advance their quality improvement journey.The core of improvement science is using a roadmap in an iterative manner. The authors thoroughly explain key driver diagrams and eloquently link these to the iterative tests of change. Again, we are most appreciative of the demonstrations of how they actually implement and opera-tionalize these tools in their respective organizations.We are inundated by payers, the govern-ment regulations, and the public with requests for more and measures. The end of UCLCLLCLFig. 9.20 Single data point outside of the control limits

192 Chapter Review Questions 1. True or False – The three basic elements that constitute the framework of a key driver dia-gram are: Aim Statement, Drivers, and Interventions?Answer: True 2. Multiple Choice – What does “SMART” stand for when discussing a “SMART” Aim (goal)? A. Specific, Measurable, Attainable, Relevant, and Time-bound B. Standard, Measurable, Articulate, Range, Testable C. Standard, Mindful, Attributable, Relevant, and Testable D. Statistical, Meaningful, Attainable, Real, TangibleAnswer: A 3. Why is it important to make a prediction about a PDSA cycle?Answer: Making a prediction about the success of an intervention, before the testing, establishes a level of confidence in the inter-vention affecting the process or outcome and assists the team in determining the testing pri-ority of the interventions. By comparing the actual results to the predictions, shared learn-ing can occur. 4. True or False – Special cause variation should be investigated because this type of variation is unexpected and don’t exist in the system all the time.Answer: True 5. Which of the following should be included on any type of run chart or control chart? A. Labels on the axes and the chart B. An arrow describing the desired change of direction C. A clear and equivalent time series and tick marks appropriate for the data set D. All of the aboveAnswer: DReferences 1. 2017 CDC National Center for Health Statistics. Secondary 2017 CDC National Center for Health Statistics 2017. https://www.cdc.gov/nchs/fastats/asthma.htm 2. Patient Protection and Affordable Care Act, 42 U.S.C. § 18001 (2010). 3. Langley GJ. The improvement guide: a practical approach to enhancing organizational performance. 2nd ed. San Francisco: Jossey-Bass; 2009. 4. Reed JE, Card AJ. The problem with plan-do-study- act cycles. BMJ Qual Saf. 2016;25(3):147–52. https://doi.org/10.1136/bmjqs-2015-005076. 5. Silver SA, McQuillan R, Harel Z, et al. How to sustain change and support continuous quality improvement. Clin J Am Soc Nephrol. 2016;11(5):916–24. https://doi.org/10.2215/CJN.11501015. 6. Kliger J, Singer S, Hoffman F, O’Neil E. Spreading a medication administration intervention organiza-tionwide in six hospitals. Jt Comm J Qual Patient Saf. 2012;38(2):51–60. 7. Provost LP, Murray SK. The health care data guide: learning from data for improvement. 1st ed. San Francisco: Jossey-Bass; 2011. 8. Donabedian A. The quality of care. How can it be assessed? JAMA. 1988;260(12):1743–8. https://doi.org/10.1001/jama.260.12.1743. 9. Lloyd RC. Quality health care: a guide to develop-ing and using indicators. 1st ed. Sudbury: Jones and Bartlett Publishers; 2004. 10. Perla RJ, Provost LP, Murray SK. The run chart: a simple analytical tool for learning from variation in healthcare processes. BMJ Qual Saf. 2011;20(1):46–51. https://doi.org/10.1136/bmjqs.2009.037895.the chapter nicely builds on the need to have pertinent and solid measures with how to best use data. It is not expected that this chapter is completely thorough; indeed, the Editors refer the reader to Lloyd Provost and Sandra Murray’s definitive and expan-sive textbook on data for quality improve-ment [7]. However, the authors of this chapter demonstrate the value of data, how to be wary of data, and how to best use data to create measures that matter.Each and every chapter in this textbook is a value-add. This chapter is crucial. We strategically placed this as the ninth chap-ter so that the improvement scientist is primed at this point of their reading journey to become committed to the quality improvement methodologies as outlined in this chapter.

193© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 R. K. Shah, S. A. Godambe (eds.), Patient Safety and Quality Improvement in Healthcare, https://doi.org/10.1007/978-3-030-55829-1_10Designing Improvement Teams for SuccessNicole M. Leone and Anupama Subramony Opening Question/ProblemQuality improvement opportunities in complex healthcare settings are omnipresent; this chapter describes the importance of creating multidisci-plinary teams and leading these teams effectively and efficiently. Intentional thought on creating an effective team with a burning platform will greatly aid in achieving the aim of the initiative. IntroductionThe creation of effective teams is a key element of all organizations, though may be especially important in healthcare settings, given the com-plexity of healthcare systems. In general, a team N. M. Leone · A. Subramony (*) Cohen Children’s Medical Center, Zucker School of Medicine at Hofstra/Northwell, Department of Pediatrics, New Hyde Park, NY, USAe-mail: [email protected]; [email protected] Objectives• Describe the importance of high- functioning teams in driving quality improvement in healthcare settings.• Generate strategies to build and lead effective teams using improvement sci-ence tools and organizational tactics.• Use principles from psychology of change to create a burning platform to drive change.Vignette 10.1Over the last few years, the infection con-trol team of a children’s hospital within a larger hospital system was focused on help-ing teams decrease central line-associated bloodstream infections and catheter- associated urinary tract infections. In this time period, the number of nosocomial C. difficile infections has risen. At a system- level quality meeting, C. difficile rates for the children’s hospital system are pre-sented, showing a rate significantly higher than the benchmark. The pediatric Chief Medical Officer returns from the meeting and charges the quality team to create an action plan and reduce rates as soon as possible.

194is defined as a group working collaboratively and interdependently to achieve a common goal, to which they are all held accountable [1]. In many healthcare organizations, teams are developed using staff with varying knowledge, skills, and experience to help solve complex problems and create pragmatic, innovative solutions. Keeping these teams on track to achieve their aims requires care and purpose. This chapter will review the development of improvement teams, how to make these teams productive, and how to keep the teams strategically aligned with a burning platform. Evolution of Teams for Quality ImprovementQuality improvement initiatives have their roots in teams. The methodologies that constitute improvement science were originally used in the scientific approach to improving the efficiency of manufacturing processes [2]. Key leaders include pioneers such as Frederick Winslow Taylor, W. Edwards Deming, Walter Shewhart, Joseph M. Juran, and Taiichi Ohno who in aggregate developed and perfected ways for organizations, specifically manufacturing, to be more effective and efficient in producing error-free products [2–7]. Central to these methodologies was the use of teams to analyze a problem, design tests of change, test solutions, and determine and mea-sure metrics of success. What Is a Team?Underlying improvement projects is a well- crafted team who work together to carry out a change process. Members of the team work col-laboratively for a shared purpose, with shared responsibility for achieving results. Teams can be used when the problem to be addressed is com-plex, when learning of the system is a necessary prerequisite, when there is no clear answer to a problem, when innovative ideas are needed, and when cross-collaboration among differing disci-plines are necessary. For teams to be successful, they need clear and attainable goals, an appropri-ately scoped initiative, expertise, and resources from across the organization [1]. An important distinction is between a team and a working group. A working group’s output is a sum of what the individuals in the group attain; in contrast, a team’s performance represents both individual work and collective work that represents the joint contribution of multiple team members [8]. Working groups are used for information sharing, to provide counsel, and to help individuals improve their performance on discrete initiatives. They usually include a clearly defined leader, involve individual work products, and are mea-sured by their influence on others; in contrast, teams focus include multiple leaders in shared leadership roles, have both individual and mutual accountability, develop collective work products, include open-ended discussions and active problem- solving group discussions, and are mea-sured by assessing their collective work products [8] (Key Points Box 10.1).Vignette 10.2The Chief Medical Officer of the pediatric hospital has apprised the Chief Quality Officer of the respective hospital about the increased C. difficile rate and the discus-sion of this at a system-level meeting. Understanding the urgency to fix the issue, she quickly alerts the medical directors of the inpatient units to stop sending tests for C. difficile unnecessarily. The leadership of the Oncology unit expresses some reluc-tance at the shotgun approach to a problem without systematically assessing the poten-tial myriad contributing factors.Key Points Box 10.1Teams are a group of people who work col-laboratively and interdependently towards an aim; the output of a team represents col-lective work jointly produced by multiple team members.

195 Identifying Team Members in an OrganizationTeams play such a significant role in driving improvement in organizations that the strategic design and crafting of a team are vital. The team’s inclusion of experts in the process that needs to be changed is a necessary first step (commonly referred to as subject matter experts which can be any level of employee). In fact, studies done at the Hawthorne factory in the mid-twentieth century showed that engaging frontline staff in redesigning a process improved both the efficiency of the process and improved and sustained reliability to the process [4]. Deferring to the expertise of the frontline staff to be able to improve a process with practical context, as opposed to having leaders’ remote to the frontline, was thought to be a contributor to the success. Walter Shewhart, who developed the statistical process control chart, a key tool in improvement science, made a strong case for engagement of frontline staff in improvement activities. He championed the development and deployment of statistical process control charts to be displayed on manufacturing floors to allow frontline staff to identify special cause varia-tion, stop the line, conduct just-in-time analy-ses, and find and fix issues as they arose [4]. This frontline engagement approach to improv-ing the efficiency of a process was transforma-tive, and its effect persists in improvement science methodologies [9]. A team cannot be assembled without pertinent subject matter experts.When building a team, understanding how stakeholders view potential improvements is imperative to developing the appropriate team composition. While it may be easier to identify stakeholders who are enthusiastic and will drive change, it may be more challenging to find those who are indifferent or opposed (commonly referred to as contrarians) and engage them in an improvement team. Effective teams are not cre-ated with all those individuals that may be the easiest to work with and the most convenient. Indeed, highly effective teams include staunch supporters and similarly strong contrarians.A stakeholder analysis illustrates the position of each stakeholder with regard to the project, can be used as a tool to understand why they may not be at the level of engagement necessary, and then helps the team determine how much effort should be spent moving that commitment from its current status to that which is necessary to achieve their goals [1]. A representative team composition can be built using the rubric shown in Fig. 10.1. This rubric categorizes stakeholders into four main categories: those with high influ-ence and high interest, high influence and low interest, low influence and high interest, and low influence and low interest. Those with high influ-ence and high interest should be fully incorpo-rated into the team, as they have the resources and passion to drive the project forward. There should be a respectful collaboration with stake-holders of more mixed influence and interest, particularly those who are not initially supportive but offer high potential for cooperation (the team may direct specific efforts to engage these Meet their needsLess importantShow considerationKey playerMeet their needsLess importantShow considerationKey player• Formal or informal influence• Frequent contact to engage and request expertise to increase level interest• Crucial to success• Inform through general communication• Harness interest to be active in conducting action plans• Can serve as ambassadorPower/InfluenceInterestFig. 10.1 Stakeholder analysis

196individuals). Those with high influence and more neutral positions may have resources that could be offered to the team and should be informed and engaged. The team that identifies stakehold-ers who are non-supportive with low potential for cooperation can be prepared to deal with these individuals [1, 10].An alternate tool that can be used to identify key team members is the SIPOC tool, which maps a process and includes all Suppliers, Inputs, Processes, Outputs, and Customers [Fig. 10.2] [1]. Defining the suppliers and inputs of a pro-cess, as well as the customers who will use the outputs of the process, can be helpful in more complex projects to assure all key stakeholders, especially those who are not immediately appar-ent, are considered [1]. Using this tool to identify hidden team members can prevent the inadver-tent exclusion of a member that would potentially be impactful. Identifying key stakeholders late in the game and necessitating a latecomer to “catch up” could possibly derail work already in prog-ress [1, 11]. The SIPOC tool may have the added benefit of starting out the improvement team with a narrow focus on a single process, avoiding the common pitfall of working on a problem that is too broad. In addition to identifying team mem-bers, these tools can also aid in identifying other valuable stakeholders who may not be part of day-to-day operations, but may prove to be valu-able allies as the team progresses; these other stakeholders should of course be kept abreast of progress on the team and can be used as consul-tants to the project in an as needed capacity [1] (Key Points Box 10.2). Considerations Around Designing a TeamWhile these tools may assure that a team accu-rately reflects the need of the project, it is also equally important to be mindful of the size of the SuppliersInputsProcessOutputsCustomer• Patient• Attending Physicians• Resident Physicians• Nurse Managers• Bedside Nurses• Nursing Assistants• Patient Care Assistants• Antibiotic Stewardship• Environmental Services• Infection control• Microbiology department• Antibiotics• Cleaning rooms and equipment• Infection control techniques• Laxatives• Indications to test patients for C, diff• Lab specimens for C. diff testing• C. diff PCR testPatient with positiveC. difficile test result1. Predisposing factors: • Antibiotic overuse • Inadequate cleaning of room/equipment • Poor infection controls techniques • Laxative use2. Patient at risk for C. diff infection3. Patient develops loose stools4. Medical team setermines C.diff testing indicated5. Nursing sends stool specimen to lab6. C. diff PCR test run in microbiology lab7. C. diff test result• Nosocomial C.difficile infection• Patients• FamiliesFig. 10.2 SIPOC toolKey Points Box 10.2Developing cross-functional teams for improvement projects requires thoughtful consideration; specific tools such as a SIPOC can help provide structure to designing a team for success.

197team. Smaller teams may be more agile, more innovative, and ultimately, more effective. Examples of these teams include many internet start-ups including WhatsApp and Amazon, whose CEO Jeff Bezos coined the “two-pizza rule,” suggesting that any team that could not be fed by two pizzas was too big [12]. Strategies to assure the optimal size of teams include break-ing down improvement teams into smaller proj-ects, empowering team members to be the decision- makers, building a trust culture to allow team members to be agile in decision-making, focusing on informal ways of sharing informa-tion as opposed to taking formal minutes or cre-ating presentations which can be less efficient, building platforms where team members are actively communicating (collaboration plat-forms), and effectively using technology to pro-gram manage [13].Over and beyond choosing the right constit-uents for and the size of the team, other consid-erations in creating effective teams include understanding how members of these teams will work together. There is science behind the concept of which elements make teams more effective at their work, including problem-solving. While the functional role of team members is important, the psychological role that a team member plays may be equally important [14]. Teamwork is dependent on ele-ments such as affective states, behavioral pro-cesses, and cognitive states of teams, much of which is influenced by the personality types of individual team members [14]. When teams are being created, it is imperative to craft a well-balanced mix of team member personalities including those team members who are results-oriented, relationship- focused, process and rule followers, innovative and disruptive think-ers, and pragmatic [15]. A mix of personality types allows for effective teamwork that trans-lates inputs (expertise, capabilities, knowl-edge) into outputs (process changes, products). The diversity of a team cannot be understated; it is well demonstrated that diverse groups in terms of race, ethnicity, and gender are more rigorous in problem-solving [16] (Key Points Box 10.3).How teams work together is integral to how they will function. Increasingly, it is becoming apparent that effective teams have several key components: mutual trust among team members rooted in emotionally intelligent behaviors, a clear group identity, and a belief that the team is more effective working together compared to working alone. Central to this is the idea is the concept that self-awareness and regulation of emotions translate into the group setting; teams should work on setting standards for behaviors that help foster trust, group identity, and a feeling of group efficacy [17]. Roles and Responsibilities for Team MembersOnce stakeholders are identified, role delineation is imperative. The core team is comprised of sev-eral important team roles (which should be delin-eated on the project charter): team member, team leader, coach, and senior sponsor.Team members contribute to the overall suc-cess of the team by sharing their knowledge and experience during team meetings and participat-ing in implementing changes. As previously dis-cussed, it is imperative to include members of the clinical team, particularly the frontline staff or those in the know of current work processes, who will be able to give context to key drivers to achieve an aim within a clinical setting. These members include physicians, nurses, and other clinical staff. Sometimes, these individuals may be seen as informal leaders in their work environ-ments and may have key relationships that allow Key Points Box 10.3Designing a team should take into account key stakeholders, level of engagement, skillset, and personality types of individual team members; teams should be the right size to be able to work together and have mutual trust that the team will be more effective working together compared to team members alone.

198them to operate as key influencers to promote change. Teams should include subject matter experts to give content expertise and provide deep knowledge in the area for improvement. Many teams can and should include patients and/or family members on their improvement teams. Inclusion of the patient voice lends credence to the improvement project and assures that the changes are patient-centered (Fig. 10.3).Even with team members whose members have formal authority within an organization, effective teams need a clearly defined leader who runs the day-to-day operations, leads meet-ings, and may become a liaison to clinical and executive leadership. Often times, the team leader is not the team member with the highest title or most power in the organization. The team leader serves as the communication link between the team and the rest of the organization. It is the leader’s responsibility to maintain the data related to the project, as well as supervise prepa-ration of reports and presentations; the team leader is the public and organizational face of the team and is accountable [1]. Returning to the C. difficile reduction vignette, a physician in the hematology- oncology unit who is enthusiastic and results- oriented may be a clear leader for the project.If the team is an improvement team, there may be an improvement coach (referred by many dif-Fig. 10.3 Roles and responsibilities for improvement projects [1]. Reprinted with permission from Scholtes et al. [1]

199ferent titles in different organizations). The coach is someone experienced in quality improvement who can bring in improvement science as needed; of course, it is unreasonable for them to teach quality improvement principles to a team. Oftentimes, in improvement work, the team leader may also fill the role of coach, depending on their background and skillset. An effective coach should advise how to best apply improve-ment science methodology to the problem, demonstrate appropriate data collection and anal-ysis techniques to ensure accurate interpretation of results, and then encourage decisions be made based on the evidence gathered [1]. In the previ-ous C. difficile example, the Chief Quality Officer, or perhaps a project manager or data ana-lyst within the hospital’s quality division, would fill this role.Lastly, an executive sponsor is oftentimes necessary, especially when there may be chal-lenges in carrying out specific interventions, a significant change needs to occur outside the limits of the team’s authority, or considerable resources may be needed. The executive sponsor is a leader in the organization who serves to lift barriers by providing resources or influencing change and will ensure alignment of the proj-ect’s aim with the organizational goals, thereby increasing the likelihood of success [1]. In the case vignette for this chapter, the C. difficile reduction initiative, the Chief Medical Officer is the logical executive sponsor. Memorializing both the aim of a project and key team members in a project charter or other formal document solidifies team roles and responsibilities as well as communicates to senior leadership the scope of an improvement project; some organizations take the additional step of having the executive sponsor sign this formal document as a team charter [18] (Key Points Box 10.4). Running Effective MeetingsRunning effective meetings is a crucial skill in improvement activities and one of the many important responsibilities of the team leader; indeed an entire chapter can be written on just this topic of teamwork. Time spent in meetings has to be of value; if meetings are ineffective, in addition to the time wasted in a meeting, engage-ment and goodwill in working on the team can dissipate quickly. A careful assessment and Key Points Box 10.4Teams are made up of several important key roles including, for example: team members, team leader, team coach, and senior or executive sponsor.Vignette 10.3Rethinking her initial shotgun approach, the Chief Quality Officer assembles a team to develop an improvement project to decrease nosocomial C. difficile rates. She invites the head of the antimicrobial stew-ardship team, the director of infection con-trol, a quality improvement consultant, and the data analyst for quality to a meeting. They ask the Chief Medical Officer to be the executive sponsor. They review the cur-rent state and develop an aim and key driv-ers. In a kickoff planning meeting, the quality improvement consultant suggests the team produce a SIPOC tool to help identify additional team members. They identify a lab technician and an antimicro-bial stewardship champion as additional members and, after stakeholder analysis, recognize they currently have minimal interest or engagement with this issue. They plan to have the Chief Medical Officer socialize the project to each of them and develop a communication plan to increase their interest in the project. Later, during walk-rounds, the Chief Quality Officer meets a passionate nurse who she witnesses peer coaching a surgical fellow on proper infection control practices. The Chief Quality Officer approaches the nurse to ask her perspective on the biggest safety issues involving nosocomial C. difficile, and she eagerly voices her suggestions and concerns.

200purposeful leadership in meetings are integral, therefore, in both running effective meetings and assuring the team achieves its aims.Preparation for a meeting is vital in assuring that the meeting is a success. A first and impor-tant step is to establish clear goals for a meeting. Given the time that a meeting takes and the intru-sion into day-to-day work, a deliberate approach to agenda creation can dramatically improve both the efficiency of a meeting and the engagement of team members [19]. A key element in laying the groundwork for a meeting is to elicit input from team members on what topics need discus-sion in a larger forum; involving team members in agenda creation has the added benefit of improving engagement and attention during the meeting. Agenda items should be pertinent to the entire team; they should represent issues that involve a coordinated, interdependent response. One recommendation is to list agenda items as questions to focus participants’ discussion to answer specific questions. In addition, each agenda item should have a purpose – to share information, solicit input for a decision, or to col-laborate to make a decision; clarity around this can aid team members in effectively participat-ing. Team leaders should consider that it may be more efficient to share information through alter-nate routes rather than use meeting time. Distributing the agenda ahead of the meeting and being prescriptive regarding what people should prepare will ensure that the time spent together is most efficient. Assuring that each agenda item has a specific time allotted to it as well as a facili-tator identified ensures that team members under-stand their specific role in a meeting [19]. In addition to timing specific agenda items, consid-erations should be made around duration of the meeting; meetings that are 30 or 45 minutes and not stacked back-to-back (in succession) may give team members the ability to tackle other business in between meetings and allow meetings to start on time which in and of itself will increase efficiency of the team [20, 21].The team leader must lay the groundwork for collaboration for the team, especially in interdis-ciplinary teams. It is imperative for improvement teams to understand the global aim for improve-ment efforts, at each step of the way. Using the initial minutes of a meeting to re-establish the aims of a specific project and the rationale for the meeting may be a strategy to prime team mem-bers for the meeting. While it may be repetitive, understanding and confirming the “why” of the project develops team bonds and underscores the rationale to spend time on the initiative. For example, in healthcare environments, teams that include patient/family stories or safety stories help set the tone for the meeting and tie team members to the mission of both the improvement effort and the organization.Facilitation of meetings is an important role of the team leader. Starting with introductions may help promote collaboration and respectful com-munication, especially in teams where traditional hierarchies may impact the way team members interact with each other. To assure that all mem-bers are treated respectfully and those who are not considered formal leaders feel empowered to speak, simple icebreakers or other tools to pro-mote collaboration may be important. Informal chitchat may help to break the ice and pull a team together – leaving time at the beginning of a meeting to promote socialization has been shown to promote team work [22]. Team leaders are responsible for facilitating the discussion during a meeting. Generally speaking, leaders of meet-ings should take an inquisitive approach – asking questions, probing for answers, modeling active listening, and drawing out reluctant participants. To assure all ideas are brought forth, one consid-eration is to allow for individuals to respond to specific questions via web-polling or by writing down ideas before sharing as a group [20]. An example of this is “brain writing” which instructs team members to individually reflect and write down ideas before sharing [1]. Other strategies to keep meetings efficient include controlling the size of the meeting so that there is robust discus-sion and yet everyone feels engaged, corralling the conversation so that it is pertinent and does not go off-tangent as well as preventing a few from monopolizing the conversation [23].In this era, use of devices such as cell phones and laptops may interrupt and impede free flow of conversation and decrease members’ atten-

201tiveness to the project. The team leader should be encouraged to set ground rules regarding using devices. Perhaps the onus on controlling wandering attention falls to the team leader to run a tight, focused, meeting; indeed, the value of the meeting should compel attention and hopefully prevent drifting attention. Furthermore, limiting membership in team meetings to core people can decrease device use and inattention [24].However well planned a meeting is, if the meeting does not conclude with a clear action plan, many of the gains from the meeting may be lost. One helpful strategy is concluding a meeting 5–10 minutes before the end of the allotted time to do a recap and confirm an action plan and assigned responsibilities. Assigning roles as explicitly as possible with specific tasks and due dates and then sending out a structured meeting summary to memorialize the discussions into a document serves to commit the team to the action promised [22] (Key Points Box 10.5). Creating a Burning PlatformMultidisciplinary teams may need assistance from leaders to develop true cohesion. Creating a sense of urgency behind the desired changes serves to unify team members and make it easier to attain buy-in from stakeholders. The burning platform metaphor is based on a true event and has been used for many years to illustrate a high level of urgency leading to change [25]. In July of 1988, a catastrophic explosion occurred on the Piper Alpha oil-drilling platform in the North Sea off the coast of Scotland, where over 200 crew members were employed. A superintendent, 1 of Key Points Box 10.5To run an effective meeting, team leaders should prepare purposefully, set team norms for behavior, and facilitate to pro-mote collaboration, especially in diverse teams.Vignette 10.4The Chief Quality Officer calls a kickoff meeting for a new C. difficile reduction team. In addition to the members from the previous meeting, she invites the nurse who she met on the hematology-oncology unit to provide clinical expertise, a fellow and physician who practice on the unit, a repre-sentative from housekeeping, the medical director of the unit, and a pediatric resident. She designated a quality consultant and data specialist for her team. She asks the physician and nurse to co-lead the team with the support from the quality consul-tant and coaching from the Chief Quality Officer. They develop the scope of the proj-ect and sign a charter, cosigned by the Chief Medical Officer who serves as the executive sponsor.The team started with introductions, and the clinical leaders shared the problem and shared a patient story about an adolescent boy with cancer who was close to discharge when he developed nosocomial C. difficile, which prolonged his hospitalization, and he had to miss his senior prom. The team leaders use an inquisitive approach to stim-ulate discussion around ideas for key driv-ers and ultimately focus on working on lab stewardship as well as improving house-keeping effectiveness as early PDSA cycles. Despite hesitation of some staff to initially speak up due to one of the physi-cians monopolizing the conversation, the team leaders have enthusiastically requested input from all team members. Over time, nosocomial C. difficile rates dropped by half by changing multiple pro-cesses: cleaning of rooms, nursing docu-mentation of stools, algorithms for testing, infection control practices, and overall antibiotic utilization. After going 30 days without an infection, the team walks the units to congratulate the frontline on this milestone.

202only 63 men to survive, recounted the decision he faced shortly after the explosion: jump approxi-mately 15 stories off the platform into extremely frigid ocean waters or remain on the burning plat-form. Though he knew jumping into the cold water was extremely risky and would likely lead to his death were he not to be rescued quickly, he believed staying on the platform would lead to certain death. He chose to make a frightening and potentially fatal decision because he believed the status quo, or resisting the change, was too costly. We can see similar sentiments during our imple-mentation of process improvements: major changes can be frightening and risky to some; executing these changes often requires true deter-mination to act [25]. While a structured approach to creating a team is important, it may only go so far; leading change is as much about understand-ing the culture of an organization as it is about the psychology of the people that make it up. In healthcare organizations, similar to other larger organizations that use cross-functional teams, it is imperative to understand strategies to manage diverse teams from different backgrounds [26].All too often, improvement leaders will encounter resistance during the course of their project, stemming from concerns with loss of control, fear of the unknown, increased work-load, and a lack of confidence that their current skills will translate to success with a new process [27]. Effective quality improvement efforts must have strategies to negotiate this resistance, built on psychology of change principles. The Institute for Healthcare Improvement developed a Psychology of Change Framework which describes activating people’s agency through unleashing intrinsic motivation, co-designing people-driven change, co-producing via an authentic relationship, distributing power, and adapting in action [28] (Fig. 10.4).Unleashing intrinsic motivation involves enabling team members to use reasons that are personally motivating by themselves in order to drive change, not being forced to carry out an improvement in which they have no interest [28]. Crafting a powerful narrative behind the team’s purpose and goals, or offering time for members to share personal stories, can unite team members and launch improvement efforts [29]. Incorporating concepts from motivational task design can help access people’s intrinsic motiva-tions, which is generally more sustainable than Fig. 10.4 The IHI Psychology of Change Framework [28]. Reprinted from www.IHI.org with permission of the Institute for Healthcare Improvement (IHI), ©2019

203extrinsic motivation, such as earning monetary rewards or avoiding punishment [30]. Psychologists have shown that intrinsic motiva-tors include the experience of meaningfulness (the task is important), the feeling of ultimate responsibility (the outcome is dependent on the team member’s performance), and visible results (allowing the team member to get real-time feed-back on their performance) [30].Co-designing people-driven change speaks to the idea that the team consists of the people that it is most likely to affect – building a team with the right key stakeholders including frontline par-ticipants avoids the consultant model, whereby improvement teams swoop in and swoop out after the project is completed [30]. Using the SIPOC tool can be key to ensuring change is truly people-driven by identifying and including all stakeholders to develop initial improvement ideas together.Co-production in authentic relationships is about fostering respectful and responsible team dynamics, avoiding hierarchies and promoting dialogue even regarding contentious issues [28]. This concept values different perspectives and promotes an environment where everyone feels safe sharing their opinion. One of the key tools in building these relationships is the one-to-one meeting, in which leaders focus on asking open- ended questions, intently listening to the responses, and ending with a strategic exchange of resources and a clear commitment to the next steps [31].Distribution of power allows for a team to work collaboratively on a level playing field, avoiding any natural work-related hierarchies from drifting into the improvement work. Leaders should be aware of their own implicit and explicit biases and how these may lead to power imbalances within the team [28]. Ideally, improvement teams should adopt a distributed leadership structure, where responsibilities are shared among several smaller groups which may be divided based on task or location [31]. Sharing power and holding team members accountable for tasks in a positive way is motivational, relay-ing an urgency to act which is crucial to improve-ment work [32].And finally, adaption in action – actions that motivate teams to move forward, even when actions may lead to failure. Inherent in the PDSA cycle is the ability to learn from prior actions, particularly failures, and adapt the test of change for further improvement. Making this continuous learning visible to team members can be reassur-ing that their efforts are integral to the advance-ment of the improvement process, strengthening their continued participation [33] (Key Points Box 10.6). Resources• The Team Handbook online resources www.teamhandbook.com [1].• The IHI has numerous templates and tools for project management for quality improvement www.ihi.org [34].• The Healthcare Improvement Skills Center 3.0 has multiple downloads and links for quality improvement http://www.improve-mentskills.org/courseinfo/resources.cfm [35].• Harvard Business Review has numerous arti-cles strategies for team building and running effective meetings https://hbr.org/ [36].Key Points Box 10.6Strategies to create a burning platform include understanding people’s psycholog-ical nature and interests in working with a team to drive improvement.Editors’ CommentsWe challenge readers to show us any quality improvement project done by a single individual. Quality improvement is done with teams, by teams, and for teams. This chapter is at the core of qual-ity improvement. The improvement team may not be apparent (usually working behind the scenes), but without it – noth-ing substantive will occur. However, suc-

204 Chapter Review Questions 1. What is a team?Answer: A team is a group of people that work collaboratively on a complex problem; they work interdependently for a shared pur-pose. For teams to be successful, there needs to be established goals, right-sized scope, expertise, and resources from the larger organization. 2. How can the design of a team impact its effec-tiveness in achieving an aim?Answer: Identifying key stakeholders across the organization using tools such as a stakeholder analysis or SIPOC can assure that the right people are on the team. Considering team member attributes such as personality types and diversity can aid in making sure a health mix of people contribute to a team. 3. What are the different roles in a team?Answer: Key team members include the team leader, team coach, team members, and sometimes an executive sponsor; each of these roles has clear responsibilities tied to the over-all aim of the team. 4. What are some strategies that can be used to lead effective meetings?Answer: Running effective meetings involves appropriate preparation in terms of agenda creation, communication both prior to and after meetings, and successful facilitation of discussion. 5. What are some examples of how a leader might use intrinsic motivation to activate peo-ple’s agency?Answer: Sharing a personal narrative that relates to the leader’s dedication to the improve-ment work at hand can demonstrate the motiva-tion for the initiative. Stories that have specific details and use vivid imagery are often more effective. Also, providing opportunity for team members to share their own experiences and values can unify the team and elicit emotions that can help motivate efforts [37]. Another way to intrinsically motivate team members is by using the design concepts of meaningful-ness, responsibility, and results. A leader can replicate these conditions by ensuring team cess is not an accident. Building and leading teams can be taught and needs to be learned and appreciated. This chapter serves as a primer for the reader to under-stand the value of teams, how to con-struct teams, and how to work within teams.A significant learning from this chapter is the need to have role delineations for teams; without team members understand-ing the role they are going to play, a team will not function at the optimum level. The authors expound upon this point and help the reader understand the different roles and the intrinsic value in each and how they contribute to the overall success of the initiative.Teams need meetings to be productive. The bane of many administrators’ time is meetings. The authors give wonderful examples and strategies on how to make meetings effective (e.g., a distinct period at the end of each meeting to summarize the meeting, discuss action items, and assign responsibilities).The authors end where any quality improvement initiative should begin, the burning platform. If the leader, or even the team, cannot articulate the burning plat-form and the imperative for action, then others will not follow, and the quality improvement initiative will not have a foundation to start its work.Often times, improvement scientists perseverate on the data and necessary improvement charts yet forget about the teams, how to run meetings, how to create a burning platform, etc. – it is these fac-tors that probably lend the biggest part to the success of a quality improvement ini-tiative. One can find many platitudes regarding teams, the Editors strongly believe that the success of a quality improvement initiative is directly corre-lated to its focus on creating the right team and platform.

205members understand how the task fits in with the overall goal, are given autonomy in its exe-cution with the knowledge that the outcome is their direct responsibility, and are able to access real-time results to receive the feedback neces-sary to make changes [30].References 1. Scholtes PR, Joiner BL, Streibel BJ. The team hand-book. 3rd ed. Methuen: GOAL/QPC; 2018. 2. Taylor FW. The principles of scientific management. New York: NY, USA and London, UK: Harper & Brothers, LCCN 11010339, OCLC 233134. Also available from Project Gutenberg 1911. 3. Anderson JC, Rungtusanatham M, Schroeder RG. A theory of quality management underlying the Deming management method. Acad Manag Rev. 1994;19(3):472–509. 4. Best M, Neuhauser D. Walter A Shewhart, 1924, and the Hawthorne factory. Qual Saf Health Care. 2006;15(2):142–3. 5. Juran J. A history of managing for quality: the evo-lution, trends, and future directions of managing for quality. Milwaukee: ASQC Quality Press; 1995. 6. Steven S, Bowen HK, Ninca MY. Decoding the DNA of the Toyota production system. Harv Bus Rev. 1999;77(5):96–106. 7. Liker JK. The Toyota way: 14 management principles from the world’s greatest manufacturer. New York: McGraw-Hill; 2004. 330p. 8. Katzenbach JR, Smith DK, et al. The discipline of teams. Harv Bus Rev. 1993;71:111. 9. Deming WE. Improvement of quality and productiv-ity through action by management. Natl Product Rev. 1981;1(1):12–22. 10. Varvasovszky Z, Brugha R. A stakeholder analysis. Health Policy Plan. 2000;15(3):338–45. 11. Wackerbarth SB, Strawser-Srinath JR, Conigliaro JC. The human side of lean teams. Am J Med Qual. 2015;30(3):248–54. 12. Jeff Bezos’ GV. “Two pizza rule” for productive meet-ings – Business Insider. Business Insider; 2013 [cited 2019 Jun 20]. Available from https://www.businessin-sider.com/jeff-bezos-two-pizza-rule-for-productive-meetings-2013-10 13. DeBrusk C. Get things done with smaller teams [Internet]. MIT Sloan Manag Rev. 2018 [cited 2019 Jun 19]. Available from https://sloanreview.mit.edu/article/get-things-done-with-smaller-teams/ 14. Bell ST, Brown SG, Colaneri A, Outland N. Team composition and the ABCs of teamwork. Am Psychol. 2018;73(4):349–62. 15. Winsborough D, Chamorro-Premuzic T. Great teams are about personalities, not just skills. Harvard Business Review 2017;25. 16. Apfelbaum E, Mangelsdorf ME. The trou-ble with homogeneous teams [Internet]. MIT Sloan Manag Rev. 2017 [cited 2019 Jun 19]. Available from https://sloanreview.mit.edu/article/the-trouble-with-homogeneous-teams/ 17. Druskat VU, Wolff SB. Building the emotional intelli-gence of groups. Harv Bus Rev. 2001;79(3):80–90, 164. 18. IHI. Quality improvement project charter developing an improvement project charter; 2018. Available from IHI.org 19. Schwarz R. How to design an agenda for an effective meeting. Harv Bus Rev. 2015, March 19. Available from https://hbr.org/2015/03/how-to-design-an-agenda-for-an-effective-meeting 20. Rogelberg S. Why your meetings stink – and what to do about it strategies for engagement why your meetings stink – and what to do about it. Harv Bus Rev. 2019, January/February. https://hbr.org/2019/01/why-your-meetings-stink-and-what-to-do-about-it 21. Jay A. How to run a meeting. Harv Bus Rev. 1976;52(2):43–57. 22. Axtell P. 5 common complaints about meetings and what to do about them. Harv Bus Rev. 2018, June 4. https://hbr.org/2018/06/5-common-complaints-about-meetings-and-what-to-do-about-them 23. Gallo A. The seven imperatives to keeping meetings on track; 2013. https://twishort.com/4F3ec 24. Axtell P. The most productive meetings have fewer than 8 people. Harv Bus Rev [Internet]. 2018. Available from https://hbr.org/2018/06/the-most-productive-meetings-have-fewer-than-8-people?autocomplete=true 25. Conner D. Managing at the speed of change: how resilient managers succeed and prosper where others fail [Internet]; 1993 [cited 2019 Jun 20]. Available from https://books.google.com/books?hl=en&lr=&id=Iu2LDQAAQBAJ&oi=fnd&pg=PR11&dq=Managing+at+the+speed+of+change:+How+resilient+managers+succeed+and+prosper+where+others+fail&ots=2EjSqqCU-E&sig=o5OPSGOQw5ZNfT1kExjxmdYV0Ak 26. Dyar C. How to lead effective cross-functional teams [Internet]. MIT Sloan Manag Rev. 2019 [cited 2019 Jun 20]. Available from https://sloanreview.mit.edu/article/how-to-lead-effective-cross-functional-teams/ 27. Kanter MR. Ten reasons people resist change. Harv Bus Rev [Internet]. 2012. Available from http://blogs.hbr.org/2012/09/ten-reasons-people-resist-chang/ 28. Hilton K, Anderson A. IHI psychology of change framework to advance and sustain improvement; 2018. www.ihi.org 29. Ganz M. Public narrative, collective action, and power. In: Accountability through public opinion [Internet]. The World Bank; 2011 [cited 2019 Jun 27]. p. 273–89. Available from http://elibrary.worldbank.org/doi/10.1596/9780821385050_CH18 30. Hackman JR, Oldham GR. Motivation through the design of work: test of a theory. Organ Behav Hum Perform. 1976;16:250–79.

206 31. Ganz M. Organizing: people, power, change. Organ Handb. 2014;Harvard Un:15–20. 32. Wageman R, Nunes D, Burruss J, Hackman J. Senior leadership teams: what it takes to make them great [Internet]. 2008 [cited 2019 Jun 20]. Available from https://books.google.com/books?hl=en&lr=&id=jl21CgAAQBAJ&oi=fnd&pg=PR9&dq=Senior+Leadership+Teams:+What+it+Takes+to+Make+Them+Great&ots=IluqEfUXaK&sig=v-6q_rtmwunfTzWIpE29-conupc 33. Edmondson AC. Strategies for learning from failure. Harvard business review 2011;89(4):48–55. 34. Institute for Healthcare Improvement [Internet]. [cited 2018 May 16]. Available from http://www.ihi.org/resources/pages/default.aspx 35. Healthcare Quality Improvement Skills Center CME: HISC Resources – Online Healthcare CME [Internet]. [cited 2019 Jun 27]. Available from http://www.improvementskills.org/courseinfo/resources.cfm 36. Harvard Business Review – ideas and advice for lead-ers [Internet]. [cited 2019 Jun 27]. Available from https://hbr.org/ 37. Odugbemi SL, Taeku T. Accountability through public opinion; 2011 [cited 2019 Jun 27]. Available from https://elibrary.worldbank.org/doi/pdf/10.1596/978-0-8213-8505-0#page=301

207© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 R. K. Shah, S. A. Godambe (eds.), Patient Safety and Quality Improvement in Healthcare, https://doi.org/10.1007/978-3-030-55829-1_11Handoffs: Reducing Harm Through High Reliability and Inter-Professional CommunicationKheyandra D. Lewis, Stacy McConkey, and Shilpa J. Patel IntroductionOptimal continuity of care between patients and providers requires a strong foundation in com-munication. With increases in transitions of care, structured communication has become integral in the education of all healthcare providers to improve patient safety. This education has been primarily focused on provider types as distinct disciplines and specialties; however, patients intersect multiple provider types when accessing medical care, thus emphasizing the need for a collaborative inter-professional approach. Miscommunication Can Lead to Adverse EventsNo matter one’s discipline, providers must make effective communication a dedicated practice. Lapses in communication due to incomplete, inaccurate, or omitted information are leading causes of adverse events, including sentinel events [1]. Up to two-thirds of sentinel events have been linked to inadequate commu-nication, of which half were attributed to poor transitions of care between providers [2] (Key Points Box 11.1).K. D. Lewis (*) Drexel University College of Medicine, St. Christopher’s Hospital for Children, Department of Pediatrics, Philadelphia, PA, USAe-mail: [email protected] S. McConkey Adventhealth for Children, Graduate Medical Education, Department of Pediatrics, Orlando, FL, USAe-mail: [email protected] S. J. Patel University of Hawaii, John A. Burns School of Medicine, Kapi’olani Medical Center for Women & Children, Department of Pediatrics, Honolulu, HI, USAe-mail: [email protected] Objectives• Define handoffs, and review common areas where lapses in communication can lead to patient harm.• Summarize the history of handoffs in graduate medical education.• Explore standardized handoff models for written and verbal communication.• Identify strategies for implementation of effective handoffs for inter- professional teams.

208 Defining HandoffsThe vulnerable state that transitions of patient care pose can be mitigated by conducting hand-offs. Handoffs, also referred to as handovers or sign-out, involve the process of transfer and acceptance of patient care information and responsibility from one provider to another [3–5]. Providers may include a range of healthcare workers, such as physicians, nurses, and advanced practice providers (Key Points Box 11.2).Handoffs characteristically involve two roles: the sender and receiver. The sender transmits patient information and releases patient care to the receiver who receives the patient information and assumes responsibility of the patient. Ideally, this exchange of information occurs face to face. Both roles should demonstrate active listening and participation; handoffs should allow opportu-nity for discussion and clarification of informa-tion [6]. Active listening, paired with the fresh perspective of the receiver, has been shown to reduce fixation errors [5]. The overarching goal is for both the sender and receiver to develop a shared mental model, “the perception of, under-standing of, or knowledge about a situation or process that is shared among team members through communication” [7] (Fig. 11.1). Handoffs Throughout the HospitalThere are a wide variety of handoffs that occur during a hospitalization, and each one has the potential for errors and communication deficits. Handoffs occur between a variety of profession-als – between those in the same profession, such as nurse to nurse during shift change, as well as inter-professional handoffs between different professional types, such as nurse to radiology technician, and inter-unit handoffs (i.e., operat-ing room (OR) staff to intensive care unit (ICU) staff) (see Table 11.1).The wide variety of handoffs are important to recognize as the language of medicine is the same across healthcare professionals, but the communication priorities may vary between pro-vider types. For instance, what is prioritized in a nurse-to-nurse handoff (e.g., reviewing orders, wound care specifics, intravenous line flushes) varies from what is prioritized in physician-to- SenderReceiverSHAREDMENTALMODELFig. 11.1 Creation of a shared mental model between sender and receiverTable 11.1 Examples of handoffs by location and pro-vider typeTypeProviderShift changePhysician to physician or nurse to nurseTemporary coverageNursing coverage for a break or a surgical technician scrubbing out during a surgical caseAcross staffingPrimary care provider to an on-call providerAcross specialtiesAnesthesiologist to surgeonAcross settings/organizationsEmergency department to intensive care unitProvider typesHealthcare provider to caregiver at a nursing home facilityKey Points Box 11.1Miscommunications can occur when the information delivered is as follows [1]: 1. Inaccurate 2. Incomplete 3. Not timely 4. Misinterpreted 5. Not requiredKey Points Box 11.2Handoffs involve the process of transfer and acceptance of patient care information and responsibility from one provider to another [3–5].

209physician handoffs (e.g., information about diag-noses and specific treatment goals/plans). As expected, there is variation in the information provided based on provider workflow and responsibilities.Walking through a hospitalization for a patient may help demonstrate possible points of commu-nication vulnerability.For Jessica, the patient in the case, the first handoff occurs between the physicians and nurses in the emergency department and the inpatient physician and nursing staff. At this juncture, it is critical to communicate where the patient is in the course of their management, e.g., when/which medication doses were last given, what testing has been completed, and what treatments are outstanding or need to be followed up. Ideally, all data (e.g., lab results) are available to all pro-viders via the electronic health record (EHR). However, delays in charting due to competing priorities may result in the lack of a shared men-tal model, thereby leading to duplication of thera-pies or delays in care.Once a patient is admitted, and the results of the testing, therapeutic interventions, and the physician and nurse assessments are complete, a plan of care is created by the inpatient unit team. In many hospitals, nurses and physicians who care for hospitalized patients provide care during scheduled shifts. Thus, nurses and physicians handoff patient care to the providers on the next shift (e.g., the day shift hands off to the night shift). Notably, nursing handoffs have evolved over time to include bedside handoffs which incorporate the patient and or family [8]. This practice provides a patient-centered approach to care by incorporating the input of the patients and their families and ensuring they are aware of and in agreement with treatment plans and proce-dures. As mentioned previously, nursing hand-offs have a different focus than physician handoffs, and both are equally important to patient care. Ideally, a shared mental model is achieved with agreement between provider types for the severity of illness, plan of care, action items, and contingency plans for the next shift.The next handoffs for Jessica are between the inpatient medical team and the radiology team. Nursing must provide adequate information regarding the patient’s condition to alert the radi-ology staff to possible issues with procedures: primary medical problem (e.g., pneumonia), per-Vignette 11.1Jessica is a 15-year-old female, with a his-tory of asthma, who was in her usual state of health until she developed fever, worsen-ing cough, and increased work of breathing in the setting of a 1-week history of cough and runny nose. She has had poor oral intake and a physical exam that is notable for crackles over the left lower lung field. She has a fever and an oxygen saturation of 100%. Jessica is admitted for community- acquired pneumonia with concern for dehydration. She is started on empiric intravenous (IV) antibiotics.Vignette 11.3Due to Jessica’s worsening condition, the decision is made to obtain a chest X-ray in the radiology suite.Vignette 11.2Jessica is admitted to the inpatient unit, where she is examined by the nursing and physician teams. She is given IV fluids with continuation of the antibiotics that were started in the emergency department. Over the course of her first day of hospital-ization, she develops increased work of breathing, pain in her chest, and a slow drop in oxygen saturation to 90%. She is started on supplemental oxygen by nasal cannula.

210tinent medical history (e.g., asthma), allergies, oxygen requirements, and sedation needs. It is critical that the team that is accepting the patient, no matter how briefly, be aware of the critical needs of the patient: Do they require oxygen? Do they have allergies to contrast?If a surgical procedure is required during the hospitalization, several other handoffs must occur: First is a communication of the patient’s history and diagnoses to the consultants – the surgeon as well as the anesthesiologists. Both consultants must be aware of issues that are spe-cific to the patient (e.g., history of asthma), including information that will likely be present in the medical record such as diagnostic results, but again could be overlooked or not yet present in the medical record in urgent situations. For transitions of care such as these, it is most help-ful to communicate a cogent patient summary, highlight the current diagnoses and pertinent past medical history, and discuss any intra-pro-cedure needs (e.g., obtaining a specimen culture of the pleural fluid).Following the procedure, the events, findings, information regarding intraoperative medica-tions, IV fluids, chest tube drains, and plans must be communicated from the surgeons to the physi-cians who are caring for the patient after surgery. Similarly, the postanesthesia unit nurses must transfer information to the receiving intensive care unit nurses. Ideally, for areas such as the intensive care unit, this communication is face to face and with all members of the inter- professional team [9] (Key Points Box 11.3).The transition between units is often typically between clinicians of the same type, e.g., physi-cian to physician and nurse to nurse. The same procedures and protocols that apply to change of shift handoffs should be utilized with this transi-tion as well, such as time of next scheduled medi-cation dosing and wound care instructions for the chest tube site (Key Points Box 11.4).Vignette 11.4The chest X-ray reveals the previously noted pneumonia but now with a new para-pneumonic effusion (infected fluid in the lung). Jessica continues to have increased work of breathing, fevers, and oxygen desaturation, requiring increased respira-tory support. The decision is made to surgi-cally drain the fluid collection in her lung.Vignette 11.6Jessica makes a rapid recovery. On hospital day 4, the chest tube is discontinued, and she is transferred from the intensive care unit back to the general inpatient unit.Key Points Box 11.3The surgical team must communicate with the medical team how they would like the chest tube managed including parameters for removal. Delays in communication can result in adverse events and unnecessary utilization of resources, resulting in increased costs.Vignette 11.5Jessica has a left-sided chest tube placed without complication. Postoperatively, she is transferred to the intensive care unit for further management.Key Points Box 11.4Handoffs from physician to physician should be timely and utilize the same framework and information that is required for inter-shift handoffs. If a patient who is transferred from the intensive care unit to the inpatient unit clinically deteriorates, the general inpatient team needs to be aware of the intensive care unit course and current clinical needs.

211Following the resolution of the illness that caused the admission, the patient must be readied for discharge. The final handoffs to occur in a hos-pitalization are from the hospital to the discharge location – whether that be home (where the receiver of the handoff is the primary care pro-vider), a rehabilitation center, nursing home, or another hospital. Each of these locations requires a different type of communication, which must be timely, concise, and accurate. Delays in commu-nication with primary care providers can result in the lack of appropriate follow- up or medical man-agement, resulting in readmission. When hand-offs are suboptimal, delays in treatment and adverse events can occur. Additionally, poor handoffs can impact patient and provider satisfac-tion, cause prolonged hospital stays, and contrib-ute to increased cost of care [9]. Evolution of HandoffsDespite the seemingly straightforward approach, high-quality handoffs are a complex process, and prior to the 2006 Joint Commission National Patient Safety Goal, handoffs were not formally taught nor required [1]. The complexity of hand-offs is further elevated by the frequency of pro-vider transitions, particularly in institutions that have medical trainees. Beginning in 2003, the Accreditation Council for Graduate Medical Education (ACGME) mandated a reduction in the length of time trainees could provide contin-uous care in the hospital, resulting in a substan-tial increase in change-of-shift transitions and ultimately the number of handoffs [10, 11]. Additionally, nurses are also vulnerable to pro-longed shift durations. A study by Scott and col-leagues showed that on average nurses work between 8 and 12 hours, and the risk of error nearly doubled when they worked 12.5 hours or more [12]. Transitions of Care and Graduate Medical EducationA large proportion of the most complex and ill patients in our healthcare system are cared for in institutions where physicians and other members of the healthcare workforce receive clinical train-ing. This has a significant effect on the quality of handoffs in three ways. Increased Frequency of HandoffsSince the advent of the ACGME’s work hour restrictions in 2003, hospital systems that utilize Graduate Medical Education (GME) trainees as their frontline workforce have been required to adapt their schedules. Restrictions in the number of hours worked per week and a decrease in the duration of shifts created condensed schedules with increased numbers of providers caring for each patient. Shifts that were previously 24 hours in length changed to 12-hour shifts, resulting in a dramatic surge in the number of handoffs between physicians. A typical teaching hospital might have up to 4000 individual patient handoffs occurring in the course of a day [1]. Lack of Standardized Training in HandoffsMany physicians functioning as the faculty for graduate medical education trainees were not for-mally trained in the process of giving handoffs. Therefore, they lack knowledge of the standard-ized training curriculum for handoffs used in GME. Furthermore, faculty may lack compe-tency in assessment and coaching of residents and fellows during observations of handoffs. While changing, many medical and nursing schools have not yet adopted a standardized cur-riculum and training of their students in provid-Vignette 11.7Jessica no longer requires supplemental oxygen and is able to take all medications by mouth. She is instructed to complete the course of antibiotics and to follow up with her primary care physician in 2 days.

212ing effective handoffs, resulting in the burden of training falling on hospital systems that sponsor residency training [13]. Variable Clinical Experience of ProvidersWithout proper supervision, inexperienced phy-sicians may not be capable of recognizing impor-tant clinical findings that could result in a delay in care or improper care of their patients. The hospital system is especially vulnerable at the time surrounding change of shift, which is now happening more frequently due to duty hour restrictions/shortened shift lengths. Written and verbal handoffs may be disorganized, unpriori-tized, or fail to paint an accurate picture of the patient’s condition and needs. This potentially results in delays in care due to prolonged hand-offs or “receiver fatigue” where critical clinical information is lost in a presentation filled with extraneous information.In an effort to continuously improve the pro-cess of training future physicians and address the persistent lack of improvement in the quality and safety gaps surrounding patient care, the ACGME devised the Next Accreditation System (NAS) to include evaluation of the Clinical Learning Environment (CLER) of the hospital systems that are training residents and fellows. They identified six areas of focus during annual CLER visits, which include (1) identification and intervention in patient safety issues, (2) quality and perfor-mance improvement efforts, (3) supervision of trainees, (4) professionalism, (5) management of burnout and fatigue (resident wellness), and (6) transitions in care [14]. Many of the tenets of pro-viding quality handoffs have been incorporated into the CLER evaluation process. Areas that are evaluated include the training of residents, fel-lows and faculty in a common clinical site-based process for handoffs, knowledge of transition of care policies among all physicians, presence of efforts to assess and continually improve hand-offs, and the participation of faculty, residents, fellows, inter-professional teams, and families in the handoffs process. Strategies for Effective HandoffsAvoiding communication failures during hand-offs can be lessened by standardizing the content communicated between the sender and receiver in both verbal and written formats [1]. Handoffs can improve communication if the information is consistent and delivered in a predictable format. Critical information, such as illness severity, code status, vital signs, allergies, medications, pertinent events leading up to illness or hospital-ization, ongoing assessment, pertinent diagnostic test results, plan of care with action items, and contingency plans, should be included in hand-offs [1] (Key Points Box 11.5).Verbal handoffs should be timely, conducted face to face, and occur in a location that is free of excess noise and distraction. Written handoff tools complement the verbal handoff communi-cation and facilitate opportunities for detail and for clarification when information is disparate. When possible, handoffs should include all mem-bers of the inter-professional team, which can promote ongoing discussion and ensure the maintenance of a shared mental model [15] (Fig. 11.2).Key Points Box 11.5Handoffs should include (1) illness sever-ity, (2) code status, (3) vital signs, (4) aller-gies, (5) medications, (6) pertinent events leading up to illness/hospitalization, (7) ongoing assessment, (8) pertinent diagnos-tic test results, (9) action items, and (10) contingency plans [1].

213 Handoff ModelsThere are several handoff mnemonic models used to structure both written and verbal handoffs (see Table 11.2 for examples). One commonly used mnemonic is SBAR, which stands for Situation, Background, Assessment, and Recommendation. SBAR is a handoff communi-cation tool developed by the US Navy and has since been adopted in many healthcare settings and is most frequently used by nurses. SBAR was designed to communicate urgent patient informa-tion in a relatively quick manner [2].An extensively studied model, the I-PASS Handoff program, uses the organizational framework of a mnemonic as an anchor for an interventional bundle that includes strategies for team communication. The I-PASS Handoff pro-gram includes seven core elements: (1) the I-PASS mnemonic (I = illness severity, P = patient summary, A = action list, S = situa-tion awareness and contingency planning, and S = synthesis by receiver), (2) a workshop for teaching team communication through the use of TeamSTEPPs and handoff techniques, (3) skills training through simulation and role-play-ing exercises, (4) independent study module, (5) faculty development, (6) a direct observation tool for feedback, and (7) campaign for adop-tion and sustaining practice. This study demon-strated a reduction in preventable adverse events by 30% and medical errors by 23% when imple-mented by nine children’s hospital [17]. The I-PASS Handoff program has since been adopted by more than 50 hospitals, studied in many iter-ations and is frequently referred to as the gold standard for effective handoff communication between physicians [1, 12]. It has also been adapted effectively for nursing shift report across varied clinical settings [18] (see Table 11.3) (see Fig. 11.3).The right incation is:Protective of sensitive patient related informations as perthe Health Insurance Portability and Accountability Act(HIPAA)The right information is:Up to date and includes: current patient condition,treatments, concerns, and anticipated changes for the nextshiftTranscribed using EHR tools to avoid errorsOrganized in a standardized formatThe right style of communications is:Face to faceProvided in both verbal and written formFocused on the creation of a shared mental modelConducive to questions and opportunities for clarificationThe right people include:All members of the inter-professional team involved in thepatient’s care; including patient and familyFaculty educators at training programs who can performworkplace assessments and provide feedback on handoffqualityQuiet, free of distractions and nonemergent interruptionsConsistent, same time, same placeFig. 11.2 The “rights” of effective handoffs [1]Table 11.2 Handoff mnemonics examples [1, 16]ISBARIdentifySituationBackgroundAssessmentRecommendationsSIGNOUTSick/do not resuscitateIdentifying dataGeneral hospital courseNew events of the dayOverall health statusUpcoming possibilities/planTasks to completeHANDOFFSHospital locationAllergies/adverse reactionsNameDo not attempt to resuscitateOngoing medical problemsFacts about hospitalizationFollow-upScenariosPSYCHPatient information/backgroundSituation leading to hospital courseYour assessmentClinical informationHindrance to discharge

214 Implementing and Sustaining Effective Handoff ProgramsImplementation of a standardized handoff system requires support from hospital administrative leadership for the rollout of the program, educa-tional efforts, time needed to train staff, and, most importantly, resources needed to provide workplace-based assessments with feedback to achieve the desired behavior change. Dedicated time and physical space for handoffs are also important. New space may be needed to have a quiet, uninterrupted discussion, and it may be necessary to have extra staff to cover patient needs while handoffs are occurring.Information such as accurate weights, aller-gies, code status, and location are critical to all handoffs. Integration of the EHR for a printed/written handoff tool that supplements/supports verbal communication allows for seamless infor-mation transfer and avoidance of transcription errors. Nursing and physician handoff written documentation differ in various aspects of focus, with physician handoffs being more related to contingency plans and action items, while nurs-ing handoffs focus on different action items that include when medications are to be given, fluid intake and output, activities of daily living, and pain management (pain scales and PCA (patient- controlled analgesia) orders). There is common ground between nursing and physician written handoff elements. A Continuity of Care Document is a potential EHR-based framework to support the supplemental written document for use in handoffs by multiple provider types [19].Once a handoff system is established in a hos-pital, it will need ongoing evaluation and support using a continuous improvement approach to keep the process at the front of the clinicians’ minds and prevent attrition of adherence to stan-dard handoff protocol. Sustaining quality improvement efforts is difficult in any setting and often more challenging in the healthcare setting. Fryman and colleagues outlined the use of qual-ity improvement cycles involving direct observa-tion of handoffs, and audits of the use of the Fig. 11.3 Example of written/printed handoff tool using the I-PASS modelTable 11.3 The I-PASS handoff model with overview of elements [17]IIllness severityAlerts the receiver to the patient clinical status: stable, watcher, or unstablePPatient summaryProvides an overview of the patient’s pertinent past medical history, events leading up to hospitalization, and interim hospital courseAAction listTasks that require completion for the next shiftSSituation awareness and contingency planningPreparatory considerations for a change in clinical status. Should be relayed in “if, then” statementsSSynthesis by receiverOpportunity for clarification and inquiry to ensure shared mental model between sender and receiver

215written handoff tools from the EHR demonstrated success in sustaining change in their system [20]. Future Direction of  Inter- Professional HandoffsThe Joint Commission and the CLER standards specifically mention inter-professional handoffs as a future standard. Calling for the development of interdisciplinary handoffs between obstetrics and neonatology physicians, Vanderbilt and col-leagues described how a common handoff would greatly benefit the neonate and mother’s health and outcomes [3]. There has not been a signifi-cant amount of research on this topic in the litera-ture, but Kostoff and colleagues showed improvement in pharmacy students’ self- perception of inter-professional competence by using the SBAR format for communication between pharmacists and other disciplines [2]. Similarly, Solan and colleagues demonstrated that multidisciplinary handoffs involving resi-dents and charge nurses improved perceptions of communication [21]. The gold standard for com-munication at transitions of care would include training entire hospital systems to perform hand-offs in a standardized, highly reliable fashion, with sharing of information via an EHR- generated handoff document to support the ver-bal communication. SummaryHandoffs provide a unique opportunity to enhance inter-professional communication. Given the increased frequency in which patients intersect different healthcare providers across disciplines, settings, and organizations, deliber-ate education in structured communication is essential to patient safety. There are several stan-dardized frameworks, such as I-PASS, that can be used to structure both written and verbal hand-offs. No matter what framework is used, it is important to maintain anchoring elements such as illness severity and contingency plans and ensure time for clarification and synthesis between sender and receiver. When handoffs include all members of the inter-professional, clinical teams are more effective in developing shared mental models regarding their patients, and improvement in patient safety follows.Editors’ CommentsIt is well accepted in the healthcare safety and quality realms that communication breakdowns are a key contributor to adverse outcomes and harm. A component of robust communication is handoffs. Care transi-tions are a vulnerable period for patients as they move within systems and through dif-ferent types of care delivery models. During these transitions in care, the reli-ance on accurate, timely, relevant handoffs cannot be overstated.In this chapter, the authors approach the topic by demonstrating the burning plat-form: communication is crucial, and hand-offs can save lives. They present a nice history of handoffs in medicine from the perspective of graduate medical education; this historical perspective has relevance as the lessons and strategies can be extrapo-lated to other care settings as well.The authors move the concept of hand-offs further by demonstrating key compo-nents of handoffs and what is considered to be crucial information. The mnemonics as well as the evidence-based materials help frame the value-add of handoffs within the context of care delivery. Once the improve-ment scientist knows the parts of the hand-off which are important, then they can decide on the best tool to implement for handoffs.There are handoffs that can be as simple as a department specific tool that is main-tained by house staff to proprietary tools described by the authors that can be imple-mented and scaled within an organization.The end of the chapter considers the next iteration of handoffs and weaves in the importance of inter-professional teams.

216 Chapter Review Questions 1. What are handoffs and what is the intended goal?Answer: Handoffs involve the process of transfer and acceptance of patient care infor-mation and responsibility from one provider (sender) to another (receiver) with the goal of creating a shared mental model. 2. 2: True or false – communication failures can be decreased by standardization of handoffs.Answer: True. Both written and verbal handoffs can improve communication if infor-mation is presented in a consistent and pre-dictable format. 3. What type of information should be included in handoffs?Answer: Critical information, such as ill-ness severity, code status, vital signs, aller-gies, medications, events leading up to illness or hospitalization, ongoing assessment, perti-nent diagnostic test results, plan of care with action items, and contingency plans, should be included in handoffs [1]. 4. List the four “Rights of Effective Handoffs”.Answer: (1) The right location (quiet, con-sistent time/place). (2) The right people (members of inter-professional team and fac-ulty educators that can assess handoff). (3) The right information (organized in standard format, up-to-date using EHR tools for tran-scription). (4) The right style of communica-tion (face to face, includes both verbal and written handoffs, allows for questions/clarifi-cations, and creates a shared mental model). 5. Define the I-PASS mnemonic elements. What advantages have been demonstrated from the I-PASS Handoff program?Answer: I-PASS stands for illness severity, patient summary, action list, situation aware-ness and contingency planning, and synthesis by receiver. The program is a standardized bundle that provides strategies to enhance team communication. Implementation of the program at nine children’s hospital was shown to decrease preventable errors by 30% and medical errors by 23%.References 1. Inadequate hand-off communication [Internet]. Jointcommission.org. 2017 [cited 29 March 2019]. Available from https://www.jointcommission.org/sentinel_event_alert_58_inadequate_handoff_com-munication/ 2. Kostoff M, Bukhardt C, Winter A, Shrader S. An interprofessional simulation using the SBAR com-munication tool. Am J Pharm Educ. 2016;80(9):157. https://doi.org/10.5688/ajpe809157. 3. Vanderbilt A, Pappada S, Harper D, Papadimos T. Increasing patient safety with neonates via handoff communication during delivery: a call for interpro-fessional health care team training across GME and CME. Adv Med Educ Pract. 2017;8:365–7. https://doi.org/10.2147/AMEP.S129674. eCollection 2017. 4. Davis J, Roach C, Elliot C, Mardis M, Justice E, Riesenberg L. Feedback and assessment tools for handoffs: a systematic review. J Grad Med Educ. 2017;9(1):18–32. https://doi.org/10.4300/JGME-D-16-00168.1. 5. Patterson E, Roth E, Woods D, Chow R, Gomes J. Handoff strategies in settings with high conse-quences for failure: lessons for health care operations. Int J Qual Health Care. 2004;16(2):125–32. 6. Catalano K. Hand-off communication does affect patient safety. Plast Surg Nurs. 2009;29(4):266–70. https://doi.org/10.1097/PSN.0b013e3181c20136. 7. TeamSTEPPS Fundamentals Course: Module 5. Situation monitoring [Internet]. Agency for HealthCare Research and Quality. 2012 [cited 22 April 2019]. Available from https://www.ahrq.gov/teamstepps/instructor/fundamentals/module5/slsit-monitor.html 8. Riesenberg L, Leitzsch J, Cunningham J. Nursing handoffs: a systematic review of the literature. Am J Nurs. 2010;110(4):24–34; quiz 35–6. https://doi.org/10.1097/01.NAJ.0000370154.79857.09.The editor’s organizations have embraced the concept of inter-disciplinary teams, and the initial results demonstrate transforma-tional capacity. The role of handoffs will be ever more important when inter- professional teams become the norm. This chapter presents a stepwise approach to understanding the intrinsic value of hand-offs all the way to describing techniques to implement handoff programs within your healthcare system.

217 9. Patterson E, Wears R. Patient handoffs: standardized and reliable measurement tools remain elusive. Jt Comm J Qual Patient Saf. 2010;36(2):52–61. 10. Accreditation Council for Graduate Medical Education. ACGME Common Program Requirements. https://www.acgme.org/Portals/0/PFAssets/ProgramRequirements/CPRs_2017-07-01.pdf. Accessed June 17, 2018. 11. DeRienzo C, Karen F, Barfield M, Gopwani P. Handoffs in the era of duty hours reform: a focused review and strategy to address changes in the accreditation Council for Graduate Medical Education Common Programs Requirements. Acad Med. 2012;87(4):403–10. https://doi.org/10.1097/ACM.0b013e318248e5c2. 12. Handoffs and signouts [Internet]. Patient Safety Primer. 2019 [cited 17 April 2019]. Available from https://psnet.ahrq.gov/primers/primer/9/handoffs-and-signouts 13. Shahian D, McEachern K, Rossi L, Chisari R, Mort E. Large-scale implementation of the I-PASS hando-ver system at an academic medical centre. BMJ Qual Saf. 2017;26(9):760–70. https://doi.org/10.1136/bmjqs-2016-006195. Epub 2017 Mar 9. 14. CLER pathways to excellence: expectations for an optimal clinical learning environment to achieve safe and high quality patient care [Internet]. Accreditation Council for Graduate Medical Education; 2014 [cited 20 April 2019]. Available from http://www.acgme.org/portals/0/PDFs/CLER/CLER_Brochure.pdf 15. Mamykina L, Jiang S, Collins S, Twohig B, Hirsh J, Hripcsak G. Revealing structures in narratives: a mixed-methods approach to studying inter-disciplinary handoff in critical care. J Biomed Inform. 2016;62:117–24. https://doi.org/10.1016/j.jbi.2016.03.025. Epub 2016 Apr 7. 16. Brownstein A, Schleyer A. The art of handoffs: a mnemonic for teaching the safe transfer of criti-cal patient information. Resid Staff Physician [Internet]. 2007 [cited 30 April 2019];53(6). Available from https://www.mdmag.com/journals/resident-and-staff/2007/2007-06/2007-06_02 17. Starmer A, Spector N, Srivastava R, West D, Rosenbluth G, et al. Changes in medical errors after implementation of a handoff program. N Engl J Med. 2014;371:1803–12. https://doi.org/10.1056/NEJMsa1405556. 18. Patel SJ, Durbin M, Lai B, Theel C, Methered T, Sataraka C, et al. Improving nurse communication using the I-PASS handoff bundle. Abstract presented at Institute for Healthcare Improvement National Forum; 2018. 19. Collins S, Stein D, Vawdrey D, Stetson P, Bakken S. Content overlap in nurse and physician handoff artifacts and the potential role of electronic health records: a systematic review. J Biomed Inform. 2011;44(4):704–12. https://doi.org/10.1016/j.jbi.2011.01.013. 20. Fryman C, Hamo C, Raghavan S, Goolsarran N. A quality improvement approach to standardization and sustainability of the hand-off process. BMJ Qual Improv Rep. 2017;6(1). pii: u222156.w8291. https://doi.org/10.1136/bmjquality.u222156.w8291. eCol-lection 2017. 21. Solan L, Yau C, Sucharew H, O’Toole J. Multidisciplinary handoffs improve perceptions of communication. Hosp Pediatr. 2014;4(5):311–5. https://doi.org/10.1542/hpeds.2014-0005.11 Handoffs: Reducing Harm Through High Reliability and Inter-Professional Communication

219© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 R. K. Shah, S. A. Godambe (eds.), Patient Safety and Quality Improvement in Healthcare, https://doi.org/10.1007/978-3-030-55829-1_12Safety II: A Novel Approach to Reducing HarmThomas Bartman, Jenna Merandi, Tensing Maa, Tara C. Cosgrove, and Richard J. Brilli Opening Question/ProblemThe most common approach to improving safety in all industries, and especially in healthcare, is learning from errors and harm. This “find and fix” approach is termed “Safety I.” After an untoward event (or sometimes a near miss) occurs, a subsequent analysis is performed to identify where individuals and/or systems failed, with steps outlined to prevent event recurrence [1–3]. While the Safety I approach has led to dra-matic safety improvements, Safety I has multiple shortcomings [4]. First, neither learning nor improvement happens until after an undesired event. Second, as individuals and systems improve to prevent recurrent errors, remaining errors/failures become “one-offs,” each unusual and unique such that learning from prior events is uninformative. Third, focusing on what went wrong leads to more rules and regulations, trend-ing toward rigid systems which cannot respond to the unexpected (assuming people follow the “rules”). Finally, since in every industry humans complete or supervise most activities, focusing on human error with the necessary enforcement of performance expectations can demoralize staff, thus potentially limiting a valuable resource – the human mind – from contributing to error reduction.Another common approach to improving safety is the failure modes and effects analysis T. Bartman (*) Nationwide Children’s Hospital, Division of Neonatology, Columbus, OH, USAe-mail: [email protected] J. Merandi Nationwide Children’s Hospital, Pharmacy Department, Columbus, OH, USAe-mail: [email protected] T. Maa · R. J. Brilli Nationwide Children’s Hospital, Division of Pediatric Critical Care Medicine, Columbus, OH, USAe-mail: [email protected]; [email protected] T. C. Cosgrove Nationwide Children’s Hospital, The Heart Center, Columbus, OH, USAe-mail: [email protected] Objectives• Highlight the differences between Safety I and Safety II approaches• Understand how Recognize, Respond, and Learn function as Safety II pillars• Understand how individual factors, rela-tionship and interactions, structural and environmental factors, and innovative approaches impact Safety II practice in a healthcare microenvironment

220Key Points Box 12.1Monitor: Being able to see what is happen-ing in a situation. Requires valid informa-tion about the conditions and presentation of that information to those who can exe-cute further Safety II steps. The level of detail and timescale of monitoring may depend on the situation and the role of the individual monitoring.Anticipate: Being able to use informa-tion about the situation to develop expecta-tions about what might happen next. May include assigning probabilities to different events.Respond: Being able to take action to prepare for anticipated future events or change the course of events. Because pro-tocol/policy determines actions under nor-mal circumstances, responding in Safety II fashion typically involves deliberate “devi-ation” from protocol or real-time innovation.Learn: Being able to learn from moni-toring/anticipating/responding events. May involve both learning about how to handle the identical situation in the future, but more importantly learning about how to improve monitoring, anticipating, and responding.Vignette 12.1While driving through a neighborhood, you see a soccer ball roll into the street just ahead. Within a split second, you anticipate a child might soon dash into the street, so you take an unusual action by putting your foot on the brake and slowing down or maybe even stopping. Soon after you stop, a child dashes out into the street to retrieve the ball. You wipe your forehead in relief at (FMEA). FMEA involves proactively identifying potential problems and then quantifying their likelihood of occurrence, the odds the problem will escape detection, and the severity of harm the event might cause [5]. A scoring system prompts system/protocol redesign to minimize the threat from potential events which are highest risk, highest likelihood, and most likely to escape detection [6]. However, FMEA has limited value in error prevention because it is usually narrow in scope, does not address all potential errors, and usually primarily focuses on problems predict-able well in advance.We believe the way forward is a new approach – Safety II [7]. Hollnagel, who initially developed the Safety II concept, describes a model for Safety II with four components: actions which he refers to as “potentials” [8]. These actions are Monitor, Anticipate, Respond, and Learn. (Key Points Box 12.1)These four components/actions are inter- dependent: they often occur in parallel, and improving capabilities in one step can improve the ability to successfully perform the others. We believe that, especially in bedside clinical care, monitoring and anticipating are so tightly linked, they constitute one action we term “recognizing,” followed by responding (e.g., taking action, or deliberately deciding no action is necessary), and learning how to improve our ability to recognize and respond.We believe adding Safety II to current harm prevention strategies will lead to improved out-comes for the following reasons: (1) responses to all possible scenarios cannot be put into protocol because of the complexity of healthcare systems; therefore, flexibility and resilience will always be needed to cope with unanticipated conditions; (2) mindfulness, situational awareness, and clinical judgment add the power of human intelligence to rote following of expected procedures; and (3) allowing people to find “work-arounds” or alter-native ways to perform their normal tasks can sometimes improve efficiency and safety simultaneously.The following scenario illustrates Safety II in a nonclinical situation:

221In this case, you were able to monitor because you were looking out the windshield and not tex-ting (good situational awareness). Seeing the soc-cer ball led to anticipating – having a strong suspicion a child was about to run in front of you (probably informed by past learned experience). Responding entailed slowing down and/or stop-ping the car suddenly – perhaps in a manner typi-cally considered dangerous and against “preferred” behavior. Finally, you learned how to better monitor (eliminate a texting habit) and anticipate (expect children playing on this street or even while driving in general). Illustrated here, no accident or error occurred, and in fact, an acci-dent was likely avoided. Additionally, important learning occurred, reducing the accident risk in the future.One other core feature in Safety II is resil-ience. Elements of resilience are foresight (pre-dicting something untoward will happen), coping (preventing something untoward from becoming worse), and recovery (ability to return to normal functioning once something untoward occurs) [9]. While related to individ-ual psychological resilience, system resilience involves the ability of the system or individual to perform under varying conditions, e.g., responding appropriately to both negative and positive conditions [10]. Resilience is central to how error is avoided and success obtained. Safety II considers the human component of systems as necessary to maximize flexibility and resilience [7, 11], whereas Safety I sees human variation as a liability requiring design out of the system.the near catastrophe avoided. Further, you make note children are playing in a particu-lar driveway on this street and decide to drive more slowly when coming down this street in the future. You also note that some-times you text and drive and are grateful you were not texting in this situation.Vignette 12.2CJ arrived in the Emergency Department (ED) with persistent fevers, headaches, sore throat, and emesis. After sending appropriate studies, empiric antibiotics were started for presumptive meningitis. CJ was admitted to the pediatric intensive care unit (PICU) for further management due to altered mental status.In the PICU, she developed septic shock. A new murmur led to diagnosis of native valve Methicillin-Susceptible S. aureus (MSSA) endocarditis. Following a complicated hospital course, she was even-tually ready for discharge with plans for a continuous nafcillin home infusion. When the physician began to place the order in the electronic medical record (EMR) for nafcillin administered as a continuous infu-sion, the option was not available.Due to her system knowledge, the clini-cal pharmacist recognized multiple risks in this unusual situation. She worked with informatics to immediately build an order in the EMR. Because this is an unusual dos-ing method for inpatients, she notified the verifying pharmacist that he would receive an order to verify an outpatient continuous nafcillin infusion. “Anticipating” that inpa-tient pharmacists might not have the knowl-edge or experience in preparing the infusion, she contacted the IV room pharmacist to discuss medication preparation details; for example, the medication must be drawn up using an exact normal saline volume and placed in a specific bag, and because nafcil-lin is stable at room temperature for only 24 hours, the bag needed refrigeration prior to administration. Finally, she sent a com-munication to all pharmacy staff to ensure their awareness of this variation in standard practice and that future orders might include continuous nafcillin infusions.

222 Recognize (Monitor and Anticipate) Improvement StrategyThe first step in using the Safety II approach in a healthcare setting is recognizing what might hap-pen next. Recognition combines observing (mon-itoring) for signals and using that data to anticipate.If individuals cannot monitor and interpret their surroundings, everything becomes a sur-prise. An individual’s role determines the moni-toring breadth, depth, and timescale. Microsystem managers may monitor their particular unit over hours, days, or weeks, while bedside providers monitor a patient moment-by-moment.Successful monitoring depends on multiple factors. Is data available describing the situation or environment? Is that data available to the individual(s) responsible for the monitoring and in a timely manner? Does the individual respon-sible for monitoring have sufficient skill/experi-ence to interpret the data presented? Finally, is the monitoring individual alert, non-distracted, and able to focus on the situation (mindful and situationally aware)?We performed qualitative research to identify individual or system characteristics that may contribute to Safety II application in our PICU [12]. The study identified 19 themes, grouped into 4 domains, which appear to improve recog-nizing (monitoring and anticipating), responding, and learning (Fig. 12.1). Characteristics (themes) that improve an individual’s ability to monitor more effectively include an aptitude to pay atten-tion to detail (focus) and to assume a more global perspective (thinking beyond one’s role and to be more sensitive to signals). The ability to monitor individual patients, as well as the overall state of the unit, is also affected by structural and envi-ronmental factors, including familiarity with and proximity to coworkers; patient number, acuity, and intensity; and shift resource availability. Thus, monitoring may improve by eliminating non-value-added tasks to decrease distraction, streamlining mundane tasks, and introducing moments during the day dedicated to performing monitoring and anticipating. Finally, providers must be alert to their mental state and the thoughts in their head (mindfulness) and the environment/situation around them (situational awareness) [13]. In healthcare, mindfulness and situational awareness are characterized by actively observ-ing oneself, the patient, and the problem [14] and then being able to convert the flood of data around us into useful and actionable information.Tightly interwoven with monitoring are fore-sight and anticipating dangers. How do individuals anticipate the future, and how do they attach vari-ous probability levels to possible future outcomes? One possibility is that individuals with experience recall previous situations and apply heuristics such as recognizing “I’ve seen this before, and I remem-ber what happened next.” Another possibility is that individuals know what aspects of their current observations do not reflect a prior experience (because no two situations are identical), leading them to a “sixth sense” which causes them to go into higher alert or prompt further investigation and inquiry. Supporting these hypotheses, the PICU providers in our qualitative research [12] observed that colleagues proficient at anticipating have more experience and expertise. The providers The clinical situation described – “need-ing to order, dispense, and deliver a medi-cation in a novel way” – required recognizing the situation and anticipating its risks and responding. The outcome was good and no adverse medication-related event occurred. The final step was ensuring individuals and the institution learned from this unique situation and thus increased the odds of success in both similar and dissimi-lar future situations. After this case was over, debriefings were performed to iden-tify how to avoid this problem or closely related problems in the future and to iden-tify what allowed this pharmacist to recog-nize a developing problem and respond in the way she did.

223with substantial experience drew on their memo-ries to remember similar situations and were more reliably able to predict the future.For people with less experience, we can accel-erate experience acquisition through simulation, especially if used to practice uncommon or unfa-miliar situations. Using simulation may also effectively teach responding skills (discussed later). Strategies for improving a system’s antici-patory ability may include ensuring availability of individuals with anticipation skills and having the microsystem deliberately take moments to pause and anticipate/predict the future (e.g., dur-ing handovers). Anticipated Results of the ImprovementTaking steps to reduce distractions and improve mindfulness/situational awareness enables indi-viduals and organizations to more effectively monitor patients, units, and the organization as a whole [15, 16]. Focusing on mindfulness, being present, and taking time to anticipate that things may not go as expected will impact an individu-al’s ability to see the “accident waiting to hap-pen.” This is in contrast to working in autopilot mode and being forced to respond/recover more often than desired, i.e., being reactive instead of proactive. Ideally, improvements in monitoring will allow information to flow to individuals, keeping in mind that too little information will miss important signals while information over-load will increase noise. The shortest possible time lag between data acquisition and its presen-tation to decision-makers gives those individuals more time to anticipate an event before it occurs and to initiate a response. Later, when learning occurs, the ability to identify leading indicators (data which accurately predicts the future) will improve.Fig. 12.1 Interrelationship between the 4 Safety II components and 19 themes from qualitative research [12]DriverIndividualCharacteristicsRelationships andInteractionsStructural andEnvironmental FactorsInnovation ApproachesMonitoringAnticipatingRespondingLearning• Attention to Detail• Taking a GlobalPerspective• Taking a GlobalPerspective• Experience andExpertise• Familiarity andProximity• Number, Acuity, andIntensity of Patients• Shift ResourceAvailability • Taking Control• Staying Calm andMaintaining Focus• Experience andExpertise• Appreciating theConsequences of Mistakes• Personal Relationships• Teamwork• Culture of Questioning• Communication• Training to IntroduceCultural Values• Relying on Teamworkif Something Novel isConsidered• Teams Responding toChallengingCircumstances• Skepticism• Bringing AtypicalApproaches from otherMicroenvironments• Careful Examinationand Feedback afterErrors are Made

224 How the Improvement Worked in Context of the CaseIn our case, the pharmacist recognized potential problems which could put the patient at risk. More specifically, her recognition (combining monitoring and anticipating) occurred when she saw that the desired antibiotic therapy could not be ordered in the current EMR. This led to her anticipating multiple steps where future errors could occur, and getting the patient the proper therapy would require adaptive/novel responses by multiple microsystems within the hospital. Struggles/Limitations/OpportunitiesA key limitation to recognizing (monitoring and anticipating) is knowing both what to monitor and correctly interpreting the date being moni-tored (i.e., turning raw data into actionable infor-mation). Asking healthcare providers simply to “monitor more inputs” is unrealistic, and thus the Safety II learning step is to become better at monitoring the right things. Using the philosophy “a picture is worth a thousand words,” a graphical information display (such as vital signs or PEWS scores) can potentially allow providers to more efficiently monitor patient status without adding significant workload burden. Visual monitoring systems require leveraging data from the elec-tronic medical record and likely expertise from information technology (IT) and clinical infor-matics specialists.Limitations which could preclude frontline healthcare providers from anticipating potential harm include but are not limited to workload, dis-tractions (from patients, families, coworkers), and fatigue. Anticipating the future is a deliberate act requiring both time and mental energy. To foresee harm, individuals must be mindful and have situational awareness of their current sur-roundings. In our clinical example, the team uti-lized a “stop and resolve” mind-set to determine actions needed prior to using a continuous nafcil-lin infusion, thus increasing odds of things going right. If the clinical pharmacist in our case had not had the time to “stop and resolve” what was needed prior to moving forward with an unfamil-iar therapy, the potential for error and harm would have been substantial. Safety II requires a con-scious effort and deliberate actions to ensure a successful outcome.Finally, a major limitation of recognition is that predicting the future will always be imper-fect. Consequently, individuals and systems may be reluctant to perform in a proactive manner (anticipation) if they do not feel the anticipation prediction is accurate or likely. Consequently, if an action is taken to head off an anticipated untoward outcome, and the untoward event never occurs, one cannot be certain the proactive action avoided an untoward event. One opportunity to improve the efficacy of anticipatory behavior may be through predictive analytics, which uti-lizes “big data” and statistical analyses to develop predictive models about future outcomes and thus can assist human decision-making [17]. Respond Improvement StrategyThe third Safety II component involves respond-ing to a situation once monitoring and anticipa-tion suggest an action is required. At this point, an individual, team, or system has made a delib-erate decision that the current protocol or policy is not appropriate and following the usual or expected practice may lead to error or harm. Multiple questions then arise: How confident are the individual or team that following the expected plan will lead to error/harm? How do they know that any alternative path is safe (or at least safer than the expected path)? Among multiple possi-ble actions, how does the individual or team choose the optimal path? Do nonidentical but similar past situations provide guidance? Will punitive action follow an innovative response? If the alternative actions still lead to error/harm, will a retrospective review conclude that the indi-vidual or team had intentions to take the safest action(s)?Central to responding is creativity. Often called “thinking outside the box,” in reference to

225a psychological experiment from the 1960s, merely telling people they need to think outside the box does not improve their creative ability [18]. Recent research suggests that the presence or absence of particular neural networks predicts an individuals’ ability to think creatively [19]. In our qualitative research, when asked how safety successfully occurs in the PICU, interviewed staff mentioned the ability to respond with inno-vation and creativity more often than other Safety II actions (Monitor, Anticipate, Learn) (Fig. 12.1). Personal characteristics or demeanors most often related to enhancing “responding” included stay-ing calm, working in multidisciplinary teams, expecting rapid-fire questioning, and seeking ideas from outside the microsystem. Anticipated Results of the ImprovementCreating an environment where intended varia-tion in practice is acceptable within limits, with the intention to avoid devolving into randomness or chaos, allows individuals and teams to perform at their highest level and feel empowered to respond to changing circumstances. Ideally, if monitoring and anticipating are working well, instances where responding is needed will be infrequent, and the magnitude of responses will likely be less. How the Improvement Worked in Context of the CaseIn our case, because the clinical pharmacist antic-ipated multiple problems that could result from needing an unusual drug delivery method for continuous nafcillin, she was able to initiate a preemptive response. Aiding the effectiveness of her response was her understanding of the com-plete process, from ordering to drug delivery. She was able to “anticipate and implement” strategies to minimize the potential for error – thus enhanc-ing the probability that things would go right. Specifically, her response included clarifying with infectious disease experts that a continuous nafcillin infusion was the intended treatment plan. She worked with various disciplines to build an order in the EMR. She educated indi-viduals who would be involved in verifying and preparing the medication order. Lastly, she implemented error-proofing strategies, commu-nicating specific instructions for medication stor-age and administration with the nursing staff. Struggles/Limitations/OpportunitiesA limitation to an individual’s or team’s ability to respond creatively is microsystem and organiza-tional culture. Almost by definition, responses involve “going off script” or protocol. In an orga-nization where prior variations in practice resulted in punitive action, individuals may be unwilling to alter their behavior. They may even take the atti-tude that “I’m just going to do what I’ve been told to do, and if something bad happens it is manage-ment’s fault.” Developing a culture wherein employees can thoughtfully vary practice in response to conditions may require leaders to spend time on the front lines demonstrating appre-ciative inquiry (i.e., focusing on what works and what people care about, through discovery, dreaming, designing, and deploying) [20, 21].Other potential limitations stem from some still unanswered questions. For example, is the ability to innovate an inherent psychological skill or something that can be learned? If only certain individuals have the ability to respond creatively, should a team have a critical mass of these indi-viduals at any given moment? Can creative indi-viduals be identified prospectively? Finally, are there ways to assess the effectiveness of creative thinking among individuals (i.e., the person who consistently identifies the “right” path, compared to the person who just creates more problems)? Learn Improvement StrategyThe ability to recognize and respond are related. By learning from experiences, individuals may

226be better informed about which cues to monitor, thus improving their potential to anticipate and respond. Healthcare is a complex sociotechnical system that is continuously changing, creating new situations that are often not predictable and which lead to planned and unplanned adaptations [22]. The ability to learn from responses that went well and improve performance is a key dif-ference between Safety I and II. This shift in approach from responding to past untoward events to being proactive and learning from what and why things go right can support an organiza-tion’s potential to handle a wider variety of conditions. Anticipated Results of the ImprovementIndividuals and organizations learn from not only what goes wrong but also what goes right. At a basic level, when things go right, one can step back to praise those who did well and learn how to respond to the same conditions in the future (“learning from excellence”) [23, 24]. At a higher order of thinking (requiring cognitive process-ing), learning in Safety II will provide generaliz-able knowledge about how to better recognize (monitor and anticipate) and respond to all pos-sible conditions encountered. How the Improvement Worked in Context of the CaseIn our case example, we had a clinical pharmacist who was able to recognize and respond, and no errors occurred. If she and the system took no further action beyond heading off harm in this dangerous situation, the learning opportunity to improve future system responses is lost. In our vignette, “doing things right” was followed by intentional steps to learn. Questions asked and answered included how do we ensure that in the future, continuous intravenous nafcillin is an expected ordering option, and what other medi-cations cannot be ordered in our EMR? This pro-active performance by the pharmacist led to organizational learning, resulting in a novel EMR protocol for ordering continuous antibiotic infu-sions, including but not limited to nafcillin. A more generalizable level of learning can happen when we identify how the pharmacist was able to function in this way and create improvements that make the ability to recognize and respond more likely in the future. Struggles/Limitations/OpportunitiesResources allocated for the sole purpose of learn-ing are often viewed as an expense rather than an investment. Because drawing a direct connection to patient-level outcomes is difficult, the educa-tional budget is frequently the first cut. Fairbanks warns, “Management initiatives must be under-taken sensitively and carefully to avoid underap-preciating the value of apparently nonproductive resources that are contributing to resilience potential and which might be otherwise mis-judged as waste” [25].In addition to learning from what goes right in actual clinical situations, the increased use of simulation can increase learning opportunities. Simulation-based education allows for reproduc-tion of high-risk low-frequency events. Experiential learning occurs by immersing teams in high-fidelity scenario-based simulation with deliberate exposure to disturbances, prompting inexperienced practitioners to learn trade-offs and consequences while managing these distur-bances. Allowing the team to replay the same scenario and apply newly learned behaviors or explore different solutions creates learning rein-forcement. Most learning occurs during focused debriefing immediately following a simulation event. Appreciative inquiry during debriefing can explore methods and frames of mind that prompted innovative or positive productive behaviors. Directed immediate feedback and the opportunity to practice teamwork and communi-cation can contribute to decreased cognitive load, improved adaptive capacity, and a wider range of conditions with sustained high performance. In these ways, simulation allows the opportunity for providers to learn the skills of monitoring, antici-

227pating, and responding without putting patients at risk. (See Key Points Box 12.2).Key Points Box 12.2 Summary• Safety I is the process of learning and responding after an error has occurred. While important, it is ultimately limited in eliminating all patient harm.• Safety II does not replace Safety I. Safety II is the process of learning from what goes right, which offers far more opportunities for spreading improvement.• Applying Safety II utilizes four compo-nents (steps/potentials): Recognize (including Monitor and Anticipate), Respond, and Learn. Each of these is dependent on the others.• Research has identified both individual and system traits which affect the ability to implement the four Safety II compo-nents [12].Editors’ CommentsWe have seen dramatic improvement in patient safety and care quality over the past decade. However, too often safety improve-ments reach a plateau before we have reached the goal (presumably zero patient harm). For example, hand hygiene compli-ance might stall at 98%; serious safety events continue at a low but non-zero rate; a small number of blood stream infections continue to occur. Often our initial belief is that if we keep doing the same things we have been doing to improve, but just do more of it, we will finally get to where we want to be. However, a well-known saying (controversially) attributed to Albert Einstein is that “the definition of insanity is doing the same thing over and over again and expecting a different result.” Are we insane in healthcare? Will continuing to employ the same strategies that enabled us to have drastic reductions in key safety and quality measures finally get us to perfec-tion? Does our current approach – forcing more standardization – have a theoretical limit in complex systems that are continu-ously evolving?Safety II may be part of the solution. The application of Safety II is in the early stages and the authors of this chapter, from Nationwide Children’s Hospital (NCH) in Columbus, Ohio, lead this work in pediat-ric healthcare. The chapter presents a thor-ough and foundational understanding of why Safety II has emerged and how it cre-ates a different approach to improving safety. Readers of this chapter should appreciate the differences between the cur-rent safety improvement strategies in many of our organizations and how Safety II implores us to think differently. The pillars of Safety II presented in the chapter and their descriptions are important for the reader to understand and be able to discuss. The authors eloquently demonstrate how the four components of Safety II (Monitor, Anticipate, Respond, Learn) can coexist with our Safety I strategies (retrospectively analyze and fix) as the two strategies are not mutually exclusive.The best methods for actually imple-menting Safety II thinking and approaches in an organization remain underexplored. The intent of this chapter is to introduce this way of thinking and use NCH as an exemplar of how to ingrain a different mind-set than what we have currently (Safety I). As we strive to reach zero harm, we must embrace different techniques, with Safety II as a prime candidate for the way forward.

228 Chapter Review Questions 1. Which of the following are key differences between Safety I and Safety II? A. Safety I focuses on what went wrong. Safety II focuses on what went right. B. Both Safety I and Safety II see humans as a liability, to be “designed” out of systems. C. Safety I tends to focus on making systems more rigid, while Safety II focuses on making systems more flexible. D. All of the above. E. A and C.Answer: E is correct – Safety II sees human foresight and ingenuity as an asset toward improving safety. The key features of Safety II are that by allowing flexibility/adaptation/resilience to complex or unexpected circum-stances, we can proactively prevent errors from ever occurring. 2. What are the four main potentials/components of Safety II? A. Monitor, Anticipate, Respond, Learn B. Monitor, Avert, React, Leave C. Investigate, Restrict, Enforce, Discipline D. Monitor, Reason, Action, LessonAnswer: A. Hollnagel proposes that Safety II involves four integrated actions as listed in (A). We suggest that two of these – Monitor and Anticipate – might be seen as “Recognize.” 3. Which of the following statements is false regarding the shortcomings of Safety I? A. Learning does not occur until after a criti-cal event has occurred. B. Over time, errors become unusual and unique making learning from events challenging. C. The result is often increased rules and regulations. D. Human error is not considered in the anal-ysis of events.Answer: D. The first three answers are all problems with the Safety I approach. This does not mean that Safety I is useless, but that addition of Safety II to our toolkit will improve safety further. Safety I often assesses for human errors leading to harm. 4. What are the three main components of resilience? A. Toughness, Plasticity, Recoil B. Foresight, Coping, Recovery C. Anticipation, Flexibility, Recoil D. Mindfulness, Anticipation, RecoveryAnswer: B. “Resilience” is the ability of an individual or system to function under circum-stances beyond the usual or outside conditions for which the system was designed. Therefore, coping and then recovering to normal function are required. 5. Which of the following statements is true regarding the weaknesses of a FMEA (failure modes and effects analysis)? A. Has limited value in error prevention as the scope is often too broad B. Primarily focuses on preventing predict-able problems C. Is a core tool in Safety II methodology D. Often addresses all potential errors preemptivelyAnswer: B. While FMEA is useful, the pro-cess is still limited because of the requirement to imagine well in advance things that might go wrong and then make strategic decisions about which possible failure modes to design out of a system. Safety II allows for coping with previously unimaginable circumstances effectively. 6. Which of the following safety approaches praises individuals who perform well with the attempt to learn how to respond to the same conditions in the future? A. Learning from experience B. Learning from praise C. Learning from positivity D. Learning from excellenceAnswer: D. Learning from Excellence describes reporting and analysis of actions individuals took to succeed in a situation. Then the analysis is used to improve safety in similar situations in the future.

229 7. Which of the following is true regarding the difference between Safety I and Safety II methodologies? A. The intention is for Safety II to replace Safety I as it is more effective at prevent-ing safety events in the healthcare setting. B. Safety I efforts are focused on the primary prevention of events, while Safety II eval-uates events after they have occurred. C. Safety I considers deviation in actions to be a liability, while Safety II considers intentional variation by humans as posi-tive and necessary. D. Safety I involves deviation from the proto-col, while Safety II stresses the impor-tance of following institution policies and procedures.Answer: C. Safety II does not replace Safety I. However, Safety II recognizes that flexibility in actions can help a system to “bend and not break.” 8. Predictive analytics, which leverages previ-ously acquired data to develop predictions about the future, is an example of which of the four main Safety II components? A. Anticipate B. Learn C. Resilience D. RespondAnswer: A. The anticipating step of Safety II requires the ability to predict the future. While humans may do this based on experi-ence, heuristics, or “gut instinct,” technologi-cal advances in predictive analytics may augment our ability to know when an event is about to occur.References 1. Patient safety primer: root cause analysis. Agency for Healthcare Research and Quality. 2019. https://psnet.ahrq.gov/primers/primer/10. Accessed 9 May 2019. 2. Charles R, Hood B, Derosier JM, Gosbee JW, Li Y, Caird MS, et al. How to perform a root cause analysis for workup and future prevention of medical errors: a review. Patient Saf Surg. 2016;10:20. https://doi.org/10.1186/s13037-016-0107-8. 3. Joint Commission Resources. Root cause analysis in health care: tools and techniques. 6th ed. Oakbrook Terrace: Joint Commission Resources; 2017. 4. Kellogg KM, Hettinger Z, Shah M, Wears RL, Sellers CR, Squires M, et al. Our current approach to root cause analysis: is it contributing to our failure to improve patient safety? BMJ Qual Saf. 2017;26(5):381–7. https://doi.org/10.1136/bmjqs-2016-005991. 5. McElroy LM, Khorzad R, Nannicelli AP, Brown AR, Ladner DP, Holl JL. Failure mode and effects analy-sis: a comparison of two common risk prioritisation methods. BMJ Qual Saf. 2016;25(5):329–36. https://doi.org/10.1136/bmjqs-2015-004130. 6. Ashley L, Armitage G. Failure mode and effects anal-ysis: an empirical comparison of failure mode scoring procedures. J Patient Saf. 2010;6(4):210–5. 7. Hollnagel E, Wears RL, Braithwaite J. From safety-I to safety-II: a white paper. Published simultaneously by the University of Southern Denmark, University of Florida, USA, and Macquarie University, Australia: The Resilient Health Care Net; 2015. 8. Hollnagel E. Safety-II in practice: developing the resilience potentials. London: Routledge, Taylor & Francis Group; 2018. 9. Westrum R. A typology of resilience situations. In: Hollnagel E, Woods DD, Leveson N, editors. Resilience engineering: concepts and precepts. Burlington: Ashgate Publishing; 2006. p. 59. 10. Hollnagel E. Resilience Engineering. 2016 http://erikhollnagel.com/ideas/resilience-engineering.html. Accessed 24 Apr 2019. 11. Dekker S. The field guide to understanding human error. Burlington: Ashgate; 2006. 12. Merandi J, Vannatta K, Davis JT, McClead RE, Brilli R, Bartman T. Safety II behavior in a pediatric inten-sive care unit. Pediatrics. 2018;141:6. https://doi.org/10.1542/peds.2018-0018. 13. Jeffcott SA, Ibrahim JE, Cameron PA. Resilience in healthcare and clinical handover. Qual Saf Health Care. 2009;18(4):256–60. https://doi.org/10.1136/qshc.2008.030163. 14. Epstein RM. Mindful practice. JAMA. 1999;282(9):833–9. 15. Brady PW, Muething S, Kotagal U, Ashby M, Gallagher R, Hall D, et al. Improving situation aware-ness to reduce unrecognized clinical deterioration and serious safety events. Pediatrics. 2013;131(1):e298–308. https://doi.org/10.1542/peds.2012-1364. 16. Endsley MR. Toward a theory of situation awareness in dynamic-systems. Hum Factors. 1995;37(1):32–64. https://doi.org/10.1518/001872095779049543. 17. Suresh S. Big data and predictive analytics: appli-cations in the care of children. Pediatr Clin N Am. 2016;63(2):357–66. https://doi.org/10.1016/j.pcl.2015.12.007. 18. Burnham CA, Davis KG. The nine-dot problem: beyond perceptual organization. Psychon Sci. 1969;17(6):321–3.

230 19. Beaty RE, Kenett YN, Christensen AP, Rosenberg MD, Benedek M, Chen Q, et al. Robust predic-tion of individual creative ability from brain func-tional connectivity. Proc Natl Acad Sci U S A. 2018;115(5):1087–92. https://doi.org/10.1073/pnas.1713532115. 20. Stavros JM, Torres CB. Dynamic relationships: unleashing the power of appreciative inquiry in daily living (focus book). 1st ed. Chagrin Falls: Taos Institute Publications; 2005. 21. Stavros JM, Torres CB. Conversations worth hav-ing: using appreciative inquiry to fuel productive and meaningful engagement. Oakland: Berrett-Koehler; 2018. 22. Holden RJ, Rivera-Rodriguez AJ, Faye H, Scanlon MC, Karsh BT. Automation and adaptation: nurses’ problem-solving behavior following the implementa-tion of bar coded medication administration technol-ogy. Cogn Technol Work. 2013;15(3):283–96. https://doi.org/10.1007/s10111-012-0229-4. 23. Cawsey MJ, Ross M, Ghafoor A, Plunkett A, Singh A. Implementation of learning from excellence initia-tive in a neonatal intensive care unit. Arch Dis Child Fetal Neonatal Ed. 2018;103(3):F293. https://doi.org/10.1136/archdischild-2017-314737. 24. Kelly N, Blake S, Plunkett A. Learning from excel-lence in healthcare: a new approach to incident report-ing. Arch Dis Child. 2016;101(9):788–91. https://doi.org/10.1136/archdischild-2015-310021. 25. Fairbanks RJ, Wears RL, Woods DD, Hollnagel E, Plsek P, Cook RI. Resilience and resilience engi-neering in health care. Jt Comm J Qual Patient Saf. 2014;40(8):376–83.

231© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 R. K. Shah, S. A. Godambe (eds.), Patient Safety and Quality Improvement in Healthcare, https://doi.org/10.1007/978-3-030-55829-1_13Bundles and ChecklistsGary Frank, Rustin B. Morse, Proshad Efune, Nikhil K. Chanani, Cindy Darnell Bowens, and Joshua Wolovits13Vignette 13.1Over the course of just a few years, a ficti-tious large free-standing children’s hospital began to notice an increase in patient- related safety events with dire conse-quences, including two wrong-site surgeries and a retained foreign body after surgery. At the same time, the hospital was experiencing a large volume of hospital- acquired infections and significant clinician- driven variation in perioperative care. While this hospital’s staff believed in Chapter Objectives• Describe the similarities, differences, and appropriate applications of check-lists and bundles.• Describe a variety of improvement opportunities amenable to checklists and bundles.• Review key literature findings support-ing the use of checklists and bundles.G. Frank Children’s Healthcare of Atlanta, Atlanta, GA, USAe-mail: [email protected] R. B. Morse (*) Nationwide Children’s Hospital, Columbus, OH, USAe-mail: [email protected] P. Efune Children’s Health Dallas, University of Texas Southwestern, Department of Anesthesia and Pain Management, Dallas, TX, USAe-mail: [email protected] N. K. Chanani Quality and Outcomes, Cardiac Service Line, Children’s Healthcare of Atlanta, Emory University School of Medicine, Department of Pediatrics, Atlanta, GA, USAe-mail: [email protected] C. D. Bowens University of Texas Southwestern, Children’s Health Dallas, Department of Pediatric Critical Care, Dallas, TX, USAe-mail: [email protected] J. Wolovits UT Southwestern Medical Center, Cardiac ICU, Children’s Health, Dallas, TX, USAe-mail: [email protected]

232 Opening Question/ProblemIn 1990, James Reason described the Generic Error Modeling System (GEMS) taxonomy for error classification which focused on cognitive factors in human errors [1]. The GEMS model was based on Jens Rasmussen’s three major cat-egories of error: skill-based slips and lapses, rule- based mistakes, and knowledge-based mistakes [2]. Despite an increasing understanding of human error, healthcare was slow to adopt meth-odologies to reduce the likelihood of error. In the year 2000, the Institute of Medicine report “To Err is Human,” highlighting the human and eco-nomic toll of medical error [3].Bundles and checklists are two simple tools, which when judiciously and appropriately applied, can be used to reduce the likelihood of slips and lapses as well as rule-based mistakes. Through a series of case vignettes and a review of the literature, this chapter describes the use of checklists and bundles in healthcare to improve patient safety.Hales et al. define a checklist as a “list of action items, tasks or behaviors arranged in a consistent manner, which allows the evaluator to record the presence or absence of the individual items” [4]. Perhaps the most well-described use of checklists pertains to aviation, where their use can be traced back to 1935 when a Boeing B-17 Flying Fortress crashed shortly after takeoff at a US Army Corps demonstration flight, killing the pilot and co-pilot. At the time, the B-17 was the most complex airplane in aviation history, yet analysis of the wreckage did not demonstrate any mechanical issues. Rather, human error was to blame. The pilot simply forgot to release the flight control gust locks, causing the airplane to nose dive into the ground. After the crash, the checklist was introduced and soon became a mandatory tool for all pilots in the Boeing fleet. The use of checklists in aviation extended from military aircraft to commercial aircraft, and the current enviable accident rate within aviation includes checklists as a key factor.From a human factors’ perspective, cockpit checklists are intended to achieve six objectives [5]: 1. Provide a standard foundation. 2. Provide a sequential framework. 3. Allow mutual supervision (cross checking). 4. Dictate the duties of each crew member. 5. Enhance a team concept. 6. Serve as a quality control tool.Each of these attributes of cockpit checklists is equally relevant to healthcare. For example, providing a standard foundation for care delivery allows clinicians to focus on the unique attributes of individual patients or situations rather than spending mental energy on basic standards. Another key concept is “mutual supervision” or “200% accountability” which allows all caregiv-Vignette 13.2The opening slide read “Keep Me Safe, Heal Me, Be Nice to Me.” The audience uncomfortably shifted in their seats as the Quality Department proceeded to share the stories of two patients that required second operations due to wrong-site surgeries. The presentation then shifted to other prevent-able hospital-acquired conditions, such as catheter-associated blood stream and uri-nary tract infections, ventilator-associated pneumonias, and pressure injuries. The rate of hospital-acquired conditions was on the providing safe, high-quality care, they also valued individual autonomy and decision- making and resisted many forms of stan-dardization. While outcomes were generally quite good, persistent periodic safety events coupled with evolving litera-ture formed a burning platform for change.rise throughout the organization, most notably in intensive care units. These pre-ventable events resulted in higher morbid-ity, longer hospital stays, and increased healthcare utilization and costs. The last slide read “First Do NO Harm.”

233ers, regardless of degrees or seniority, to provide team-member checking. If performed consis-tently, teams become stronger and quality improves.One of the first people to recognize the poten-tial for checklists to transform the quality and safety of healthcare was Dr. Peter Pronovost, a critical care physician at Johns Hopkins Hospital. In 2001, Dr. Pronovost introduced the concept of a checklist to reduce catheter-related blood stream infections at Johns Hopkins. At the time, these infections were often considered to be a known complication of care and were not neces-sarily thought to be preventable. Dr. Pronovost’s work helped change this mindset. After initial local success, Dr. Pronovost was approached by the Michigan Health and Hospital Association to implement checklists in intensive care units across the state through a project commonly known as the Keystone Initiative. In ground-breaking work published in the New England Journal of Medicine and lauded in many layper-son presses, 103 ICUs demonstrated a decrease in the median rate of catheter-related bloodstream infections from 2.7 (mean 7.7) infections per 1000 catheter days at baseline to 0 (mean 2.3). These improvements were sustained during 18 months of follow-up [6] and led to a paradigm shift in the approach to infections and other hospital- acquired conditions.In 2006, Dr. Atul Gawande, a general and endocrine surgeon at Brigham and Women’s Hospital in Boston, was approached by the World Health Organization (WHO) to join a group of people tasked with reducing avoidable death and harm from surgery [7]. At the time, more than 200 million operations were performed annually around the world with complication rates esti-mated to be 3–17%. Furthermore, more than a million people died each year due to complica-tions of surgery. The WHO group considered numerous approaches to improving surgical safety. Training programs, pay-for-performance models, and the development of benchmarks and guidelines were all considered potential opportu-nities to improve safety. At the time, the routine use of a surgical checklist and surgical “timeout” had not yet been considered.Inspired by the pioneering work of Dr. Peter Pronovost, Dr. Gawande suggested that an effec-tive intervention needed to be simple, measur-able, and transmissible. The eventual result was the WHO Safe Surgery Checklist which outlined 19 checks including 7 before anesthesia (e.g., patient allergies), 7 after anesthesia but before incision (e.g., correct procedure and site), and 5 at the end of the operation (e.g., equipment counts). In a pilot study involving eight hospitals with varying levels of sophistication, the rate of major complications dropped from 11% to 7%, and the inpatient death rate following major oper-ations fell from 1.5% to 0.8% [8]. In the ensuing decade, the Safe Surgery Checklist was imple-mented in thousands of hospitals around the world. Indeed, at present, it is hard to find a patient having surgery without the use of checklist.Vignette 13.3The ICU leadership team met to address the problem of the increasing rate of hospital- acquired conditions. The scope and severity of the problem were recog-nized. Brainstorming sessions ensued with a focus on receiving input from key stake-holders. These stakeholders included front- line ICU clinicians (residents and nurse practitioners), bedside nurses, respiratory therapists, nursing leadership, and attend-ing physicians. The staff highlighted the significant variability among providers regarding management of patients, particu-larly in addressing the prevention of hospital- acquired conditions on daily rounds. Audit data from daily rounds revealed that these subjects were not being consistently discussed. For example, not all providers rationalized the use of lab work or established the necessity of indwelling devices. After a series of brainstorming sessions with stakeholders, a list of round-ing topics was developed. Stakeholder meetings continued until consensus was achieved on each of the rounding topics,

234In the past decade, the use of checklists has become increasingly common in healthcare. For example, Weiss et al. examined the use of “prompted” checklists in the medical intensive care unit of a tertiary care university hospital. Physicians were prompted to address six parameters from a daily rounding checklist if those parameters were missed during rounds. A control group used the same checklist without prompting, and both groups were compared to a pre- intervention group. Patients in the prompted checklist group were found to have increased median ventilator-free duration, decreased empirical antibiotic and central venous catheter duration, and increased rates of deep vein thrombosis and stress ulcer prophylaxis. Remarkably, prompted group patients had lower risk-adjusted mortality, whereas control group patients had the same outcomes as the pre-inter-vention group. As an expected caveat, the authors noted that the availability of a checklist alone is not sufficient and that the manner of checklist imple-mentation is critical [9].Several studies have demonstrated the poten-tial for checklists to improve quality of care and rounding efficiency in pediatric intensive care units (PICUs). A PICU Safety Checklist incorpo-rated into the electronic medical record at Children’s Hospitals and Clinics of Minnesota resulted in improvements in quality and safety metrics such as invasive device use, medication costs, antibiotic and laboratory use, and compli-ance with standards of care [10]. Efune et al. described how the use of a rounding template at Children’s Health Medical Center Dallas (see Table 13.1, sample rounding checklist), which included evidence-based ICU standards of care, improved the reliability of discussing patient care goal elements during rounds [11]. Authors at Stanford University demonstrated that a “daily patient goal sheet” in the PICU improved com-munication between healthcare providers, helped nurses identify the in-charge physicians, and was helpful for patient care [12].In an analysis of 29 articles regarding safety checklists in medicine, Thomassen et al. found that checklists improved guideline compliance, improved human factors (i.e., the interaction between humans and other elements of a system), and reduced the incidence of adverse events. Four of the included studies demonstrated a decrease in mortality, and none reported an increase in adverse events. The authors con-cluded that safety checklists are effective tools in various clinical settings [13].These studies clearly demonstrate that the use of checklists can be effective in a variety of dif-ferent clinical venues and healthcare systems. However, for maximal clinical impact, checklists must be kept simple and focused. A bloated checklist is likely to be ignored or only partially completed.establishing that these core topics were vital to patient management and safety, and merited daily discussion on rounds. These core topics were incorporated into a daily rounding checklist. This checklist was then implemented in the ICU. Patient care goals were also incorporated to improve commu-nication between the medical team, bed-side staff, and families. All front-line clinicians, attending physicians, and bed-side staff received education and training on the checklist. The front-line clinicians were tasked with completing the checklist on every patient during rounds, although the attending physicians were ultimately responsible for its completion. A checkbox within the electronic medical record to document that the daily rounding checklist had been performed was implemented to serve as a way to track compliance.Vignette 13.4Six months later, the hospital-acquired condi-tion rate in the ICU was re-examined. Despite initial optimism from ICU leadership regard-ing the implementation of the checklist, the rate of hospital-acquired conditions failed to improve. Input from key stakeholders revealed that the project had potentially expe-rienced “scope creep” from the addition of patient goals into the rounding checklist. To address this, the team decided to narrow the focus to a set of evidence- based, condition-specific, “bundle elements.”

235Table 13.1 Sample rounding checklistMRNDay1234567Date__/__/____/__/____/__/____/__/____/__/____/__/____/__/__SedationGoalSedation holiday□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/AParalytic holiday□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/ASigns of withdrawal□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/ALinesSite/day/remove?Site/day/remove?Site/day/remove?Site/day/remove?Site/day/remove?Site/day/remove?Site/day/remove?CVLArt lineFoley care qshift?□Y □N□Y □N□Y □N□Y □N□Y □N□Y □N□Y □NMedsDiscont/chg to enteral?□Discussed□Discussed□Discussed□Discussed□Discussed□Discussed□DiscussedDrug levels□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/AAdjustment for organ system dysfunction□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/AGI prophylaxis□Y □N□Y □N□Y □N□Y □N□Y □N□Y □N□Y □NDVT prophylaxis□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/AResp/VentGoal blood gas□N/A□N/A□N/A□N/A□N/A□N/A□N/AGoal satERT performed□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A(continued)

236Table 13.1 (continued)MRNERT result□P □F □N/A□P □F □N/A□P □F □N/A□P □F □N/A□P □F □N/A□P □F □N/A□P □F □N/AETT leak□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/AHold feeds for ext□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/APatient careHOB elevated/oral care□Y □N□Y □N□Y □N□Y □N□Y □N□Y □N□Y □NSkin regimen□Y □N□Y □N□Y □N□Y □N□Y □N□Y □N□Y □NTurning q2h?/bath q24h?□Y □N□Y □N□Y □N□Y □N□Y □N□Y □N□Y □NBowel regimen□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/A□Y □N □N/AInitiate or advance feeds□Y □N□Y □N□Y □N□Y □N□Y □N□Y □N□Y □NNasal ETT or nasal NIPPV? Nares ok?□Y □N□Y □N□Y □N□Y □N□Y □N□Y □N□Y □NLabsLab schedule□Reviewed□Reviewed□Reviewed□Reviewed□Reviewed□Reviewed□ReviewedGoalsGoal BP□N/A□N/A□N/A□N/A□N/A□N/A□N/AGoal fluid balance – net even/pos/neg?Care coordAnticipated PICU d/c dayMilestones necessary for d/c

237The concept of a healthcare “bundle” was developed in 2001 by the Institute for Healthcare Improvement (IHI). In a joint initiative with the Voluntary Hospital Association, the Idealized Design of the Intensive Care Unit (IDICU) initia-tive aimed at improving the reliability of critical care processes to improve outcomes at 13 partici-pating hospitals. Collaborators initially found wide variation in care processes and were not able to improve outcomes despite strong participation in the collaborative. An alternative approach was sought, and “bundles” were developed for patients on ventilators and those with central lines.Bundles are defined by the IHI as, “A small set of evidence-based interventions for a defined patient segment/population and care setting that, when implemented together, will result in signifi-cantly better outcomes than when implemented individually.” According to the IHI, bundles have the following properties [14]: 1. The bundle has three to five interventions (ele-ments), with strong clinician agreement. 2. Each bundle element is relatively independent. 3. The bundle is used with a defined patient pop-ulation in one location. 4. The multidisciplinary care team develops the bundle. 5. Bundle elements should be descriptive rather than prescriptive, to allow for local custom-ization and appropriate clinical judgment. 6. Compliance with bundles is measured using all-or-none measurement, with a goal of 95% or greater.A 2009 article by Lachman and Yuen described some additional characteristics of healthcare bundles [15]. For example, they note that each bundle intervention should be scientifi-cally grounded (i.e., evidence-based) and that experts believe that the bundle elements are essential to improving outcomes. They also note that not all possible therapies are included in a particular bundle and that bundle elements should not be “forced” if clinically inappropriate or contraindicated. However, consensus can also drive common- sense inclusion of bundle ele-ments in the absence of evidence. For example, the use of a parachute when jumping from a plane is often cited as a necessary element with-out a clinical trial [16].In one of the first large studies demonstrating the value of bundles in healthcare, Resar et al. demonstrated a 44.5% reduction in ventilator- associated pneumonia rates among 35 hospitals that consistently adhered to ventilator bundle ele-ments [17]. Since then, several studies have dem-onstrated the value of bundles in improving patient safety and reducing hospital-acquired conditions. In a study of 29 pediatric intensive care units across the United States, catheter- associated bloodstream infection rates were reduced by 43% through the implementation of two central line care bundles; one focused on central venous line insertion, and one focused on maintenance [18]. Another quality improvement collaborative involving nine Department of Veterans Affairs hospitals yielded a 48% decrease in catheter-related bloodstream infections and a 41% decrease in ventilator-associated pneumonia by engaging multidisciplinary teams to imple-ment evidence-based ventilator and central line insertion bundles, team rounds, and a daily patient ICU bedside checklist [19].In a partnership with the Institute for Healthcare Improvement, the Surviving Sepsis Campaign incorporated bundles into the diagnosis and treat-ment of patients with severe sepsis and septic shock. The results have been dramatic. For exam-ple, a study of two adult hospitals demonstrated that lack of adherence to the 6-hour sepsis bundle yielded a relative risk of in-hospital mortality of 2.12 [20]. More recently, a meta- analysis of 50 observational studies demonstrated that hospitals that implemented performance improvement pro-grams were able to increase adherence to the 6- and 24-hour sepsis bundles with a resultant decrease in mortality (OR = 0.66) [21].Similar results have been demonstrated in pediatrics. After the death of a pediatric patient with sepsis, New York State mandated an evidence- based approach to sepsis treatment in 2013 [22]. An amendment of Title 10 of the New York State Codes, Rules, and Regulations required hospitals to develop and implement sep-sis protocols for the early recognition and treat-ment of sepsis and to report data regarding

238performance on key measures to the New York State Department of Health. While protocols var-ied by hospital, all protocols were required to include a pediatric bundle consisting of three interventions to occur within 1 hour: blood cul-tures, broad-spectrum antibiotics, and a 20 ml/kg bolus of intravenous fluids.In 2018, a cohort study was published in the Journal of the American Medical Association describing the impact of the New York State mandated bundle for 1179 patients aged 18 years and younger with sepsis and septic shock treated at 59 hospitals. In this cohort, completion of all bundle elements within 1 hour was associated with a lower risk-adjusted odds of in-hospital mortality (odds ratio, 0.59, P = 0.02). Of interest, completion of individual bundle elements within 1 hour did not lead to lower risk-adjusted mortal-ity [23]. This important result points to the “all- or- none” nature of bundles. In other words, bundle elements work synergistically and must all be completed in order to obtain the desired outcome. Clinicians should not choose which bundle elements to complete at any particular encounter. Furthermore, it also demonstrated the importance of consistently implementing the entire bundle without omitting patients.Another innovative use of bundles was dem-onstrated by the Solutions for Patient Safety (SPS) network (https://www.solutionsforpatient-safety.org/), a cohort of pediatric hospitals dedi-cated to eliminating patient and employee harm. The SPS approach to reducing hospital-acquired conditions (HACs) involves the identification and dissemination of evidence-based bundles that are developed by expert panels. Each hospital in the collaborative targets 90% reliability for each bundle and reports both outcome and process (i.e., bundle reliability) data for each HAC.The SPS approach to reducing harm in hospi-talized children has proven to be effective. For example, in a cohort of 33 SPS hospitals partici-pating in a collaborative to reduce pressure inju-ries (PIs), stage 3 PIs declined from 0.06 to 0.03 per 1000 patient days, and stage 4 PIs declined from 0.01 to 0.004 per 1000 patient days after implementation of a 5 element bundle [24]. In the first 3 years of the SPS network, significant harm reduction occurred in 8 of 9 hospital-acquired conditions (HACs) [25], and the network has now grown to more than 130 hospitals. The SPS operational definitions and bundles are available online [26].Bundle implementation by itself is not ade-quate to improve outcomes. Rather, bundle ele-ments must be performed reliably. The importance of consistent bundle adherence was demonstrated in a study of 984 adult ICUs in 632 hospitals. Most of these ICUs had central line bundle poli-cies, but only 69% reported excellent adherence to at least one bundle element (excellent adher-ence was defined as implementing the bundle in ≥95% of patients). ICUs that demonstrated excellent adherence to at least one element had an incidence rate ratio (IRR) of 0.77, and ICUs that had excellent adherence to all five elements had an IRR of 0.67. No association between bun-dle compliance and outcomes for CLABSI rates was found for hospitals that had a written bundle policy but did not measure bundle compliance or had measured compliance of less than 75% [27].Bundle compliance is measured using an “all- or- none” methodology. In a commentary in the Journal of the American Medical Association, Thomas Nolan and Donald Berwick described three advantages of an “all-or-none” approach to Vignette 13.5Despite initial enthusiasm for the bundles, the hospital-acquired condition rate in the ICU continued to be unacceptably high. Review of the literature demonstrated strong evidence supporting bundle imple-mentation, yet ICU leaders were frustrated with the lack of improvement. To better understand how the bundles were being used, a measurement plan was developed. Audits were performed to examine compli-ance with bundle elements. While a few of the individual elements had compliance rates above 90%, most elements demon-strated lower rates of compliance. Moreover, when bundle compliance was measured using an “all-or-none” approach, bundle reliability was found to be well below 50%.

239performance measurement [28]. First, this approach more likely reflects the interests and desires of the patients. In other words, patients and their families likely expect that all evidence- based measures are reliably performed. Second, all-or-none measurement fosters a systems approach to improvement rather than relying on individual performance. Finally, all-or-none measurement offers a more sensitive scale for assessing improvements. Whereas it may be rel-atively easy to achieve 95% compliance with individual elements, an all-or-none approach often highlights the frequency with which cer-tain bundle elements or steps in a process are skipped.As outlined by Gurses et al., several contribu-tory factors impact bundle adherence [29]. These factors include a lack of clarity around the spe-cific goals (tasks) for the patient; confusion about how to complete a step of a guideline (methods); who should do what (individual responsibilities); a lack of understanding of the exceptions to the guideline; and the sanctions (if any) if the proto-col is not followed. Every healthcare system needs to consider each of these factors when building a bundle.To achieve excellent bundle compliance, each element of the bundle must be assessed for acces-sibility. Limiting the bundle to the most salient elements highlights the value of each one. Following a rigorous selection process, including multidisciplinary consensus, can help to promote the high reliability needed for bundle adherence.Measurement of bundle adherence, often referred to as bundle reliability, is a key compo-nent of any improvement efforts involving bun-dles. The creation and distribution of a bundle are not in and of itself sufficient. Rather, bundle adherence should be measured in a continuous feedback cycle. One approach to measurement is through automated monitoring of documentation in the electronic medical record. While this approach is often used, it does not provide an opportunity for real-time feedback.Kamishibai, or K-cards, is an approach to improving bundle reliability that has been encouraged by the Solutions for Patient Safety network. This quality tool is named after a form of storytelling that was used by Buddhist monks in the twelfth century to express stories of moral significance. The monks would show drawings on pieces of paper and tell a story related to the drawings [30]. Reflecting the moral gravity of high-quality manufacturing, the Toyota Production System adopted the moniker Kamishibai and built boards as visual controls for important processes and tasks. Each task or step in a process is displayed on a double-sided card (red and green). If the step is completed appropri-ately, the card is placed on the board with the green side showing. If a problem is encountered or a step is skipped, the card is placed with the red side showing. This allows leaders to easily recognize if critical tasks are being executed con-sistently. As part of their standard work, leaders routinely view the board to assess the status of the process in question.In healthcare, K-cards can be used for audit-ing and visual control of bundle reliability. The individual bundle elements are listed on the front (green) side of the card, and a description of each element is provided on the back (red) side (see Fig. 13.1, Sample CLABSI K-card). The auditor observes the process in question using the K-card as a reference. If all elements of the bundle are completed correctly, the clinician is praised, and the card is placed back on the board with the green side up. If any element is not completed as intended, the auditor provides real-time feed-back. For each K-card audit that is completed, a record is kept of which individual items were missed. The card is then placed back on the board with the red side up (see Fig. 13.2, K-card audit board). The board is often hung in a common area allowing all clinical staff to easily under-stand the performance of their unit relative to the bundle. When all K-cards for a given audit cycle are completed, the results are tabulated and recorded.While the implementation of K-cards to increase bundle reliability in healthcare is a rela-tively new concept, a growing body of anecdotal and peer-reviewed evidence supports this approach. Most notably, K-cards are a key com-ponent of the Solutions for Patient Safety approach to eliminating patient harm in chil-dren’s hospitals, as previously described. A recent peer-reviewed article describes how one pediatric unit within the American Family Children’s Hospital in Wisconsin was able to

240CLABSI Safety CardASK NURSE TO SHOW YOU IN CHART, DOCUMENTATION OF:1. Was necessity of the line discussed at least once within the past 2 completed shifts? • Does the patient have a temporary or long term line?Temporary: Intended for short term use (</= 30 days); Sutured in orstat lock applied; Common locations – umbilical, neck, femoralLong Term: Intended for long term use such as chemotherapyor TPN (> 30 days); Common types: IVAD, Broviac, and Tunneled • If Temporary. Ask nurse to show you documentation of discussion: (N/A if admitted to hospital or line placed within last 24 hours) • If Long Term, Question is N/A : go to the next question2. Are daily care elements documented at least once within the past 2 completed shifts? • If CHG ordered, did the patient receive a CHG bath at least once within the past 2 completed shifts? (N/A if not ordered, medically contraindicated, admitted to hospital or line placed within last 24 hours, or family refusal despite documented provider discussion) • Is there documentation of a linen change at least once within the past 2 completed shifts? (N/A if admitted to hospital or line placed within last 24 hours or if medically unstable for linen change)GO TO BEDSIDE WITH NURSE AND OBSERVE:1. Was the line maintained according to bundle? • Dressing is clean, dry and occlusive • Dressing and tubing dates of change are labeled appropriately • Tubing is away from potential contamination (ostomy bag, diaper) • Curos caps are covering all access points on all lines (N/A if assessed as choking hazard)2. Assess for risk factors necessitating the need for a protective overlay/drape. • Risk factors may include: femoral line placement, line near excessive secretions or emesis, line placed near ostomy, or patient with excessive stooling. • If any risk factors are identified, is there a protective overlay/drape placed correctly? (N/A if no risk factors)NEXT STEPS:Place GREEN side of card showing If ALL elements arecompliant at time of observation. • Give in the moment praise for keeping patients safe. • Complete Safety Board documentation and Rounding for Children’s Health survey.Place RED side of card showing if ANY elements arenon-compliant at time of observation. • Discuss with nurse plan to correct any non-compliant elements in a timely manner. • Complete Safety Board documentation and Rounding for Children’s Health survey.Identify an RN who is caring for a patient with a central line.CARD 1Fig. 13.1 Sample CLABSI K-card. Used with permission from Children’s Health © 2019

241reduce their CLABSI rate from 1.83 to 0 per 1000 line days after the implementation of K-cards [31]. (Key Points Box 13.1)Fig. 13.2 K-card audit board. Incomplete cards are placed on the left. Completed cards are placed on the right with the green side showing only if all bundle elements were completed correctly. Used with permission from Children’s Health © 2019Vignette 13.6End users (bedside nurses, clinicians) in the ICU commented to the project team that the bundle elements were currently not a nor-mal part of their workflow and so were often forgotten. To improve bundle compli-ance, the team turned their attention back to the rounding process. The bundle elements were incorporated into the daily progress note using both free text and data from the electronic medical record. The daily prog-ress note was chosen as the format in which to implement the bundle elements because it served as a script for patient presentations by the majority of ICU front-line providers. Audits revealed an increase in bundle com-pliance; however, the “all-or-none” bundle compliance rate remained below 50%. A second “plan-do- study-act” (PDSA) cycle was implemented, and the bundles with their individual elements were displayed on laminated “badge buddies” which were subsequently distributed to the critical care medicine physicians in training. These senior trainees facilitated bedside rounds and were tasked with ensuring the bundles were appropriately discussed for each patient on morning rounds. Audits were repeated, and this time the “all-or-none” compliance rate increased above the team’s goal of 95%. Through the incorporation of bundle elements into the current ICU work-flow, the project team was able to improve bundle compliance and reduce the inci-dence of hospital- acquired conditions [11].Key Points Box 13.1• A checklist is a “list of action items, tasks or behaviors arranged in a consis-tent manner.”• A bundle is “a small set of evidence- based interventions for a defined patient segment/population and care setting that, when implemented together, will result in significantly better outcomes than when implemented individually.”• Bundle compliance is measured using an “all-or-none” methodology.• Kamishibai refers to a form of storytell-ing that was used by Buddhist monks in the twelfth century. Kamishibai, or K-cards, is an approach to improving bundle reliability that has been encour-aged by the Solutions for Patient Safety network.

242 Chapter Review Questions 1. True or false: Bundles and checklists are the same.Answer: False 2. Describe some similarities and differences between bundles and checklists.Answer: A checklist is a list of action items, tasks, or behaviors arranged in a consistent manner. A bundle is generally limited to 3–5 elements and is measured in an all-or-none manner. In healthcare, checklists and bundles are both tools that can be used to standardize care, avoid unnecessary variation, and reduce the potential for human error. 3. Describe some practical applications of checklists in healthcare.Answer: While there are numerous poten-tial applications for checklists, the most widely reported are the surgical checklist and the ICU rounding checklist. 4. Describe some practical applications of bun-dles in healthcare.Answer: Bundles have most frequently applied to common hospital-acquired condi-tions such as pressure injuries, medication errors, central line associated bloodstream infections, and surgical site infections. 5. Which of the following does not describe a bundle?Editors’ CommentsBundles and checklists save lives; the edi-tors have seen it with their own patients, their practices, their hospitals, and their own family members.We reiterate, unequivocally, bundles and checklists save lives.These repetitive declarations cannot be truer now and in the future with the bur-geoning data and the speed at which infor-mation comes to healthcare providers. There needs to be a way to organize infor-mation and prioritize data. If we can agree that bundles and checklists were trendy, en vogue, helpful, etc. in this decade, then in the next 5–10 years, bundles and checklists will be imperative, ubiquitous, and compulsory.This chapter is a keystone for this text-book; the reader will be able to understand the role of bundles and checklists in health-care and their practices and how to opera-tionalize these tools. The author’s objectives are to ensure the reader under-stands on a cursory level the difference between bundles and checklists. The edi-tors have often time utilized the quality improvement vernacular of “bundles” when discussing outcomes, harm, and hospital- acquired conditions with our exec-utive leadership and hospital boards.Many people intuitively understand checklists, but the concept of bundles is worth explaining; importantly, the poten-tial to partially utilize bundles cannot be underscored. The editors often hear from improvement teams that despite their efforts, they cannot move the needle on a particular outcome. When we dig deeper, we have learned that although the quality improvement leader states that the team is utilizing a bundle, they are not following and tracking all components of the bun-dle – they are only following, for example, 75% of the bundle elements (in other words, following 3 of the 4 bundle compo-nents). This is a very important point for the reader to understand – when teams are using bundles, make sure you know how much of the bundle is being implemented as then the outcomes can be put in the proper context.The authors end the chapter with tangi-ble examples of checklists. Once complet-ing this chapter, the reader should assess their microenvironment and ensure they are utilizing bundles and checklists where appropriate; the guidance and pearls from this chapter should serve as a scaffold for the improvement scientists.

243 A. Typically has three to five elements. B. Each element is relatively independent. C. Should be developed by a multidisci-plinary team D. Bundle elements should be prescriptive without room for local interpretation. E. Compliance should be measured “all-or-none.”Answer: D 6. What is the “all-or-none” concept that is used to describe bundles?Answer: A bundle is only considered to be correctly performed if all bundle elements are completed. Bundle reliability is reported as the percentage of bundles audits in which all elements were performed. 7. How are K-cards used to measure bundle reliability?Answer: K-cards are a visual tool for audit-ing and displaying bundle reliability. Typically, the front of the card is green and the back of the card is red. If all elements of the bundle are performed correctly, the card is placed on the board with the green side show-ing. If any element is not correctly performed, the card is placed with the red side showing. On a regular basis, the results of the bundle audits are tabulated. Bundle reliability is reported as the percent of bundles that are cor-rectly performed.References 1. Reason J. Human error. Cambridge: Cambridge University Press; 1990. 2. Rasmussen J. Skills, rules, and knowledge: signals, signs and symbols and other distinctions in human performance models. IEEE Trans Syst Man Cybern. 1987;SMC-13:257–67. 3. Institute of Medicine. To err is human: building a safer health system. Washington, DC: The National Academies Press; 2000. 4. Hales B, Terblanche M, Fowler R, Sibbald W. Development of medical checklists for improved quality of patient care. Int J Qual Health Care. 2008;20(1):22–30. 5. Degani A, Wiener EL. Cockpit checklists: concepts, design, and use. Hum Factors. 1993;35(2):28–43. 6. Pronovost P, Needham D, Berenholtz S, Sinopoli D, Chu H, Cosgrove S, et al. An intervention to decrease catheter-related bloodstream infections in the ICU. N Engl J Med. 2006;355:2725–32. 7. Gawande A. The checklist manifesto: how to get things right. New York: Metropolitan Books; 2010. 8. Haynes AB, Weiser TG, Berry WR, Lipsitz SR, Breizat AHS, Dellinger EP, et al. A surgical safety checklist to reduce morbidity and mortality in a global population. N Engl J Med. 2009;360:491–9. 9. Weiss CH, Moazed F, McEvoy CA, Singer B, Szleifer I, Amaral LA, et al. Prompting physicians to address a daily checklist and process of care and clinical out-comes: a single-site study. Am J Respir Crit Care Med. 2011;184(6):680–6. 10. Tarrago R, Nowak JE, Leonard CS, Payne NR. Reductions in invasive device use and care costs after institution of a daily safety checklist in a pedi-atric critical care unit. Jt Comm J Qual Patient Saf. 2014;40(6):270–8. 11. Efune PN, Morse RB, Maeve S, Malone LM, Robertson TS, Darnell C. Improving reliability to a care goal rounding. Pediatr Qual Saf. 2018;3(6):e117. 12. Agarwal S, Frankel L, Tourner S, McMillan A, Sharek PJ. Improving communication in a pediatric intensive care unit using daily patient goal sheets. J Crit Care. 2008;23(2):227–35. 13. Thomassen O, Storesund A, Softeland E, Brattebo G. The effects of safety checklists in medicine: a systematic review. Acta Anaesthesiol Scand. 2014;58(1):5–18. 14. Resar R, Griffin F, Haraden C, Nolan T. Using care bundles to improve healthcare quality. IHI innova-tion series white paper. Cambridge, MA: Institute for Healthcare Improvement; 2012. 15. Lachman P, Yuen S. Using care bundles to prevent infection in neonatal and paediatric ICUs. Curr Opin Infect Dis. 2009;22(3):224–8. 16. Yeh RW, Valsdottir LR, Yeh MW, Shen C, Kramer DB, Strom JB, et al. Parachute use to prevent death and major trauma when jumping from aircraft: ran-domized controlled trial. BMJ. 2018;363:k5094. 17. Resar R, Pronovost P, Haraden C, Simmonds T, Rainey T, Nolan T. Using a bundle approach to improve ventilator care processes and reduce ventilator- associated pneumonia. Jt Comm J Qual Patient Saf. 2005;31(5):243–8. 18. Miller MR, Griswold M, Harris JM, Yenokyan G, Huskins WC, Moss M, et al. Decreasing PICU catheter-associated bloodstream infections: NACHRI’s quality transformation efforts. Pediatrics. 2010;125(2):206–13. 19. Bonello RS, Fletcher CE, Becker WK, Clutter KL, Arjes SL, Cook JJ, et al. An intensive care unit qual-ity improvement collaborative in nine Department of Veterans Affairs Hospitals: reducing ventilator- associated pneumonia and catheter-related blood-stream infection rates. Jt Comm J Qual Patient Saf. 2008;34(11):639–45. 20. Gao F, Melody T, Daniels D, Giles S, Fox S. The impact of compliance with 6-hour and 24-hour sepsis bundles on hospital mortality in patients with severe

244sepsis: a prospective observational study. Crit Care. 2005;9(6):R764–70. 21. Damiani E, Donati A, Serafini G, Rinaldi L, Adrario E, Pelaia P, et al. Effect of performance improvement programs on compliance with sepsis bundles and mor-tality: a systematic review and meta-analysis of obser-vational studies. PLoS One. 2015;10(5):e0125827. 22. Dwyer J. Cuomo plans new rules in fight against sep-sis. The New York Times. 2013, January 7. 23. Evans IVR, Phillips GS, Alpern ER, Angus DC, Friedrich ME, Kissoon N, et al. Association between the New York sepsis care mandate and in-hospital mortality for pediatric sepsis. JAMA. 2018;320(4):358–67. 24. Frank G, Walsh KE, Wooton S, Bost J, Dong W, Keller L, et al. Impact of a pressure injury preven-tion bundle in the solutions for patient safety network. Pediatr Qual Saf. 2017;2(2):e013. 25. Lyren A, Brilli RJ, Zieker K, Marino M, Muething S, Sharek PJ. Children’s hospitals solutions for patient safety collaborative impact on hospital-acquired harm. Pediatrics. 2017;140(3):e20163494. 26. Solutions for Patient Safety: Hospital Resources. 2019. https://www.solutionsforpatientsafety.org/for-hospitals/hospital-resources/. Accessed 10 Mar 2019. 27. Furuya EY, Dick AW, Herzig CT, Pogorzelska- Maziarz M, Larson EL, Stone P. Central line- associated blood-stream infection reduction and bundle compliance in intensive care units: a National study. Infect Control Hosp Epidemiol. 2016;37(7):805–10. 28. Nolan T, Berwick DM. All-or-none measurement raises the bar. JAMA. 2006;295(10):1168–70. 29. Gurses AP, Seidl KL, Vaidya V, Bochicchio G, Harris AD, Hebden J, et al. Systems ambiguity and guide-line compliance: a qualitative study of how intensive care units follow evidencebased guidelines to reduce healthcare-associated infections. Qual Saf Healthcare. 2008;17(5):351–9. 30. Niederstadt J. Kamishibai boards: a lean visual man-agement system that supports layered audits. Boca Raton: CRC Press; 2013. 31. Shea G, Smith W, Koffarnus K, Knobloch MJ, Safdar N. Kamishibai cards to sustain evidence-based prac-tices to reduce health care–associated infections. Am J Infect Control. 2018; S0196-6553(18)30992- 1:Epub2018 Dec 3.

245© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 R. K. Shah, S. A. Godambe (eds.), Patient Safety and Quality Improvement in Healthcare, https://doi.org/10.1007/978-3-030-55829-1_14Pathways and Guidelines: An Approach to Operationalizing Patient Safety and Quality ImprovementAndrew R. Buchert and Gabriella A. Butler Opening Question/ProblemDespite evidence that the healthcare industry can adapt process improvement methodologies and high reliability techniques from other leading A. R. Buchert (*) Clinical Resource Management, UPMC Children’s Hospital of Pittsburgh, Pittsburgh, PA, USA GME Quality and Safety, Wolff Center at UPMC, Pittsburgh, PA, USA Pediatrics, University of Pittsburgh School of Medicine, Pittsburgh, PA, USAe-mail: [email protected] G. A. Butler Healthcare Analytics and Strategy, UPMC Children’s Hospital of Pittsburgh, Pittsburgh, PA, USAe-mail: [email protected] 14.1A tertiary-care children’s hospital is seek-ing a solution to rising demand for services, increasing hospital census, escalating costs, and fragmented care. Concurrently, however, the hospital is growing clinical service lines and increasing its focus on patient, family, and staff/provider satisfac-tion. An assessment of the current state reveals high variability in resource utiliza-tion and length of stay, opportunities to improve key quality and patient safety met-rics including unplanned readmissions and healthcare-acquired infections such as CLABSIs and CAUTIs, and a need for a better connectedness along the entire con-tinuum of care including both pre- and post-hospitalization. The executive team turns to the safety and quality leaders of the organization to design an approach that will bridge silos and lead to wide-ranging improvements to ensure the safest, highest quality, most evidence-based, and efficient clinical care.Chapter Objectives• To demonstrate how clinical pathways and guidelines can be used as a tool to reduce patient harm, improve clinical and quality outcomes, and advance organizational strategy.• To understand the essential elements of successful clinical pathways and guide-lines to drive improvement.• To explain the process of clinical path-way and guideline development, mea-surement, and sustainability.• To highlight examples of positive orga-nizational impact from clinical path-ways and guidelines.

246organizations in safety and quality to drive improvement and decrease patient harm, the healthcare delivery system remains dependent on the human-to-human interactions between pro-viders and patients [1]. There is a need for an approach to reduce variability in care that crosses and breaks down silos and allows for clinical algorithms that incorporate both evidence- and consensus-based decision-making and that is data-driven. This chapter will focus on the utili-zation of clinical pathways and guidelines as a vehicle to drive organization-wide improvement and will explore the methods for development, measurement, and sustainability of this tool through a series of case vignettes.The burning platform for improvement in patient safety and quality continues to smolder. Since the To Err is Human report’s release in 1999 [2], more recent literature has suggested that up to 400,000 Americans die every year from medical errors, with serious harm more common than death [3]. In addition to patient harm, today’s healthcare environment is under significant finan-cial pressures, especially as healthcare in the United States remains some of the costliest in the world [4]. High reliability organizations have provided the healthcare industry with models for error reduction through process improvement and strategies for organizational culture change, and as healthcare organizations have embraced these methodologies, outcomes and quality met-rics have begun to improve [5]. But in many orga-nizations, clinical care continues to remain highly individualized with medical decision-making often tailored on a case-by-case basis [6]. Although it is unlikely that variability in health-care will ever completely disappear, nor should this be the expectation, the high reliability orga-nizations have taught us the importance of reduc-ing unnecessary variability [5].It has previously been shown that standardiza-tion of clinical care via pathways and guidelines leads to improvements in the safety and quality of care for both individual patients as well as for the population of patients with a particular diag-nosis or condition [7], and it has also been shown that reduction of variability in clinical care via pathways and guidelines also leads to reduced costs of care [8]. Thus, the development of clini-cal pathways and guidelines needs to be a key component of a healthcare organization’s efforts to improve the care experience, including quality and safety, as well as to improve the health of populations and reduce the costs of care, i.e., the Triple Aim as defined by the Institute for Healthcare Improvement (IHI) [9].In a 2010 Cochran review, researchers exam-ined clinical pathways and guidelines sampled from a large number of different healthcare orga-nizations. Although they noted significant vari-ability in definitions, settings of care, and intended impacts, they found that there were several con-sistencies across all of the pathways and guide-lines of which they sampled. These similarities included multidisciplinary clinical algorithms that translated evidence-based guidelines for care into local organizational culture (evidence- and con-sensus-based care) for specific diagnoses, condi-tions, or populations, and the algorithms were tailored to specific settings of care [10].Vignette 14.2The organization has a long-standing cul-ture of utilizing clinical effectiveness guidelines to help guide care. Over the course of the preceding 10 years, the orga-nization has developed at least 200 guide-lines for both common and rare diagnoses that have been developed and regularly updated by frontline clinical champions. The quality and safety leaders of the orga-nization appreciated that these clinical effectiveness guidelines transcended silos and represented nearly all specialties and settings of care. At the same time, however, the leaders also noted that there were no processes in place to measure if and when these clinical effectiveness guidelines were actually being utilized, nor were there pro-cesses in place to measure the outcomes of utilization of these guidelines. The leaders sought an approach that would be data-driven.

247Clinical pathways are more than tools to reduce variation, however. Successfully imple-mented pathways can also drive organizational strategy. In addition to what has been established so far as key components to clinical pathways and guidelines, it is imperative that pathways also include measurement and a data dissemination and feedback process, as well as a structured development and rollout process to ensure buy-in and support. These are the essential elements that lead to successful clinical pathways and guide-lines that have widespread impact and drive improvement throughout the organization. These elements will be detailed further throughout this chapter (Key Points Box 14.1).The traditional approach to the measurement of clinical pathways and guidelines has focused on assessing the use of and the adherence to the clini-cal algorithm, often accomplished by measuring the use of certain orders or of the accompanying order set, if one exists. Linking this data to the asso-ciated outcomes data (clinical, quality, and safety indicators) can indicate that the adherence to the pathway or guideline has an association with the improvement in the clinical, quality, and safety metrics. An advantage of this method is generally ease of measurement, i.e., the ability to measure the use of a particular order or orders, and/or the use of a particular pre-defined order set is often a simple task in most electronic medical records that can be done readily, quickly, and without much technical expertise. Additionally, many of the safety, quality, and clinical outcomes are already being measured via other methods for required reporting both inter-nally and externally, so a new process does not need to be built. Association does not prove causa-tion, however, and a significant downside to this method is that orders can and often are changed at the point of care, which may result in measure-ments that are not an accurate reflection of actual use of the pathway or guideline.An alternative method for the measurement of clinical pathways and guidelines is to measure Vignette 14.3The leadership of the organization commits to measuring the adherence to and the impact of the utilization of clinical path-ways and guidelines. The leadership engages its information technology experts and decides to build a process of measure-ment that assesses both care that is ordered and care that is actually delivered. The gen-eral surgeons have just revised their clinical pathway for the management of a common surgical diagnosis, and this provides an ideal opportunity to pilot this new measure-ment process. The adherence to the newly revised pathway is outstanding – the rec-ommended perioperative antibiotic is cor-rectly ordered nearly 98% of the time. Key Points Box 14.1 Key Elements of Clinical Pathways and Guidelines• Evidence- and consensus-based• Inter- and multidisciplinary• Integrated into the EMR and workflow• Measurement and feedback/data dis-semination process• Education and rollout process to ensure buy-in• Alignment with organizational strategyHowever, an assessment of the actual anti-biotic administration from the surgical records in the EMR indicates that although an antibiotic is administered prior to the incision 100% of the time, the recom-mended antibiotic is only administered 50% of the time. The other 50% of the time the antibiotic administered is different from the one that was ordered. Further investiga-tion reveals that the recommended antibi-otic is actually not routinely stocked on all of the anesthesia carts, and in these instances a decision is made at the point of care to administer an alternative and read-ily available antibiotic. Once this was iden-tified, the leadership was able to stock the recommended antibiotic on all carts.

248the care that is actually delivered to the patient, as opposed to the previously described method that measures care intended. Measurement of care that is actually delivered not only provides more proof to the effectiveness of the pathway or guideline, but it also provides insight into oppor-tunities for system improvements that may not otherwise be discovered.Measuring care delivered enables insight into opportunities for system improvement and often exposes hidden opportunities. Measuring care that is delivered can also shed light onto other new opportunities for improvement. Length of stay and resource utilization, as well as organiza-tional quality and safety metrics such as unplanned readmissions and opportunities to impact healthcare-acquired infections, may be related to variations in clinical care (Key Points Box 14.2).Clinical pathways and guidelines must incor-porate measurement, ideally of care delivered, and this data-driven approach can drive clinical improvement that leads to subsequent improve-ment in organizational quality, safety, and perfor-mance metrics. An essential element of this data-driven approach, however, is the integration of clinical and information technology (IT) exper-tise. Formerly, clinicians in need of data communi-cated with the IT team via email or other formal request processes, the request was triaged and pri-oritized by the IT team, and once ready, the data would be pulled and delivered back to the request-ing clinician. There were often discrepancies between what was requested and what was deliv-ered, however. For example, a data pull for patients treated for an asthma exacerbation may have come back with much lower than expected volume, and upon further investigation, it is found that only patients with a final diagnosis of asthma exacerba-tion were extracted. The clinician intended the data pull to include all of the variants that would indicate an asthma exacerbation, i.e., status asth-maticus, wheezing, respiratory distress, etc. The clinician knew this but did not realize that this was not intuitive to the IT team who are not clinicians. Vignette 14.4The implementation and measurement of a clinical pathway for a common surgical diagnosis yield a 40% reduction in median length of stay within 18 months, and it also nearly eliminates the patients receiving a central line for prolonged administration of antibiotics. Impressed and moved by this success, the perioperative stakeholders and providers seek to engage in the develop-ment of another clinical pathway, and they Key Points Box 14.2 Measurement of Clinical Pathways and Guidelines• Care intended: orders, order sets (e.g., a third-generation cephalosporin was ordered, as recommended by the clinical pathway or guideline.)• Care delivered: what the patient actually received (e.g., a third-generation cepha-losporin was ordered, as recommended by the clinical pathway or guideline, but the patient actually received a fluoroquinolone.)begin with a critical analysis of their peri-operative data. They find that for another common surgical diagnosis, there is one provider who is an outlier from all of the others, in that this particular provider’s patients are almost never admitted postop-eratively. In further investigation, they find positive deviance – this provider and the anesthesia team have taken a novel approach to the preoperative education, timing of the procedure, and postoperative pain management. The group designs and implements a new clinical pathway that incorporates and makes these novel ele-ments the new standard, leading to a 65% decrease in median length of stay. They also found a 100% reduction in 30-day unplanned hospital readmissions for patients with this particular diagnosis.

249The IT team provided back to the requesting clini-cian exactly what was asked for, but understand-ably, the IT team did not appreciate the clinical nuance of the requested patient population.Accurate and effective measurement of clini-cal pathways and guidelines requires an appreci-ation of clinical nuances. This is best achieved via a close partnership between clinical and IT expertise. The pathway and guideline develop-ment process must include both clinicians and IT team members working together at all points throughout the entire process. While the clinical experts are drafting the clinical algorithm and devising the key metrics, the IT team is concur-rently developing a blueprint for the data pull and measurement logic, i.e., what defines the patient population, which patients are eligible and ineli-gible for the pathway, and what elements must be incorporated into the logic to ensure that the mea-surement is accurate and accounts for the realm of clinical care. This process is dynamic – clini-cians provide the guidance for the patient popula-tion, eligibility, and metrics, and the IT team turns this into the logic. As the logic is tested and data is pulled, this is then reviewed with the clini-cal experts, and tweaks to the logic are made as indicated. This process happens numerous times in an iterative manner throughout the pathway development process and continues to occur fol-lowing release of the pathway, as well. The mea-surement process becomes refined over time, and the result is pathways with adherence and impact data that are trusted by clinicians and, thus, are truly able to change behavior and affect change (Key Points Box 14.3).Clinical pathways can be a key component in the implementation of organizational strategy. As previously discussed, it is important to measure care delivered and not just care intended, and successful measurement requires an integration Vignette 14.5The organization has seen the success of the surgical pathways including reductions in length of stay, reductions of usage of Key Points Box 14.3 Integration with ITClinical experts and IT experts at the table together (literally!) and working hand-in- hand throughout the entire process.central lines, and reductions in unplanned hospital readmissions. Demands for ser-vices and hospital census remain high, and the organization identifies a particular need for a solution to high-demand in one of the specialty ICUs where referrals are consis-tently outweighing capacity. Leadership embarks on a critical examination of the types of patients and diagnoses who popu-late this ICU, and they also benchmark against peer institutions. They find that there is one particular diagnosis of which the organization admits to this ICU at a rate nearly double that of their peer institutions. The organization embarks on a deep dive into these patients and uses a well- established risk stratification guideline for this particular diagnosis to assess whether the patients admitted to the ICU instead of an acute care unit were sicker or higher risk, and they find that this is not the case. They also find that the patients with this diagnosis who are admitted to the ICU have a length of stay that is nearly double that of those who are admitted to an acute care unit, and the patients in the ICU have higher resource utilization including more lab draws and more pharmaceuticals, despite having the same risk stratification and illness severity as those who are admit-ted to an acute care unit (Fig. 14.1). The organization embarks upon the develop-ment of a clinical pathway that includes guidance at the time of admission for where these patients are admitted, as well as that standardizes the management regardless of the setting of care, resulting in a 22% reduction in the median length of stay for these patients, as well as an increase in capacity of the specialty ICU.

250of clinical and IT expertise throughout the entire process. The process of pathway development itself, however, warrants its own discussion. Successful pathways are inter- and multidisci-plinary and evidence- and consensus- based, include a robust measurement and data dissemi-nation/feedback process, are aligned to organiza-tional strategy, and span the continuum of care.The process begins with the formation of the team. At a minimum, the team must consist of clinical champions who are a subset of all stake-holders. For example, a clinical pathway for the management of acute asthma exacerbations must include providers and nursing staff from the emer-gency department, from the inpatient services and acute care units, and from the intensive care units, as well as other stakeholders including respiratory therapists (RTs) and pharmacists. Ideally, there would also be representation from subspecialty stakeholders such as from pulmonology, as well as primary care and urgent care providers and nurses. In addition to clinical champions, the team is led by the project manager who has a combina-tion of a clinical background and technical knowl-edge of the EMR. The team must also include developers who will design the logic and data pull, analysts who review and perform the analy-sis of the data, informaticists to edit orders and order sets, and EMR analysts to complete the changes to clinical documentation. Additional team members include a clinical education spe-cialist to help with the buy-in and rollout process, a marketing/communications specialist to build the concomitant informational and educational materials for both internal staff and those that are patient- and family-facing, and a finance special-ist to measure the cost-savings and economic impact (Key Points Box 14.4).0510152025303540# AdmissionsLow RiskMedium RiskHigh RiskAdmitted to ICUAdmitted to Acute Care UnitsFig. 14.1 The highest risk patients were actually admitted to the acute care units, where their management involved less lab draws and resource utilization, with a shorter length of stayKey Points Box 14.4 Clinical Pathways and Guidelines Team• Clinical champions representing stake-holders, i.e., providers, nurses, pharma-cists, RTs, etc.• IT experts (data pull, logic, order set builds, dashboards, and data analytics)• Clinical education specialists (buy-in and rollout)

251After the team is formed, the pathway devel-opment process commences with evidence- gathering. This includes existing guidelines, perhaps from other organizations or specialty societies, as well as a thorough review of the lit-erature with grading of the strength of the evi-dence. We use the GRADE methodology as our system of evaluating the literature [11]. The design and development phase begins with all stakeholders coming together to devise the clini-cal algorithm based on the evidence as well as group consensus and concomitant development of the data stream including the patient popula-tion, key metrics, and plan for analysis and dis-semination. The clinical champions representing the stakeholders are responsible for taking the clinical algorithm back to their colleagues at multiple points along the way. This ensures that there is widespread buy-in and consensus for the pathway elements. Once the clinical algorithm has been designed, the IT team continues to work on the data elements to operationalize measure-ment and feedback, while the orders and associ-ated orders sets to facilitate the use of the pathway are also developed. The rollout process is a com-bination of local championing by the stakehold-ers, as well as more targeted education by the clinical education specialist. Finally, and perhaps the most influential driver of the pathway, is the data dissemination and feedback process. Self- service dashboards can be immensely useful in providing targeted data around the key metrics for each pathway, tailored to each group of stake-holders. For example, hospital leadership receives high-level summary data around the key metrics for each pathway, while providers receive data on metrics for which they have influence, and nurses receive data on other metrics for which they are the influencers (Fig. 14.2). The stakeholder group should narrow their focus to 3–4 key metrics which are chosen by organizational leadership and the pathway clinical champions and are aligned with organizational goals and strategic priorities. Finally, it should be mentioned that every pathway is an iterative cycle – It is expected that the data after implementation will likely lead Dissemination of DataClinicalChampionsFrontline Staff,i.e. Nurses, Providers,RespiratoryTherapists, etc.Quality andPatient SafetyLeadershipNursingLeadershipProviderLeadershipOrganizationalLeadershipFig. 14.2 Data dissemination is tailored to all stakeholders• Marketing/communication specialists (internal and patient-/family-facing edu-cational materials)• Finance representation (costs of care, resource utilization, financial impact)

252to changes to the pathway and/or measurement process, and so the process described here is a cycle that intentionally does not have an end date (Key Points Box 14.5).Clinical pathways and guidelines link provid-ers and settings of care, and in the process, they transcend silos. They enable the standardization of care that is necessary to improve safety and quality, and they reduce the variability in care that contributes to increased cost and decreased patient and family satisfaction, and that at its worst leads to medical errors. Clinical pathways and guidelines are a tool to drive change, and the impact of successful pathways is widespread, including patients, populations, healthcare orga-nizations, and the healthcare system as a whole. Successfully implemented clinical pathways and guidelines result in improved clinical outcomes; improved quality metrics; decreased safety events; improvements in efficiency, cost of care, and workflows; and increased patient and family engagement and satisfaction. With an integrated development process, improved job satisfaction among IT team members can also be observed as they are directly connected to clinicians and see how their work has impact at the patient level. Clinical pathways are an essential piece to the implementation of an organization’s strategic plan (Key Points Box 14.6).Key Points Box 14.6 Impact of Clinical Pathways and Guidelines• Improved clinical outcomes, including individual patients and the population of patients with this particular diagnosis or condition• Decreased safety events• Improved quality metrics, i.e., reduction of healthcare-associated infections and Vignette 14.6As the organization continues to see improvement in quality metrics, patient safety, clinical outcomes, and operational impact including decreasing length of stay, increasing capacity, and more efficient resource utilization, there are also notable improvements in patient and family satis-faction and staff and provider engagement. Patient- and family-facing educational materials associated with clinical pathways are leading to better engagement of patients and families. They know what care to expect, and why certain things are done, as well as why certain other things may not be done. They see the evidence behind the care that is being delivered, and they see the outcomes data. Providers feel that they are no longer siloed and are on the same page as their colleagues in other settings of care, and patients and families sense this, as well. Given these successes, the organi-zation now embarks upon developing clini-cal pathways and guidelines around the management of chronic diseases that are designed to maximize health and reduce exacerbations requiring ED and hospital utilization. The organization finds that clin-ical pathways and guidelines that span the continuum of care are driving a healthier population.Key Points Box 14.5 Clinical Pathways and Guidelines Development Process• Idea generation• Building the team• Evidence-gathering• Design of clinical algorithm (evidence- and consensus-based)• Design of data pull, metrics and mea-surement, and IT build including orders and order sets• Education and rollout• Measurement and feedback• Dissemination of knowledge• Repeat

253 Chapter Review Questions 1. Describe how clinical pathways and guide-lines can have impact beyond just clinical outcomes.Answer: Impact may include improve-ments in patient and family experience, reduc-tion of hospital readmissions and/or returns to the emergency department, improvements in provider experience and reduction of provider burnout, enhancements to organizational patient flow and workflow, reduction of length of stay, as well as positive economic out-comes. Additionally, pathways and guidelines may help to drive organizational strategy. 2. What are the differences between measure-ment of care intended and care delivered?Answer: Care intended represents what is ordered and is often measured by use of an order set or by pulling specific orders from the EMR. Care delivered represents the care that actually happens, i.e., the care that reaches the patient. Examples of measurement include medication administration data (MAR) data, imaging tests completed, and labs drawn. 3. Describe how measurement of care delivered can lead to improvements in systems of care.Answer: Care intended may not always rep-resent the care that actually makes it to the patient, i.e., care delivered, and thus may not Editors’ CommentsThe authors of this chapter are experts in clinical pathways and guidelines and have transformed the level of care in their orga-nization by following the rubric outlined in this chapter. The authors provide the read-ers with a recipe for planning, implement-ing, and then following up (i.e., measuring) pathways and guidelines. Many organiza-tions, due to resource constraints, chose to only do one of the above tactics with path-ways and guidelines. For example, as cau-tioned by the authors, just implementing without monitoring pathways and guide-lines will not lead to the sustained changes that can be enabled by embracing pathways and guidelines.Furthermore, the authors elaborate on two important facets of pathways and guide-lines that may not be readily appreciated by an organization that is not as well versed in the power of these improvement methodolo-gies. The value of the team cannot be under-stated. This textbook has a whole chapter dedicated to understanding the value of teams in driving change; the authors simi-larly discuss the vital role of the team throughout all stages of pathways and guide-lines. Many organizations involve teams haphazardly, intermittently, or without a firm conviction; the authors of this chapter outline the crucial role of the team and how to use the team concept throughout the life-cycle of pathways and guidelines.The face-to-face collaboration between the clinical team and the information tech-nology team cannot be mitigated. The authors of the chapter explicitly comment on how we need to move past passive com-munication to genuinely engage one other so that the clinicians and the information technology teams can have a shared mental model of the scope, objectives, and sustain-ment of pathways and guidelines.This chapter provides a roadmap for organizations that are just beginning on their journey with pathways and guidelines to those that have matured their processes and need validation or inspiration to take their work to the next plateau in the journey of continuous quality improvement.reduction of unplanned hospital readmissions• Improved efficiency, i.e., workflow enhancements, patient flow, and decreased costs of care• Drive organizational strategy14 Pathways and Guidelines: An Approach to Operationalizing Patient Safety and Quality Improvement

254fully represent the true patient care experience. Measuring care delivered allows for discovery of opportunities for improvement when the care intended and the care delivered do not align. 4. What are the elements of successful and impactful clinical pathways and guidelines?Answer: Successful and impactful clinical pathways and guidelines are as follows:• Evidence- and consensus-based• Inter- and multidisciplinary• Integrated into the EMR and workflow• Measurement and feedback/data dissemi-nation process• Education and rollout process to ensure buy-In• Alignment with organizational strategy 5. Describe the benefits of integrating clinical champions and IT expertise.Answer: Clinicians provide the clinical guidance, and the IT team turns this into the logic. As the logic is tested and data is pulled, this is then reviewed with the clinical experts, and tweaks to the logic are made. The mea-surement process becomes refined over time. The result is pathways with adherence and impact data that are trusted by clinicians and, thus, are truly able to change behavior and affect change 6. Who should be on the team for the design and implementation of a successful and impactful clinical pathway?Answer: Key stakeholders should be included. Some examples are the following:• Clinical champions, i.e., providers, nurses, pharmacists, RTs, etc.• IT experts (data pull, logic, order set builds, dashboards, and data analytics)• Clinical education specialists (buy-in and rollout)• Marketing/communication specialists (internal and patient-/family-facing educa-tional materials)• Finance representation (costs of care, resource utilization, financial impact) 7. True or false: The success of clinical path-ways is driven by measurement and data dis-semination and feedback.Answer: True 8. Based on the discussions in this chapter, which of the following are essential for widespread impact of clinical pathways? A. Measurement and feedback/data dissemi-nation B. Representation of champions representing all stakeholders C. Strategic alignment D. Evidence- and consensus-based E. All of the aboveAnswer: EReferences 1. Chassin MR. The joint commission 13 September 2013, https://doi.org/10.1111/1468-0009.12023. 2. Institute of Medicine. To err is human: build-ing a safer health system. Washington, DC: The National Academies Press; 2000. https://doi.org/10.17226/9728. 3. James JT. A new, evidence-based estimate of patient harms associated with hospital care. J Patient Saf. 2013;9:122–8. https://doi.org/10.1097/PTS.0b013e3182948a69. 4. World Health Organization. Health expenditure, total (% of GDP). 2015 World Bank Group. 2015. Available at http://data.worldbank.org/indicator/SH.XPD.TOTL.ZS. Accessed 6 May 2019. 5. Weick KE, Sutcliffe KM. Managing the unexpected: sustained performance in a complex world. 3rd ed. Hoboken: John Wiley and Sons; 2015. 6. Buchert AR, Butler GA. Clinical pathways: driving high-reliability and high-value care. Pediatr Clin N Am. 2016;63(2):317–28. https://doi.org/10.1016/j.pcl.2015.12.005. 7. Wolff AM, Taylor SA, McCabe JF. Using checklists and reminders in clinical pathways to improve hospi-tal inpatient care. Med J Aust. 2004;181:428–31. 8. Share DA, Campbell DA, Birkmeyer N, et al. How a regional collaborative of hospitals and physicians in Michigan cut costs and improved the quality of care. Health Aff. 2011;30:636–45. 9. Institute for Healthcare Improvement. IHI triple aim initiative: better care for individuals, better health for populations, and lower per capita costs. Cambridge, MA: IHI Initiatives Page; 2015. Available at www.ihi.org/engage/initiatives/tripleaim/Pages/default.aspx 10. Kinsman L, Rotter T, James E, et al. What are clinical pathways? Development of a definition to inform the debate. BMC Med. 2010;8:31. 11. Guyatt G, Oxman AD, Akl EA, Kunz R, Vist G, Brozek J, et al. GRADE guidelines: 1. Introduction- GRADE evidence profiles and summary of findings tables. J Clin Epidemiol. 2011;64(4):383–94. https://doi.org/10.1016/j.jclinepi.2010.04.026.

255© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 R. K. Shah, S. A. Godambe (eds.), Patient Safety and Quality Improvement in Healthcare, https://doi.org/10.1007/978-3-030-55829-1_15Accountable Justifications and Peer Comparisons as Behavioral Economic Nudges to Improve Clinical PracticeJack StevensAbbreviationEMR Electronic medical recordIn 2016, Meeker et al. published results in the Journal of the American Medical Association (JAMA) of a randomized cluster trial to reduce inappropriate antibiotic prescribing among those 47 primary care practices [1]. Those investigators found that two behavioral interventions – accountable justifications and peer comparisons – successfully achieved that objective. Both interventions are consistent with behavioral economics, an interdisciplinary field that utilizes insights from economics, market-ing, and psychology to enhance individual J. Stevens (*) Nationwide Children’s Hospital, Ohio State University Department of Pediatrics, Columbus, OH, USAe-mail: [email protected] 15.1In 2011, Meeker and colleagues found that antibiotics were ordered inappropriately for nearly 25% of primary care visits with diag-Chapter Objectives• To define accountable justifications and peer comparisons• To describe potential mechanisms of action for each strategy• To describe important design consider-ations when implementing these approaches• To assess the strengths and limitations of these approaches for improving clini-cal practicenoses of nonspecific upper respiratory tract infections, influenza, and/or acute bronchi-tis across 47 different practices in Los Angeles and Boston [1]. Such problematic prescribing is hardly unique to those locales. Annually across the United States, tens of millions of antibiotic prescriptions are likely not warranted; they are ordered for indications (e.g., viral infections) that do not respond to these medications [2]. Inappropriate antibiotic prescribing leads to unnecessary side effects, antimicrobial resistance, and avoidable healthcare expen-ditures. Strategies to improve antibiotic stewardship are greatly needed.

256decision-making. Applied by healthcare teams [3], governmental agencies [4], and academic institutions [5], behavioral economics features low-intensity interventions often known as “nudges.” While these interventions are not heavy-handed mandates or bans, these nudges are not neutral. These nudges intentionally encourage decision-makers to make particular choices (Key Points Box 15.1).Through three key events in 2018, behavioral economic nudges received their greatest attention to date for promoting patient safety and quality. To begin with, a commentary about the first behavioral economics team embedded within a healthcare organization – the Penn Medicine Nudge Unit – was published in the New England Journal of Medicine [3]. Next, leaders from 22 North American organizations attended the inau-gural “Nudge Units in Health Care” symposium in Philadelphia [6]. Finally, the topic for the sin-gle keynote speech at the Institute for Healthcare Improvement (IHI) scientific symposium was behavioral economics [7].The novel contribution of this chapter is to highlight specific design considerations when implementing two behavioral economic nudges – accountable justifications and/or peer compari-sons – to improve clinical practice. This chapter is intended to provide more explicit detail on these design considerations relative to previous commentaries [8, 9]. While the context for these strategies in the present chapter is antibiotic stewardship, these two behavioral economic interventions could be applied for a wide range of clinician behaviors. Each of these two behavioral economic strategies will be discussed in turn. The Meeker et al. study was chosen as an exemplary vignette to demonstrate a real-world application of behavioral economics in healthcare [1]. Accountable Justifications DescriptionAccountable justifications feature asking health-care providers to document a rationale for mak-ing questionable clinical decisions. In regard to the Meeker et al. study [1], clinicians were asked to write a justification in a free text box in an electronic medical record (EMR) system when ordering an antibiotic (e.g., penicillin) for a con-dition that would not respond to that medication (e.g., a viral infection like influenza). Clinicians could click a button to cancel the antibiotic pre-scription after the presentation of the accountable justification request. However, clinicians were notified that if they continued to order the antibi-otic and failed to provide an accountable justifi-cation in the EMR system, the words “NO JUSTIFICATION FOR PRESCRIBING ANTIBIOTIC” would be placed in the patient’s record. Those cautionary words featured both capital letters and red font as part of an actual justification request. Potential Mechanisms of ActionThis successful intervention is consistent with EAST framework posited by David Halpern from the United Kingdom [4]. His group established one of the first formal behavioral economics teams embedded within a governmental agency to address diverse public policy objectives (e.g., promoting organ donor registration, increasing tax collections). The acronym EAST stands for Easy, Attract [Attention], Social, and Timely.Regarding Easy, behavioral economics sug-gests that optimal decisions can be achieved by making the desired behavior as natural and auto-matic as possible while making the opposite Key Points Box 15.1 Design Considerations for Accountable JustificationsPhrasing – Neutral versus strongly wordedTiming – Early during the clinical encounter versus later in the clinical encounterAcceptability – Framing this intervention as (1) promoting adherence to best prac-tices, (2) gathering information regarding when guidelines are not followed, and (3) allowing professional autonomy

257behavior difficult to complete. In the present anti-biotic example, an accountable justification requires extra documentation time for clinicians when making questionable prescribing choices.Regarding Attract [Attention], behavioral eco-nomics recognizes that everyone has cognitive limits in attention and memory. Therefore, behav-ioral economics posits that salient information is sometimes needed to encourage optimal behav-ior. The EMR message asking clinicians to enter an antibiotic justification and particularly those cautionary words about failing to provide this rationale are examples of novel and vivid content.Regarding Social, behavioral economics sug-gest that sometimes interpersonal influences, as opposed to basic scientific information, might promote optimal decisions. The accountable jus-tification did not remind clinicians of biological facts they already knew, such as viral infections do not respond to antibiotics. Rather, this behav-ioral economics intervention relied on clinicians’ evaluation concerns about being judged by others who could access these rationales.Regarding Timely, behavioral economics pos-its that optimal decisions can be fostered by pro-viding people with critical information at key moments. The long gap in time between informa-tion provision and decision-making may be one of the leading drawbacks of traditional continu-ing education initiatives. While widely utilized, such continuing education programs often have small effects on clinical practice [10]. In contrast, the accountable justification draws clinician’ attention to the importance of antibiotic steward-ship just as treatment decisions are being made. A close temporal relationship between informa-tion provision and decision-making may enhance outcomes. Important Design Considerations The Phrasing of the Request for the Accountable JustificationSimply asking clinicians in a neutral fashion to “please comment” on a particular clinical deci-sion may not be a sufficient nudge to improve practice. Therefore, more strongly worded lan-guage can be tailored to remind clinicians of the importance of specific standards (e.g., advan-tages of best practices, disadvantages of alterna-tive approaches) and the consequences of not providing a justification. Similarly, allowing cli-nicians to click a selection on an EMR screen with a predetermined list of potential reasons for disregarding a particular guideline may not pro-mote thoughtful reflection. Instead, requesting a more complicated and time-intensive response, such as a free text justification, might facilitate more deliberative thinking. The Timing of the Request for the Accountable JustificationAsking clinicians to provide an accountable justi-fication as they are writing a questionable order seems like the most logical time for this strategy. In the Meeker et al. study, an encounter could not be closed until the prompt for the justification was acknowledged [1]. However, if clinicians have already orally informed patients of particu-lar decisions before placing the order in the EMR system, the accountable justification may be too late to change practice for the current patient. Nevertheless, the accountable justification may change future practice if clinicians become aware that particular decisions will again be called into question. Therefore, if there are opportunities for accountable justifications to appear earlier in the clinical encounter (e.g., while practitioners are reviewing particular presenting problem in the EMR based upon a triage nurse’s assessment), this behavioral economics strategy might have even a more significant impact. Clinician Acceptance of this Behavioral InterventionAn accountable justification could create slightly more documentation for busy clini-cians. This strategy should be framed in a non-punitive fashion. Accountable justifications have two purposes: (1) encouraging adherence to best practice standards and (2) an opportunity for organizations to learn more about the cir-cumstances under which clinicians decide against following guidelines. Regarding the lat-

258ter, organizations can learn when clinicians view guidelines as inappropriate for particular subgroups of patients. Original research or a review of the scientific literature can subse-quently be conducted to ascertain if those excep-tions are warranted. Alternatively, clinicians may document in an accountable justification that certain guidelines are impractical, thereby encouraging organizations to develop strategies to improve workflow. Clinicians May Also View Accountable Justifications as Interfering with Their Professional AutonomyClinicians should be reminded that an account-able justification is not a hard stop. In the Meeker et al. example, clinicians could still prescribe antibiotics for viral infections when they wished to do so [1]. If the accountable justification had mandated that an antibiotic could not be pre-scribed for certain diagnoses, one could envision greater clinician resentment and perhaps some gaming of the system (e.g., coding a viral illness as a bacterial infection to obviate the need to write a justification). Interestingly, Meeker and colleagues found no evidence of such gaming as a result of accountable justifications relative to a control condition [1]. Peer Comparisons DescriptionPeer comparisons feature giving a clinician periodic feedback regarding his/her perfor-mance relative to the behavior of similar health-care providers. This performance may be based upon different sources (e.g., direct observation, review of EMR documentation). Peer compari-sons are often referred to as social norms or descriptive norms because the targeted individ-ual is informed about the routine behavior of similar people (peers). Peer comparisons have empirical support for a wide range of outcomes in the behavioral economics literature, from increasing voter turnout [11] to promoting energy conservation [12].Peer comparisons should be distinguished from widely utilized “audit and feedback” inter-ventions due to the different types of performance profiles. Audit and feedback interventions feature giving a target clinician a periodic summary of her/his performance; audit and feedback inter-ventions sometimes, but not necessarily, feature performance feedback relative to a peer compari-son base. Therefore, essential design consider-ations for audit and feedback interventions [13] may not necessarily apply to peer comparisons. Furthermore, peer comparisons should be distin-guished from peer review. In the latter, a clinician is assessed by a similar healthcare provider who may not necessarily provide a performance eval-uation relative to a peer comparison base. There are also medical staff and potential medicolegal implications of referring to this as “peer review” which is beyond the scope of this chapter (Key Points Box 15.2).Key Points Box 15.2 Design Considerations for Peer ComparisonsBenchmarks – Superstar versus top-tier versus all practitioners versus average practitioner versus slightly high- performing groupSocial approval for high achievers – preventing deterioration for top performersFrequency of peer comparison feed-back – annual versus monthly versus daily/weeklyModality of peer comparison feedback – by person or through electronic meansAudience of peer comparison feed-back – private versus publicTarget behavior – selecting an action that would require only short-term interventionTarget clinicians – characteristics of responsive versus nonresponsive practitionersTarget unit – individual clinician versus larger unit (e.g., floor, healthcare organization)

259In the Meeker et al. study, clinicians received peer comparison feedback on a monthly basis [1]. Each clinician from the top decile – those prescribers who rarely wrote an inappropriate antibiotic prescription relative to their peers – received an e-mail message informing them that “You are a top performer” with a summary of how often she/he wrote an inappropriate pre-scription in that month. The remaining 90% of clinicians received an e-mail message informing them that “You are not a top performer” with a summary of how often he/she wrote an inappro-priate prescription in that month and how that compared to the top decile. Potential Mechanisms of ActionPeer comparisons may exert positive effects through three psychological mechanisms. First, peer comparisons may provide social proof to the target individual [14]. A clinician may recognize the importance of antibiotic stewardship when learning that other practitioners judiciously pre-scribe these medications. Second, peer compari-sons may enhance the self-efficacy of the target individual [15]. A clinician may be convinced that a low inappropriate antibiotic prescribing rate is feasible when learning that other prescrib-ers have already demonstrated this target behav-ior. Third, peer comparisons may motivate a clinician to improve his/her standing relative to other practitioners [16]. Many people view themselves as above average in a variety of domains. Therefore, receiving feedback on sub-optimal performance levels may motivate a pre-scriber to order antibiotics more appropriately so that these positive self-perceptions can continue to be held. Important Design Considerations Benchmarks for ComparisonsA clinician could receive feedback relative to five different standards: the superstar, the top tier, all practitioners, the average practitioner, or a slightly higher-performing group. The advan-tages and disadvantages of each standard are dis-cussed in turn.Allowing a target clinician to receive feedback on what the superstar – the highest-performing healthcare provider – is achieving may promote change as he/she learns there is room for improve-ment. However, such feedback may be disheart-ening when the performance gap is quite large. Alternatively, she/he may view the superstar’s performance with skepticism, concluding that there is something atypical about this highest performer’s patient population.Allowing a target clinician to receive feedback on what the top tier of clinicians is exhibiting may be more beneficial. People may not view themselves as the best performer, but they may view themselves as part of the top 10–20% in a particular category and indeed may strive for this categorization. The performance of this tier may be considered to be an “achievable benchmark” and may motivate him/her to bolster perfor-mance. This top decile was successfully utilized in the Meeker et al. study [1] as well as an earlier study on diabetes care [17]. Again, clinicians in the lowest tier of performance may still view a goal of being in the top group as unattainable.Allowing a target clinician to receive rank- order feedback, in which he/she is shown his/her relative standing on the continuum from lowest performing to highest performing, provides read-ily understandable peer comparisons. This rank- order feedback was more effective than average feedback in a nonclinical domain [18].Allowing a target clinician to receive feedback on what the average clinician is exhibiting may be the simplest option for implementation. However, an average may not sufficiently moti-vate improvement if a clinician cannot easily ascertain what levels of performance are inside and outside of the normal range and may uninten-tionally reinforce complacency.Allowing a target clinician to receive feedback on a slightly higher-performing group of provid-ers may motivate small but realistic levels of improvement. However, providing many differ-ent tiers of feedback depending upon a target cli-nician’s recent performance level may be logistically unwieldy for organizations.

260 Social Approval for High PerformancePeer comparisons are primarily designed to improve the performance of low achievers. However, careful consideration is needed regard-ing how to prevent deterioration of the perfor-mance of high achievers when peer comparisons are given. High achievers may subsequently lower their performance if they learn that others are not exerting the same effort in a particular domain. If both low and high achievers gravitate towards average performance after receiving peer comparisons, there may not be an overall improvement in the target behavior. An example of such an unintended effect occurred in an energy conservation study [12]. When low initial energy users received feedback that they were below the neighborhood average, many subse-quently increased their usage. Only a signal of social approval – a smiley face icon on their energy bills – helped low initial energy users to continue to conserve over time.Referring back to the case vignette from the Meeker et al. study, the top tier of clinicians received a message of social approval known as an injunctive norm [1], which refers to signaling praise or displeasure for someone else’s behavior. They were informed that “You are a top per-former.” This message was intended to encourage those prescribers who rarely wrote an inappropri-ate prescription to maintain their high level of antibiotic stewardship. Frequency of Peer Comparison FeedbackFinding the right balance between infrequent and overly frequent feedback is important. Infrequent feedback (e.g., yearly) gives clinicians rare opportunities to observe improvements in rela-tive standing. Overly frequent feedback (e.g., daily or weekly) would likely be affected by nor-mal variations in patient characteristics or clini-cian behavior, as opposed to more stable changes in practice patterns. In the Meeker et al. study, monthly feedback was utilized [1]. Similarly, Kiefe and colleagues provided peer comparison feedback once every 3–6 weeks in order to pro-mote better diabetes care [17]. Modality of Peer Comparison FeedbackReceiving feedback from an administrator or another clinician – either face to face or over the telephone – may make the peer comparisons very noticeable. However, such feedback may not be realistic to deliver on a large scale and may natu-rally engender feelings of defensiveness. In con-trast, feedback delivered electronically (e.g., e-mail), as was implemented in the Meeker et al. study [1], may require fewer resources to imple-ment but may not be as salient to target clinicians. Audience of Peer Comparison FeedbackPeer comparisons can be presented to a variety of audiences ranging from the individual practitio-ner to a clinical division to the general public. While public presentation might optimize improvement as clinicians seek to protect and enhance their reputations, such a display may have unintended consequences (e.g., leading cli-nicians to select own those patients with the most promising prognoses). Furthermore, if peer com-parisons are presented publicly, extra steps may need to be taken to allow clinicians to correct any errors in individual performance profiles before release. Target Behavior for Peer Comparison FeedbackConsiderable resources in data collection and analysis may be required to periodically provide clinicians with peer comparisons. Long-term provision of peer comparisons may be cost- prohibitive. Ideally, the target behavior should be initially enhanced by peer comparison feedback, but longer-term improvement should be main-tained by other factors. In a follow-up to the orig-inal Meeker et al. randomized trial, Linder and colleagues found that improvement in antibiotic stewardship persisted for those primary care cli-nicians receiving peer comparison feedback even during the year after that intervention was stopped [19]. Peer comparisons may have made routine responsible antibiotic prescribing. Information on relative standing and social approval may not

261be necessary for the long run once a clinician has established a particular practice pattern.A potential unintended consequence of peer comparisons is deterioration in nontarget behav-iors. If practitioners focus their attention on domains featured in peer comparison reports at the expense of other important clinician behavior, there may be no overall improvement in health-care delivery. Therefore, developers of peer com-parison interventions may wish to monitor nontarget behaviors to examine if negative spill-over effects occur. Target Clinicians for Peer Comparison FeedbackAs described in the previously mentioned mecha-nisms of action section, clinicians may respond if they are ambivalent about the utility and feasibil-ity of a particular guideline and/or if they want to compare favorably relative to their peers. However, some clinicians may not fit those char-acteristics. Clinicians who adamantly disagree with a particular clinical guideline are unlikely to change based upon peer comparisons. Alternatively, clinicians who do not care about relative standing or pride themselves on practic-ing differently from their colleagues are also unlikely to be positively influenced by peer com-parisons. Finally, clinicians who have other pri-orities (e.g., preventing long discussions with patients during hectic schedules about the futility of antibiotics for certain infections) may not be influenced by peer comparisons. As can be pre-dicted, clinician characteristics may moderate response to peer comparisons. Target Unit for Peer Comparison FeedbackMany clinical outcomes are the result of team-work as opposed to the efforts of an individual practitioner. Therefore, sometimes comparisons between larger units (e.g., hospital floors, entire healthcare facilities) are appropriate. On the one hand, such comparisons may instill cross-group competition to harness overall system improve-ment. On the other hand, such comparisons may diffuse responsibility for a particular outcome and fail to precipitate accountability and change at the individual level. ConclusionThe present chapter highlighted important design considerations in implementing accountable justi-fication and/or peer comparisons to change clinical practice. Based upon their considerable success in the Meeker et al. study and their cost- effectiveness [20], these strategies appear worthy of consider-ation. However, readers should be cognizant that much remains unknown about these strategies.To begin with, accountable justifications have not been extensively evaluated in other healthcare clinical domains besides antibiotic stewardship. However, in non-randomized studies, similar accountable justifications have been part of multi-component programs aimed at reducing inappro-priately scheduled births at 36–38 weeks’ gestation [21] and promoting adherence to bron-chiolitis guidelines [22]. In addition, peer com-parisons have often had unsuccessful results in the medical literature [23]. Furthermore, little empiri-cal research has been conducted on which param-eters maximize the effectiveness of these two strategies. Finally, accountable justification and peer comparisons should be evaluated head to head versus other behavioral economic strategies like the ones listed in Key Points Box 15.3. Careful evaluation is warranted when utilizing the promising approaches featured in this chapter.Key Points Box 15.3 Empirical Examples of Successful Behavioral Economic Strategies Not Featured in the Present ChapterPublic CommitmentsDefinition: Have clinicians (or other similar physician leaders or influencers in an organization) make public declarations of their pledges to follow evidence-based practice guidelinesExample: Posters containing photo-graphs of primary care clinicians placed in

262their examination rooms with text pro-claiming their commitment to prescribe antibiotics only when necessary [24]Loss AversionDefinition: Utilizing people’s strong dislike of losses to encourage target behaviorsExample: Giving financial bonuses up- front to clinicians at Massachusetts General Hospital that could be deducted from future compensation if performance standards were not met [25]Changing the DefaultDefinition: Altering an EMR field to make the preferred pathway occur as auto-matically as possibleExample: Changing the number of post-operative pills for an opioid prescription from 30 to 12 in a pre-populated EMR field but allowing clinicians to override this lower recommendation and prescribe a dif-ferent number of pills [26]FramingDefinition: Changing the phrasing of an option to encourage optimal behaviorExample: Messages reminding clini-cians that hand hygiene prevents “patients,” as opposed to “you [clinicians],” from catching infectious disease [27]Editors’ CommentsThe author of this chapter is a pioneer in the application of behavioral economics to quality improvement and safety. He has been embraced by his organization to use novel strategies extrapolated from other industries to drive improvement past the inevitable plateaus. In a textbook of this nature, it is impossible to cover the topic of behavioral economics in healthcare in a complete manner; rather, the author chooses to expound on two tactics that are particularly pertinent to healthcare: accountable justifications and peer comparisons.The opening vignette and introduction cite a few of the seminal works in the field of behavioral economics and healthcare. The reader, once seeing the value of incor-porating accountable justifications through the author’s examples, will suddenly see the potential applications of accountable justifications throughout the clinical realm. Indeed, this is the value of this chapter by Dr. Stevens. Once the improvement scien-tist understands the role of accountable jus-tifications, peer comparisons, and other behavioral economic strategies, the world of continuous quality improvement and improvement science becomes logical and less mysterious.Peer comparisons are already a crucial part of a provider’s thought process – how-ever, Dr. Stevens helps provide the under-standing and tools for improvement scientists to consider how to best use peer comparisons on a macro-level within a hos-pital or healthcare organization to drive towards an intended result.It is not fathomable for Dr. Stevens to cover in depth the myriad techniques and strategies of incorporating behavioral eco-nomic approaches to quality improvement in healthcare, especially since the applica-tions are relatively new and constantly being updated. The reader should be left with a deep appreciation of how the next significant gains in quality improvement and patient safety will be from incorporat-ing strategies such as those from behavioral economics.

263 Chapter Review Questions 1. Describe accountable justifications and peer comparisons.Answer: Accountable justifications feature asking healthcare providers to document a rationale for making questionable clinical decisions. Peer comparisons feature giving a target clinician periodic feedback regarding his/her performance relative to the behavior of similar healthcare providers. 2. Discuss why each of these strategies might improve clinical practice.Answer: Accountable justifications may work because they make inappropriate behav-ior more difficult, attract attention to desired behavior, use social influences, and/or provide timely reminders. Peer comparisons may work because they demonstrate that target behavior is acceptable and feasible. Additionally, peer comparisons may work because people want to compare favorably relative to others. 3. What are three important design consider-ations for accountable justifications?Answer: Timing, phrasing/content, and acceptability. 4. What are some important design consider-ations for peer comparisons?Answer: Benchmarks, signaling approval for high achievers, frequency of feedback, modality of feedback, audience for feedback, target behaviors, target clinicians, and target unit (individual clinicians versus teams). 5. From an empirical perspective, describe the limitations of the two behavioral economic strategies discussed in this chapter.Answer: Accountable justifications have little empirical data. Peer comparisons have often not been successful in other studies. Both interventions should be evaluated rela-tive to other behavioral economic strategies.References 1. Meeker D, Linder JA, Fox CR, Friedberg MW, Persell SD, Goldstein NJ, et al. Effect of behavioral interven-tions on inappropriate antibiotic prescribing among primary care practices: a randomized clinical trial. JAMA. 2016;315:562–70. 2. Fleming-Dutra KE, Hersh AL, Shapiro DJ, Bartoces M, Enns EA, File TM, et al. Prevalence of inappropriate antibiotic prescriptions among US ambulatory care visits, 2010–2011. JAMA. 2016;315:1864–1873.3. 3. Patel MS, Volpp KG, Asch DA. Nudge units to improve the delivery to health care. N Engl J Med. 2018;378:214–6. 4. Halpern D. Inside the nudge unit. London: WH Allen; 2015. 5. Behavioral Insights Group, Center for Public Leadership, Harvard Kennedy School. https://cpl.hks.harvard.edu/behavioral-insights-group. Accessed 18 Jan 2019. 6. Leonard Davis Institute of Health Economics. Report from the first national ‘nudge units in health care’ symposium. https://ldi.upenn.edu/news/report-first-national-nudge-units-health-care-symposium. Accessed 18 Mar 2019. 7. Institute for Healthcare Improvement. 2018 IHI national forum keynotes and spotlight sessions – nudg-ing patients and providers to improve quality. http://www.ihi.org/resources/Pages/AudioandVideo/2018-IHI-National-Forum-Keynotes-and-Spotlight-Sessions/aspx. Accessed 18 Mar 2019. 8. Navathe AS, Emanuel EJ. Physician peer compari-sons as a nonfinancial strategy to improve the value of care. JAMA. 2016;316:1759–60. 9. Stevens J. The promising contributions of behavioral economics to quality improvement in health care. Pediatr Qual Saf. 2017;2:e023. 10. Mansouri M, Lockyer J. A meta-analysis of continu-ing medication education effectiveness. J Contin Educ Heal Prof. 2007;27:6–15. 11. Gerber AS, Rogers T. Descriptive social norms and motivation to vote: everybody’s voting and so should you. J Politics. 2009;71:178–91. 12. Schultz PW, Nolan JM, Cialdin RB, Goldstein NJ, Griskevicius V. The constructive, destructive, and reconstructive power of social norms. Psychol Sci. 2007;18:429–34. 13. Ivers NM, Grimshaw JM, Jamtvedt G, Flottorp S, O’Brien MA, French SD, et al. Growing literature, stagnant science? Systematic review, meta- regression, and cumulative analysis of audit and feedback

264interventions in health care. J Gen Intern Med. 2014;29:1534–41. 14. Deutsch M, Gerard HB. A study of the normative and informational social influences upon individual judg-ment. J Abnorm Soc Psychol. 1955;51:629–35. 15. Stok FM, Verkooijen KT, de Ridder DET, de Wit JBF, de Vet E. How norms work: self-identification, attitude, and self-efficacy mediate the relation between descriptive social norms and vegetable intake. Appl Psychol Health Well Being. 2014;6:230–50. 16. Hoorens V. Self-enhancement and superiority biases in social comparison. Eur Rev Soc Psychol. 1993;4:113–39. 17. Kiefe CI, Allison JJ, William OD, Person SD, Weaver MT, Weissman NW. Improving quality improve-ment using achievable benchmarks for physician feedback: a randomized controlled trial. JAMA. 2001;285:2871–9. 18. Aldrovandi S, Brown GDA, Wood AM. Social norms and rank-based nudging: changing willing-ness to pay for healthy food. J Exp Psychol Appl. 2015;21:242–54. 19. Linder JA, Meeker D, Friedberg MW, Persell SD, Goldstein NJ, Doctor JN. Effects of behavioral inter-ventions on inappropriate antibiotic prescribing in primary care 12 months after stopping interventions. JAMA. 2017;318:1391–2. 20. Gong CL, Zangwill KM, Hay JW, Meeker D, Doctor JN. Behavioral economics interventions to improve outpatient antibiotic prescribing for acute respiratory infections: a cost-effectiveness analysis. J Gen Intern Med. 2018; https://doi.org/10.1007/s11606-018-4467-x. 21. The Ohio Perinatal Quality Collaborative Writing Committee. A statewide initiative to reduce inap-propriate births at 360/7–386/7 weeks’ gestation. Am J Obstet Gynecology. 2010;202:243,e1–8. 22. Sprecher E, Chi G, Ozonoff A, Cox J, Patel N, Conroy K. Use of social psychology to improve adher-ence to national bronchiolitis guidelines. Pediatrics. 2019;143:e20174156. 23. Hysong SJ. Meta-analysis: audit & feedback fea-tures impact effectiveness on care quality. Med Care. 2009;47:356–63. 24. Meeker D, Knight TK, Friedberg MW, Linder JA, Goldstein NJ, Fox CR, et al. Nudging guideline- concordant antibiotic prescribing: a randomized clini-cal trial. JAMA Intern Med. 2014;174:425–31. 25. Torchiana DF, Colton DG, Rao SK, Lenz SK. Massachusetts general physician organization’s quality incentive program produces encouraging results. Health Aff. 2013;32:1748–56. 26. Chiu AS, Jean RA, Hoag JR, Freedman-Weiss M, Healy JM, Pei KY. Association of lowering default pill counts in electronic medical record systems with postoperative opioid prescribing. JAMA Surg. 2018;153:1012–9. 27. Grant AM, Hofman DA. It’s not all about me: motivat-ing hand hygiene among health care professionals by focusing on patients. Psychol Sci. 2011;22:1494–9.

265© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 R. K. Shah, S. A. Godambe (eds.), Patient Safety and Quality Improvement in Healthcare, https://doi.org/10.1007/978-3-030-55829-1_16Diagnostic Errors and Their Associated Cognitive BiasesJennifer E. Melvin, Michael F. Perry, and Richard E. McClead Jr. IntroductionIn the fall of 2016, the National Academies of Sciences, Engineering, and Medicine (NASEM, formerly the Institute of Medicine) published a third book in the Quality Chasm Series. The first two books were To Err Is Human: Building a Safer Health System [1] and Crossing the Quality Chasm: A New Health System for the 21st Century [2]. Diagnostic error was finally intro-duced in this third book, Improving Diagnosis in Health Care [3].Diagnostic errors, whether resulting from an inaccurate or delayed diagnosis, are common in healthcare, although they are largely underappre-ciated. The NASEM report [3] estimated that “5% of U.S. adults who seek outpatient care each year experience a diagnostic error.” For years, we have known that 10% of autopsies reveal a prior clinical diagnostic error. A careful review of hos-pital adverse events reveals that 6–17% of these events involved a diagnostic error. Diagnostic errors are the causal factors in the majority of paid malpractice claims. They represent the high-est proportion of total malpractice payments and are twice as likely as other errors to result in the claimant’s death. Most concerning is the fact that each of us, as patients, will experience at least one diagnostic error in our lifetime [3].Diagnostic errors result from a variety of fail-ure modes. These failures result from inadequate collaboration or poor communication among members of the healthcare team, their patients, and the patients’ families; a healthcare system that does not support the diagnostic process; a J. E. Melvin Nationwide Children’s Hospital, Department of Emergency Medicine, Columbus, OH, USAe-mail: [email protected] M. F. Perry Nationwide Children’s Hospital, Department of Hospital Medicine, Columbus, OH, USAe-mail: [email protected] R. E. McCleadJr. (*) Nationwide Children’s Hospital, Hospital Administration, Columbus, OH, USAe-mail: [email protected] Objectives• To introduce the current understanding of the diagnostic process• To describe common cognitive biases that contribute to diagnostic errors• To illustrate how cognitive biases impact the diagnostic process and patient care through a case-based approach

266lack of feedback to clinicians regarding their diagnostic performance; and a culture that is punitive and nontransparent regarding medical errors.The NASEM report presents a conceptual model of the diagnostic process that emphasizes its complexity and iterative nature (Fig. 16.1) [3]. This model describes how a patient with a health problem engages with a medical practitioner and the healthcare system. The medical practitioner working within the system gathers information from the patient, and in collaboration with other members of the healthcare team, integrates and interprets the information to arrive at a working diagnosis. The practitioner communicates this preliminary diagnosis to the patient, and a treat-ment plan is initiated. The clinical response to the treatment is then entered into an iterative feed-back loop at the information gathering step, and the working diagnosis is confirmed or modified accordingly. Eventually, patient and system out-comes result, and learning from the diagnostic process occurs.The NASEM report defined a diagnostic error as the failure to (a) establish an accurate and timely explanation of the patient’s health problem(s) or (b) communicate that explanation to the patient [3]. This definition is patient cen-tered and reflects the iterative and complex nature of the diagnostic process. However, this defini-tion is not without problems. How far amiss from the patient’s true diagnosis must a practitioner’s diagnosis be for it to be inaccurate? Additionally, a timely diagnosis refers to a diagnosis that is not meaningfully delayed, but since the diagnostic process is iterative, some time is required to reach a final diagnosis, indicating that timeliness is context specific. For example, a patient who pres-ents with abdominal pain in the right lower abdominal quadrant may have appendicitis, but the supportive information (e.g., abdominal ultra-sonography) may be initially inconclusive. A patient can present so early in the course of appendicitis that the clinician cannot make the diagnosis. The patient may then be discharged with nonspecific abdominal pain and advised to THE WORK SYSTEM• Diagnostic Team Members• Tasks• Technologies and Tools• Organization• Physical Environment• External EnvironmentAccurate,TimelyDiagnosesOUTCOMESDiagnosticErrors and NearMissesPATIENT OUTCOMESSYSTEM OUTCOMESEffects on Quality, Safety, Cost, Efficiency, Morale,Public Confidence in the Health Care SystemLearning from Diagnostic Errors, NearMisses, and Accurate, Timely DiagnosesCommunicationof the DiagnosisTreatmentPatientEngages withHealth CareSystemPatientExperiencea HealthProblemTHEDIAGNOSTICPROCESSINFORMATION INTEGRATION & INTERPRETATIONDIAGNOSISWORKINGINFORMATIONGATHERINGTIMEFig. 16.1 Conceptual model of the diagnostic process [3]. (Reprinted from Balogh E [3]. Reproduced with permission from the National Academy of Sciences, Courtesy of the National Academies Press, Washington, DC)

267follow-up if the pain does not improve. If the patient then presents with persistent abdominal pain that leads to a conclusive diagnosis of appendicitis and subsequent appendectomy, did the original practitioner make a diagnostic error?In this chapter, we will present a clinical vignette that resulted in a delay in diagnosis and, initially, the incorrect management of a patient. We will then discuss the systems of thinking involved in the diagnostic process and the cogni-tive biases that influence the diagnosis. Finally, we will present a diagnostic autopsy of the clini-cal vignette and the failure of critical thinking that may have contributed to the cognitive biases involved with the case. How Cognitive Bias Impacts the Diagnostic ProcessMany factors contribute to diagnostic error. In a review of 122 closed malpractice claims, Kachalia et al. described four categories of fail-ures that contribute to diagnostic errors [4]. These categories of failure include patient-related fac-tors such as an atypical presentation or a poor historian, system factors such as workload and interruptions, communication factors such as handoffs and conflict, and cognitive factors such as judgment and knowledge. While recognizing that multiple categories often play a role in any given case, these authors demonstrated that cog-nitive factors contributed to diagnostic error in 96% of these cases. Another approach to assess diagnostic errors is to categorize them into the following two main domains: those resulting from cognitive errors and those related to system- based errors [5]. Although many cases may have overlap between these two domains, diagnostic errors stemming from cognitive errors present a unique problem compared to those related to sys-tem process errors, as cognitive errors are diffi-cult to measure and therefore more challenging to improve. Both of these approaches highlight the importance of the cognitive processes in diag-nostic errors. Toward a deeper understanding of these processes, Graber et al. published a litera-ture analysis that characterized three main types of cognitive errors: faulty clinical knowledge, faulty reasoning and/or decision-making, and failure to employ appropriate help when needed [6]. Cognitive biases underlie and often drive these categories of cognitive errors. (Key Points Box 16.1)Diagnostic reasoning is complex, and cogni-tive biases play an integral role. Often studied in psychology and behavioral economics, cognitive biases are also prevalent in the medical profes-sion. The automatic acceptance of cognitive biases, or “rules of thumb,” often leads to predict-able and recurrent results. These biases are fre-Vignette 16.1A teenage girl presents to an emergency room with a history of abdominal pain and vomiting. Her physical exam is unremark-able. She is given the diagnosis of gastro-esophageal reflux disease (GERD) and discharged with medication to treat her symptoms. Despite medication compli-ance, her symptoms persist. She is subse-quently evaluated by multiple providers, all of whom confirm the diagnosis of reflux and further intervene only to expand her pharmacotherapy. When her symptoms continue to persist, a practitioner ultimately diagnoses her with a psychosomatic disor-der. Subsequently, she is lost to follow-up. When she presents months later to the healthcare system for worsening of her symptoms and significant weight loss, a more thorough clinical history is obtained, and the correct diagnosis and treatment are established.Key Points Box 16.1 Diagnostic ErrorAs defined by the National Academy of Medicine, a diagnostic error includes a fail-ure to (1) establish an accurate and timely diagnosis or (2) communicate that diagno-sis to the patient.

268quently employed when making quick, snap judgments in the face of uncertainty. In the appropriate setting, cognitive biases enable rapid and efficient decisions and actions and are there-fore valuable when timeliness is imperative or during cognitively demanding tasks. However, at times, cognitive biases result from a limited capacity for information processing and may therefore lead to inappropriate or wrong conclu-sions. Over one hundred cognitive biases relating to decision-making, behavior, memory, and social biases have been described, and many of them play a critical role in diagnostic errors. Some of the more recognizable biases that oper-ate in the diagnostic process are reviewed in Table 16.1. System 1 and System 2 ThinkingWhen faced with a problem, decision-making often takes one of two forms of thinking. Fast, quick thinking allows a person to jump to a con-clusion automatically. Slower or more focused thinking involves paying close attention to the thought process, the possible solutions, and the possible outcomes. Psychologists describe these dual forms of thinking as System 1 and System 2 thinking. System 1 thinking refers to the faster, automatic, and unconscious model, whereas System 2 thinking refers to the slower, conscious, and effortful model [7, 8]. System 1, therefore, can be compared to an intuitive track, as it oper-ates rapidly and with little voluntary control. Information processed along this track includes reading a simple sentence or reading 2 + 2 =? and automatically knowing the answer is 4. System 2 thinking, however, is more deliberate and slow. Thinking along the System 2 track is useful when concentration is required, such as in filling out a tax form or solving the equation 48 × 36 =? [7].This dual process theory of thinking and decision- making is also employed when clini-cians are faced with a diagnostic challenge [8, 9]. System 1 thinking often defaults to utilization of cognitive biases to reach a conclusion. Clinicians are constantly encountering new patients and pre-sentations throughout their training and clinical experience. Over time, clinicians learn symptom patterns, also known as illness scripts. With rep-etition, quick pattern recognition allows clini-cians to utilize System 1 thinking and effortlessly categorize certain illness scripts into the correct diagnostic category. This quick, automatic think-Table 16.1 Cognitive biasesCognitive biasesDescriptionAnchoring biasThe tendency to lock onto salient features and fail to adjust the initial impression after further studiesAvailability biasThe tendency to initially think about whatever diagnosis most readily comes to mindBandwagon effectThe more other people draw similar conclusions or have comparable perceptions, the more likely subsequent individuals come to the same conclusionsConfirmation biasA self-perpetuating cycle in which there is a tendency to look for confirming evidence to support an initial diagnosis, sometimes at the expense of reviewing disconfirming evidenceDiagnostic momentum biasWhen a potential diagnosis is passed from one provider to another and as such becomes more “sticky,” until it finally is accepted as the true diagnosis despite an incomplete evaluationFraming effect biasThe tendency to be influenced by the way the problem presentsOverconfidence biasThe tendency to act on incomplete information due to overconfidence in judgments, which may result in incomplete differential diagnosesPremature closureOnce a clinician makes a diagnosis, less effort is put into trying to disprove itRepresentative biasMaking judgments about events based on personal experiences and preconceived notions rather than based on their actual likelihoodUnpacking principleA failure to elicit all relevant information; therefore, the differential diagnosis may be left incompleteVisceral biasMay result in missed diagnosis due to countertransference of positive or negative feelings toward patients

269ing may be particularly beneficial in emergent situations such as those that present to the emer-gency department. For example, rapid recogni-tion of Cushing’s triad in a trauma patient allows for immediate lifesaving interventions instead of a delay in care due to evaluation for other poten-tially non-emergent etiologies. However, this quick System 1 thinking can fail both the clini-cian and the patient in several ways. If a bias clouds the initial patient presentation such that the clinician fails to elicit the entire story or if the workup is prematurely or inappropriately com-pleted, the clinician may place the patient into the wrong diagnostic category. In this way, reliance on System 1 pathways and, indirectly, cognitive biases may lend itself to a diagnostic error [10]. (Key Points Box 16.2)In contrast, thinking slower and therefore forc-ing utilization of the System 2 pathway may improve decision-making in certain circum-stances. This approach employs deliberate, logi-cal thinking and therefore allows for deeper thought and reflection. This method may be ben-eficial in a medically complex patient or in a patient that does not neatly fit into an illness script. In medicine, however, slow and deliberate thinking is neither always necessary nor required. A large proportion of patients do fall into easily recognized patterns and diagnoses, and recogniz-ing these patterns decreases unnecessary testing and utilization of resources [11]. It is unreason-able then to suggest that System 1 thinking should never be utilized in medicine. More importantly, clinicians should recognize and acknowledge their biases and knowledge deficits. If these biases and deficits are recognized and acknowledged while thinking about the patient, System 2 think-ing can also be activated, and diagnostic errors due to cognitive bias may be prevented. Teaching Diagnostic ReasoningConscious utilization of the System 2 pathway is an important step in the improvement of clini-cal reasoning. Royce et al. have described clini-cal reasoning as “the process of applying cognitive skills, knowledge, and experience to diagnose and treat patients” [12]. Therefore, to improve clinical reasoning, a practitioner needs to develop an understanding of their underlying cognitive skills and biases, improve their knowl-edge base, and increase their diagnostic experi-ences. Experiences often come with time and volume, while knowledge, although a lifelong process, starts early in a physician’s career. Methods for teaching cognitive skills are not well described but may be less effective until a physician has a solid foundation of medical knowledge.Improved recognition and teaching of clinical reasoning skills are important in the overall goal of improving diagnostic accuracy. In several mal-practice cases attributed to diagnostic error, the errors were found to be secondary to a “failure to consider the correct diagnosis” and not due to ignorance of the correct diagnosis [12, 13]. Often, recognizing where the clinical reasoning went wrong when a diagnostic error occurred is diffi-cult. An important and fundamental component of clinical reasoning is cognitive bias. Therefore, if we teach critical thinking and debiasing tech-niques, we will improve the overall diagnostic process.Teaching metacognition, or “the capacity for self-reflection on the process of thinking and self-regulation in monitoring decision making,” works synergistically with critical thinking and debiasing technique education [12]. Together, these skills encourage the provider to think slower and spend more time utilizing the System 2 logical thinking track. Awareness of the auto-matic acceptance of results provided by a cogni-tive bias brings those biases to the forefront. This recognition allows the bias to be observed and critically considered and can lead to a conscious decision to either accept or reject the diagnosis that the bias has led the provider to. (Key Points Box 16.3)Key Points Box 16.2 Dual Process TheorySystem 1 Thinking – quick, automatic, unconscious mode of thinkingSystem 2 Thinking – slow, deliberate, purposeful mode of thinking

270However, reviews of studies teaching meta-cognition and debiasing strategies have led to a variety of outcomes. Understandably, teach-ing debiasing techniques to medical students has not demonstrated a benefit for improving diagnostic accuracy. This observation is likely related to overall knowledge deficits in medi-cal students. If knowledge is lacking, certain cognitive biases are likely not present, and diagnostic accuracy will suffer, regardless of biases [14–16]. Similarly, in diagnoses with classic presentations, the clinician often knows the diagnosis with a high degree of certainty, and teaching debiasing strategies and critical thinking will not likely alter their (correct) diagnosis. Encouraging critical thinking or analytical reasoning is, therefore, most benefi-cial for complex or atypical cases. In one study [17], medical residents evaluated computer-based cases of varying complexities. In cases with classic presentations in which the diag-nosis was clear, the authors found that no amount of analytical reasoning would change the clinicians’ minds. However, when the cases became more complex with conflicting data, cognitive biases such as premature clo-sure and subsequent diagnostic errors were more prevalent. Therefore, in complex or unusual cases, they noted that “physicians would benefit from better awareness of cogni-tive processes and the application of rigorous analytic reasoning” [12]. Several other studies have demonstrated that providing training in reflective practice, reasoning skills, and prob-abilistic decision-making do improve diagnos-tic accuracy, often by decreasing cognitive biases [18–20]. Diagnostic AutopsyTwo of the more commonly encountered cog-nitive biases are anchoring bias and premature closure bias. As described by Daniel Kahneman [7], the quick, autonomous style of System 1 thinking depends on pattern recognition to make an accurate diagnosis. However, this system of thinking is prone to cognitive bias as noted above. Anchoring bias, the tendency to lock onto salient features and failure to adjust this initial impres-sion, may be particularly apparent when patients present from other facilities, as they often arrive with a labeled diagnosis. Likewise, documented triage chief complaints may also provide the phy-sician with a “pre-diagnosis.” Anchoring from this initial diagnosis may lead to inappropriate Key Points Box 16.3 MetacognitionThe ability to critically evaluate one’s own thinking, commonly referred to as “think-ing about thinking.” It is considered a criti-cal component of learning.Vignette 16.2“Jessica,” a 17-year-old female with no sig-nificant past medical history, presents to a pediatric urgent care facility with 1 month of epigastric pain and vomiting. She describes the pain as daily, dull, and non- radiating. She has not been able to achieve any relief from her symptoms. Given the clinical history and unequivocal physical exam, the urgent care physician transfers the patient to a large quaternary pediatric emergency department for ongoing evalua-tion. In the emergency department, further clinical history reveals that the pain seems to be worse in the morning and after meals. Symptoms were consistent with GERD, so no further laboratory or radiographic evalu-ation was obtained. Review of the elec-tronic health record (EHR) did not reveal a documented differential diagnosis in the physician’s note, so it is unclear if the phy-sician considered other diagnoses. Jessica was discharged home with parental reas-surance, a prescription for ranitidine, and follow-up with her primary care doctor.

271diagnostic conclusions. Similarly, the premature closure bias, in which a provider fails to consider alternative diagnoses once a primary diagnosis is assigned (either by that provider or another), may occur in these situations. Applying this principle to the conceptual model of the diagnostic process as outlined in Fig. 16.1, a clinician’s information gathering will be limited by the premature clo-sure bias. This limitation can unintentionally result in failure to complete a thorough clinical history, failure to consider necessary diagnostic tests, and failure to consult pertinent sub- specialty services in appropriate clinical scenarios. The failure to consider a more thorough differential diagnosis likely led the clinician to obtain no additional documented clinical history or evalua-tion in this case presentation.During a busy emergency department shift, providers can see dozens of patients with similar symptoms, so they must often rely on illness scripts to help identify, triage, and treat patients. If a provider frequently sees patients with reflux symptoms, he or she would be more likely to make that diagnosis when a teenager with epigas-tric pain and vomiting presents to the emergency department. This tendency to diagnose based on previous experiences with illness scripts with similar symptoms is called the availability bias. This bias can occur when healthcare providers make decisions based on perceived frequency or likelihood of a condition due to the ease of recall rather than actual probabilities [21]. One study found that only 27.9% of 10- to 17-year-old chil-dren with symptoms of gastroesophageal reflux reported experiencing abdominal pain [22]. Therefore, although abdominal pain occurs with reflux, it is unlikely that every patient presenting to the emergency department with epigastric pain has GERD. This bias tends to lead to the overdi-agnosis of the more common conditions and underestimation of the true prevalence of rare diseases. The availability bias inherently lives in the common medical aphorism, “when you hear hoof beats, think horses, not zebras.” In Jessica’s case, multiple providers were quick to attribute chronic epigastric pain and vomiting to GERD, a routine diagnosis in pediatric and adolescent medicine. However, her actual diagnosis was less common and required more thoughtful consider-ation to make an accurate diagnosis.The patient followed up with her primary care provider as instructed to re-evaluate her com-plaints. She continued to endorse abdominal pain, nausea, and vomiting despite treatment with two prescription medications for reflux. During this visit, her high-risk adolescent behav-iors are commented upon in the physician docu-mentation, but the connection to the primary complaint was not established. The physician appropriately took a more detailed clinical his-tory but concluded that the patient was likely suf-Vignette 16.3The next day, Jessica returns to the same emergency department with continued epi-gastric pain, nausea, and vomiting. The emergency room physician assessing her in all likelihood has seen numerous adoles-cent patients with a similar presentation who were diagnosed with GERD and suc-cessfully treated. In the medical record, the provider comments on the patient’s reas-suring clinical appearance and recent diag-nosis of GERD, so a prescription for a cytoprotectant, sucralfate, is added to her regimen. She is discharged home with a continued emphasis on supportive care.Vignette 16.4The following day, Jessica sees her primary care doctor to discuss these symptoms. At that visit, the physician completes the first fully documented adolescent psychosocial screen. Jessica endorses occasional drink-ing and trying marijuana “a couple of times.” There is a concern for a possible “psychosomatic” etiology to her symptoms; however, the patient is told to continue sup-portive care with a new prescription for ondansetron and to follow- up in 1 week.

272fering from psychosomatic symptoms. This tendency to use evidence that supports your working diagnosis (e.g., benign abdominal exam) but ignore other evidence that may contradict it (e.g., history of marijuana use, no improvement on appropriate medications) is known as the con-firmation bias. This bias likely impaired the pro-vider’s diagnostic thinking and inhibited the acquisition of a more detailed history.As mentioned, this patient was a healthy, well- appearing teenager with a history of marijuana use who, despite persistent symptoms, had a benign abdominal exam. This benign exam led multiple providers to assume that this patient’s symptoms were more representative of functional abdominal pain rather than a more complicated diagnosis. Tversky and Kahneman discussed this inherent tendency to make a judgment based off something being more representative rather than more likely and labeled it the representativeness bias [21]. They illustrate this bias through an example. They describe a shy and timid individ-ual who is meek, tidy, and introverted and has exceptional attention to detail. They then ask sub-jects to assess the probability of this individual’s profession when given the choices of farmer, physician, librarian, or pilot. The representative-ness bias would stereotype this individual as best fitting our preconceived notions of a librarian. In our clinical vignette, providers stereotyped a well-appearing, healthy adolescent female with intermittent marijuana use as more likely to have functional abdominal pain than a more complex diagnosis.The 2002 NASEM report, Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care, highlighted these inherent stereo-types and implicit biases [23]. While this chapter will not go into the complex nature of current healthcare disparities, it is important to note that stereotypes and implicit bias can alter the diag-nostic process. When stereotypes compromise the healthcare system and its providers, they may detrimentally affect the entire conceptual model of the diagnostic process. These biases negatively impact the way patients and families engage in the healthcare system. They also adversely impact how providers formulate a working diag-nosis and how they communicate the diagnosis and treatment options. Patient outcomes can also unfortunately be negatively impacted.In addition to cognitive biases, other system processes can contribute to diagnostic errors. The leading paradigm of thinking that illustrates this concept is James Reason’s Swiss Cheese analogy [24]. In his model, potential hazards can reach victims when the holes (latent failures) in many different processes are perfectly aligned to result in preventable patient harm. These holes include both active failures, such as unsafe acts commit-ted by individuals in direct contact with the patient or system, and underlying conditions, such as inherent flaws within a system that can contribute to harm [24]. In medicine, the EHR is a potential hazard that can contribute a latent fail-ure as described by Reason’s Swiss Cheese analogy.The inherent flaws in EHRs are demonstrated in our clinical vignette, as provider interaction with the EHR contributed to delayed recognition of the patient’s diagnosis. One of the most widely used EHR features is its ability to highlight a patient’s medical and social histories. The pro-vider who first discovered Jessica’s history of marijuana use failed to document this history. Subsequently, multiple following providers also failed to complete this social history, and the EHR system failed to alert the providers that the social history section was incomplete. This infor-mation may have triggered future providers to either obtain a more detailed social history or consider marijuana use as a contributory factor in her diagnosis. Instead, the providers committed an error of omission, which supported the pre-vailing theory that her pain was secondary to GERD.Vignette 16.5After the patient’s visit, the physician com-pletes documentation in the EHR. The clin-ical indication for this visit is “follow-up exam,” with no other documented diagno-sis of marijuana use.

273Other inherent flaws with EHRs that require consideration are the use of highly focused order-sets and the enormous volume of documentation that EHRs create. Pre-templated ordersets, while helpful for efficiency, may bias the clinician toward premature closure, as ordersets are often very narrow in their diagnostic focus. The vol-ume of documentation in EHRs must also not be underestimated. Large quaternary medical centers will generate tens of millions of notes in a given year. This staggering statistic makes it nearly impossible for medical providers to review all documentation, especially when patients have numerous prior encounters. With increasing external pressures from administrators and insur-ance companies to expedite patient care and reduce the length of stay, providers are often unable to dedicate the necessary time and resources required for detailed chart review.Upon receiving the diagnosis of GERD, the patient had several subsequent encounters with medical providers who continued to attribute her symptoms to reflux. Her ongoing complaints of nausea, vomiting, and epigastric pain that were refractory to medications should have been a warning sign that GERD was not the correct diagnosis. At the core of the conceptual model is the question “has sufficient information been col-lected?” This question challenges providers to regularly reflect on whether key components of a proposed working diagnosis are missing. In this instance, medical providers had tunnel vision, failed to seek new information, and failed to con-sider an alternative diagnosis. Provider assump-tion of the correctness of the initial diagnosis to make subsequent medical decisions is known as the anchoring bias.Also, this mode of thinking perpetuated a sec-ondary cognitive bias, the bandwagon effect. The more other people draw similar conclusions or have comparable perceptions about a case, the more likely subsequent individuals are predis-posed to those same biases and come to the same conclusions. Encounter after encounter, this patient was labeled as suffering from GERD. Minimal diagnostic evaluations were completed on each subsequent encounter, and less attention was paid to the fact that her symptoms persisted despite being on the appropriate treatment for GERD. Rather than stopping to take a “diagnos-tic time-out” and to consider other possible explanations, the clinicians continued to treat Jessica for GERD. (Key Points Box 16.4)Vignette 16.6Twelve days after her initial clinic appoint-ment, Jessica is seen again for a follow-up visit. Her symptoms continued to be attrib-uted to underlying GERD; yet, there is no discussion or documentation at this visit of her previously identified high-risk adoles-cent behaviors. The patient is sent home with education about GERD prevention.Key Points Box 16.4 Common Types of Cognitive BiasesRepresentativeness – probability that an event or object belongs to another category based on the characteristics it shares with the categoryAvailability – making a judgment about likelihood based on how easily the event or object comes to mindAnchoring – heavily relying on an ini-tial impression to make subsequent judgmentsVignette 16.7Jessica is then lost to follow-up for several months until she finally returns to the clinic with 12-pound weight loss, persistent abdominal pain, and near-daily vomiting. For the first time, a provider documents in the EHR a detailed differential diagnosis including irritable bowel syndrome, inflam-matory bowel disease, hyperemesis from marijuana, bowel perforation, and appendi-citis. The patient is transferred to the same quaternary pediatric emergency depart-

274Cannabinoid hyperemesis syndrome (CHS) is a well-documented sequela of chronic abuse of marijuana that consists of recurrent abdominal pain with unrelenting nausea and vomiting. One of the most pathognomonic characteristics of CHS is the relief of symptoms by hot showers. While the true incidence of CHS is unknown, multiple case reports have been published in both the pediatric and adult literature [25–27]. Mitigating Cognitive BiasAs discussed, cognitive biases play an integral role in many diagnostic errors. There are over one hundred such biases that can impair our thinking. These biases are inevitable, but through debias-ing and mitigation strategies, we may be able to temper their effect.Mitigating cognitive bias involves a change in behavior and thinking. The first step is to recog-nize and accept the facts that thinking is prone to bias and that everyone makes diagnostic errors. When a clinician recognizes that a cognitive bias is impairing their diagnostic process, they can adjust their thinking to mitigate the effect of that bias on their reasoning. However, recognition of the effect that these cognitive biases have on the quality of our diagnostic thinking is influenced by other factors and conditions in which we oper-ate. These factors include the environment in which we function, the people and patients with whom we work, and our emotional state and physical well-being [28]. Sometimes it takes a string of serious safety events due to diagnostic errors or the publication of a major report such as Improving Diagnosis in Health Care to foster awareness and acceptance of our error-prone diagnostic processes. Unfortunately, some cogni-tive biases are more resistant to mitigation, and a more structured mitigation strategy is required. (Key Points Box 16.5)Croskerry [28] describes several strategies that may mitigate cognitive bias. The clinician must obtain a thorough history and physical and must develop a comprehensive differential diag-nosis that is appropriate for the clinical presenta-tion. These two seemingly simple strategies alone may mitigate several cognitive biases, including the unpacking principle and the ascertainment, anchoring, premature closure, availability, and representative biases. Unfortunately, in the rush to see an increasing number of patients, in our ment, where an extensive workup includ-ing complete blood count, chemistry panel, hepatic function panel, and inflammatory markers are unremarkable. An acute abdominal X-ray series is obtained and unremarkable. The only pertinent positive test result is a urine drug screen, which is positive for marijuana. At that time, the emergency department provider takes a more directed and detailed drug abuse his-tory and discovers that the patient has been chronically using marijuana several times a day for the past 12 months. Upon admis-sion to the pediatric general medicine ser-vice for intravenous fluid resuscitation, it was discovered that warm showers relieved her symptoms, and her family confirmed her history of taking numerous hot showers every day over the last few months. Many months after her initial presentation for epigastric pain and vomiting, Jessica is cor-rectly diagnosed with cannabinoid hyper-emesis syndrome. She established care with the Psychology Department, and symptoms resolved after cessation of mari-juana use.Key Points Box 16.5 Mitigating Cognitive BiasTo lessen the frequency of cognitive bias, individuals must first recognize the exis-tence and prevalence of these lapses in thinking. Different strategies to mitigate cognitive bias have been identified, includ-ing processes such as a diagnostic time-out, formulating a differential diagnosis, check-lists, etc.

275effort to be overly efficient with documentation and coding, and in our frustration with EHRs, we may cut corners when reviewing a case and developing a differential diagnosis. For hospital-ized patients, we think a diagnostic time-out can help enhance the breadth and quality of differen-tial diagnoses. This time-out is utilized during the initial inpatient assessment, at which time the patient care team is asked to address three ques-tions: (1) What do we think the patient has? (2) What else might the patient have? (3) What diag-nosis can we not afford to miss? The diagnostic time-out also addresses two other mitigating strategies identified by Croskerry, Rule Out the Worst Case Scenario (ROWS) and Until Proven Otherwise (UPO).Other mitigating strategies include a checklist as described by John Ely et al. [29]. A checklist for a variety of chief complaints may mitigate anchoring bias and availability bias and mini-mize memory failures. Mnemonics to aid the development of differential diagnoses are often learned in medical school, but practicing clini-cians may abandon these tools in the busy clini-cal environment. Table 16.2 demonstrates several inclusive mnemonics helpful for a variety of con-ditions. Each of the mnemonics are derived from a list of clinical systems from which a clinician can develop a differential diagnosis consistent with a chief complaint. The first letter of the mne-monic corresponds to a clinical system to consider. Measuring Diagnostic ErrorPeter Drucker (1909–2005), a business manage-ment consultant, coined the axiom among quality improvement experts “if you can’t measure it, you can’t improve it.” Increasing awareness of diagnostic error and the cognitive biases that con-tribute to their occurrence are important; how-ever, in order to act on that increased awareness, i.e., in order to decrease the frequency of diag-nostic errors, diagnostic errors first need to be made measurable.To measure diagnostic errors, detailed chart reviews can be conducted. If the NASEM report is accurate, there are multiple diagnostic errors occurring daily, and they should not be hard to find. In reality, not all diagnostic errors result in patient harm, and patients’ symptoms often resolve despite a diagnostic error. Detailed chart reviews can be time-consuming and involve mul-tiple practitioners to validate the diagnostic error identified. A more practical approach to measur-ing harmful but potentially preventable diagnos-tic errors may be the use of a diagnostic error index [30]. A mature patient safety program may identify diagnostic errors from established insti-tutional sources. For example, quality improve-ment staff can monitor diagnostic errors from reliable sources such as autopsy data, morbidity and mortality reports, adverse event reports, root cause analysis investigations, and medical record triggers. We have used the diagnostic error index Table 16.2 Differential diagnosis mnemonicsVindicate psychologyVictimsVitamin DVin di catem-pVascularVascularVascularVascularInfectiousInfectiousInfectiousInfectiousNeoplasticCongenitalTraumaNeoplasticDrugTumor/TraumaAutoimmune/inflammatoryDegenerativeInflammatoryImmunologicMetabolicImmune/IntoxicationCongenitalMetabolicInherited/congenitalCongenitalAutoimmune/AllergicSeizures/pSychNeoplasmAutoimmune/AllergyTraumaticDrug toxicityTraumaEndocrineEndocrinePsychologyMetabolicPsychological

276as a measurement tool and have shown a decrease in significant patient harm events. Culture of Safety and TransparencyOver the last 20 years, the importance of diag-nostic error recognition within the patient safety culture has come to the forefront. In its purest form, patient safety culture was conceptually defined in the 1999 NASEM report To Err Is Human as an environment that prioritizes reduc-tion of patient harm and improvement in clinical outcomes [1]. For a complex and sophisticated system such as a healthcare organization, a safety culture entails the attitudes, beliefs, and patterns of behaviors of the group and its con-stituents that relate to the organization’s commit-ment to safety [31]. These definitions highlight the fact that safety culture is an active and dynamic notion that requires the engagement of both individuals and leaders.Before an organization can improve safety, its constituents must have a work environment in which there is not a fear of retribution. A trans-parent and open culture directed toward safety is essential to healthcare and is a necessary compo-nent for the reduction of diagnostic errors and preventable patient harm. Creating this psycho-logical “safe space” is vital to successful patient safety cultures but can be very difficult to achieve. Many organizations, including the Institute for Healthcare Improvement (IHI), have long recog-nized the importance of a culture of safety in healthcare. To assist institutions who are seeking to establish or solidify their culture of safety, the IHI outlined six specific domains to guide safety endeavors [32]: 1. Establish a compelling vision for safety 2. Value trust, respect, and inclusion 3. Select, develop, and engage your board 4. Prioritize safety in selection and development of leaders 5. Lead and reward a just culture 6. Establish organizational behavior expectationsWhether utilizing these six principles or other similar concepts, healthcare organizations must first have a foundational basis which supports a positive safety culture. It is from this open and transparent safety culture that organizations can address diagnostic error and eliminate prevent-able harm occurring in their healthcare system. SummaryIn this chapter, we provide background regarding the problem of diagnostic error. We describe the diagnostic process and the types of diagnostic thinking involved. We present a case of a diag-nostic error that involved several cognitive biases and discuss the more common biases. We then perform a diagnostic autopsy and demonstrate how cognitive biases likely led multiple physi-cians to the same erroneous diagnosis. Finally, we suggest ways that diagnostic errors can be measured and that the associated cognitive biases can be mitigated.Editors’ CommentsMedical diagnosis has long been looked at as an art which was presumed to have been mastered during a clinician’s training years and is the very basis of the business of and presumed need for expansive healthcare delivery systems. Over the past two decades, the vulnerability and pitfalls of these systems have been exposed by multi-ple studies, most mentioned in the bibliog-raphy of this chapter. Our healthcare systems have become more complex and lack the framework for sustained improve-ment, accountability, communications, and other qualities seen in high functioning and high reliability organizations.Diagnostic errors appear to be focused on individual errors when they actually often reflect system issues that create situa-tions that place healthcare team members

277 Chapter Review Questions 1. What are the two characteristics that define a diagnostic error?Answer: A diagnostic error is defined as a failure to either establish a timely and accu-rate explanation of a patient’s health problem or a failure to communicate that diagnosis to the patient. 2. What is a cognitive bias?Answer: Cognitive bias refers to a cogni-tive strategy or mental shortcut that fails, lead-ing an individual to deviate from the standard norm of decision-making based on the way information or a situation is presented. 3. True or False: System 2 thinking involves the quick, autonomous way of thinking, which often happens below our level of consciousness.Answer: False 4. What are the three core elements to the diag-nostic process, as outlined by the conceptual model of the diagnostic process?Answer: The three core elements include gathering of information, information integra-tion and interpretation, and formulation of a working diagnosis. 5. A primary care doctor evaluates a 64-year-old female complaining of chest pain. Her symp-toms are very similar to those of another patient who was seen in the office yesterday with pneumonia, so the physician prescribes an antibiotic and discharges today’s patient home with a diagnosis of pneumonia and no further testing. This event may be an example of which type of cognitive bias?and our patients at risk. The story of The Tortoise and the Hare [33] comes to mind when reading about fast (type 1) and slow (type 2) thinking. Time pressures due to financial constraints, staffing shortages, changing clinical acuity, and seasonality force providers to see more patients rap-idly, a situation that does not foster slow and deliberate thinking. There are times when we want the hare to win the race such as during a resuscitation, but the tortoise needs to be in the room. If we want tortoise- type thinking to dominate daily medical activities, our systems need to be better designed and more efficient; problems and errors need to be brought to the surface quickly; organizational cultures need to be aligned for continuous improvement, trans-parency, accountability and change; and finally, regulations and financial incentives need to be realigned to promote type 2 thinking. We need to be aware that clinical pathways, while necessary for standardiza-tion and efficiency, can also foster cogni-tive biases.Diagnostic errors and their associated cognitive biases represent the fallibility of medicine. With their acknowledgment, our patients, families, and team members ask how we are redesigning our systems to improve detection and reduce diagnostic error occurrence. Unfortunately, our sys-tems with the many mergers of healthcare entities are becoming more complex, not less. We need to consider educating our students, active workforce, patients, and their families about the nature of such errors and biases and how we can together reduce their incidence. We need to under-stand the diagnostic process and the meth-ods by which we can overcome its pitfalls. Diagnostic errors represent one of the big-gest challenges for patient safety, espe-cially with the difficulty surrounding their measurement. This field will likely be receiving much more attention in the upcoming years. In 2009 David Newman- Toker and Peter Provonost [34] once elo-quently stated that: “as with most scientific advances, this will be a long and arduous journey, but the next frontier for patient safety is in plain view.” Despite the passage of 10 years, the challenges of diagnosis and critical thinking have not changed. We will need to work harder and smarter to assuage the harm caused by diagnostic errors.

278 A. Bandwagon effect B. Overconfidence bias C. Visceral bias D. Availability bias E. Selection biasAnswer: D. Availability bias 6. True or False: Marijuana is the second most commonly used substance by adolescents.Answer: True 7. What strategies can be implemented to miti-gate cognitive bias and reduce the potential for diagnostic error?Answer: Cognitive biases are an inevitable part of human decision-making. While these biases may impair our judgment and lead to diagnostic error, several strategies have been outlined to help individuals minimize their impact. These strategies include but are not limited to the creation of checklists, develop-ment of a differential diagnosis, utilization of mnemonics, and use of a deliberate pause in thinking, called a diagnostic time-out. 8. Describe concepts that have been outlined to help organizations, leaders, and individuals establish a culture of safety at their institution.Answer: A transparent and open safety cul-ture is imperative to reduce diagnostic error and preventable harm in healthcare. Institutions can work toward establishing this culture by following the six domains outlined by the Institute for Healthcare Improvement. These core domains include objectives such as establishing a compelling vision for safety, actively engaging the administrative board, prioritizing safety in the development and selection of leaders, and placing a value on trust, respect, and inclusion.References 1. Kohn LT, Corrigan J, Donaldson MS. To err is human: building a safer health system. Washington, DC: National Academy Press; 2000. xxi, 287p. 2. Institute of Medicine (U.S.). Committee on Quality of Health Care in America. Crossing the quality chasm: a new health system for the 21st century. Washington, DC: National Academy Press; 2001. xx, 337p. 3. Balogh E, Miller BT, Ball J, Institute of Medicine (U.S.). Committee on Diagnostic Error in Health Care. Improving diagnosis in health care. Washington, DC: The National Academies Press; 2015. xxvii, 444p. 4. Kachalia A, Gandhi TK, Puopolo AL, Yoon C, Thomas EJ, Griffey R, et al. Missed and delayed diagnoses in the emergency department: a study of closed malprac-tice claims from 4 liability insurers. Ann Emerg Med. 2007;49(2):196–205. 5. Gupta A, Harrod M, Quinn M, Manojlovich M, Fowler KE, Singh H, et al. Mind the overlap: how system problems contribute to cognitive failure and diagnostic errors. Diagnosis (Berl). 2018;5(3):151–6. 6. Graber ML, Kissam S, Payne VL, Meyer AN, Sorensen A, Lenfestey N, et al. Cognitive interven-tions to reduce diagnostic error: a narrative review. BMJ Qual Saf. 2012;21(7):535–57. 7. Kahneman D. Thinking, fast and slow. 1st ed. New York: Farrar, Straus and Giroux; 2011. 499p. 8. Norman GR, Monteiro SD, Sherbino J, Ilgen JS, Schmidt HG, Mamede S. The causes of errors in clinical reasoning: cognitive biases, knowledge deficits, and dual process thinking. Acad Med. 2017;92(1):23–30. 9. Evans JS, Stanovich KE. Dual-process theories of higher cognition: advancing the debate. Perspect Psychol Sci. 2013;8(3):223–41. 10. Redelmeier DA. Cognitive psychology and medical judgment: some downfalls of studying pitfalls. Med Decis Mak. 1991;11(3):169–70. 11. Meyer AND, Singh H. The path to diagnostic excel-lence includes feedback to calibrate how clini-cians think. JAMA. 2019; https://doi.org/10.1001/jama.2019.0113. [Epub ahead of print]. 12. Royce CS, Hayes MM, Schwartzstein RM. Teaching critical thinking: a case for instruction in cognitive biases to reduce diagnostic errors and improve patient safety. Acad Med. 2019;94(2):187–94. 13. Schiff GD, Puopolo AL, Huben-Kearney A, Yu W, Keohane C, McDonough P, et al. Primary care closed claims experience of Massachusetts malpractice insurers. JAMA Intern Med. 2013;173(22):2063–8. 14. Eva KW. What every teacher needs to know about clinical reasoning. Med Educ. 2005;39(1):98–106. 15. Sherbino J, Kulasegaram K, Howey E, Norman G. Ineffectiveness of cognitive forcing strategies to reduce biases in diagnostic reasoning: a controlled trial. CJEM. 2014;16(1):34–40. 16. Sherbino J, Yip S, Dore KL, Siu E, Norman GR. The effectiveness of cognitive forcing strategies to decrease diagnostic error: an exploratory study. Teach Learn Med. 2011;23(1):78–84. 17. Norman G, Sherbino J, Dore K, Wood T, Young M, Gaissmaier W, et al. The etiology of diagnostic errors: a controlled trial of system 1 versus system 2 reason-ing. Acad Med. 2014;89(2):277–84. 18. Krynski TR, Tenenbaum JB. The role of causality in judgment under uncertainty. J Exp Psychol Gen. 2007;136(3):430–50.

279 19. Mamede S, van Gog T, van den Berge K, Rikers RM, van Saase JL, van Guldener C, et al. Effect of avail-ability bias and reflective reasoning on diagnostic accuracy among internal medicine residents. JAMA. 2010;304(11):1198–203. 20. Naude JM, Burch VC. Checklist of cognitive contribu-tions to diagnostic errors: a tool for clinician- educators. Afr J Health Professions Educ. 2018;10(3):153–8. 21. Tversky A, Kahneman D. Judgment under uncertainty: heuristics and biases. Science. 1974;185(4157):1124–31. 22. Nelson SP, Chen EH, Syniar GM, Christoffel KK. Prevalence of symptoms of gastroesophageal reflux during childhood: a pediatric practice-based survey. Pediatric Practice Research Group. Arch Pediatr Adolesc Med. 2000;154(2):150–4. 23. Nelson A. Unequal treatment: confronting racial and ethnic disparities in health care. J Natl Med Assoc. 2002;94(8):666–8. 24. Reason J. Human error: models and management. BMJ. 2000;320(7237):768–70. 25. Miller JB, Walsh M, Patel PA, Rogan M, Arnold C, Maloney M, et al. Pediatric cannabinoid hyperemesis: two cases. Pediatr Emerg Care. 2010;26(12):919–20. 26. Nicolson SE, Denysenko L, Mulcare JL, Vito JP, Chabon B. Cannabinoid hyperemesis syndrome: a case series and review of previous reports. Psychosomatics. 2012;53(3):212–9. 27. Simonetto DA, Oxentenko AS, Herman ML, Szostek JH. Cannabinoid hyperemesis: a case series of 98 patients. Mayo Clin Proc. 2012;87(2):114–9. 28. Croskerry P, Cosby K, Graber ML, Singh, H. Diagnosis: interpreting the shadows. 1 ed.: CRC Press. Boca Raton, Fl; 2017. 370p. 29. Ely JW, Graber ML, Croskerry P. Checklists to reduce diagnostic errors. Acad Med. 2011;86(3):307–13. 30. McClead R, Kasick R, Perry M, Melvin J, Kersey K, Kambo M, et al. IHI ID 09 A quality improve-ment (QI) initiative to decrease diagnostic errors. BMJ Open Qual. 2018;7 https://doi.org/10.1136/ihisciabs.9. 31. National Patient Safety Foundation. Free from Harm: Accelerating Patient Safety Improvement Fifteen Years after To Err is Human. Boston, MA: National Patient Safety Foundation; 2015. Accessed November 26, 2020. Available from: http://www.ihi.org/resources/pages/publications/Free-From-Harm-Accelerating-Patient-Safety-Improvement.aspx. 32. Leading a culture of safety: a blueprint for success. Boston: American College of Healthcare Executives and Institute for Healthcare Improvement; 2017. Accessed November 26, 2020. Available from: http://www.ihi.org/resources/pages/publications/Leading-A-Culture-of-Safety-A-Blueprint-for-Success.aspx. 33. Tortoise and the hare in the classic treasury of aesop’s fables (illustrated by Don Daily). Philadelphia: Running Press; 1999. p. 43–46. 34. Newman-Toker DE, Pronovost PJ. Diagnostic errors – the next frontier for patient safety. JAMA. 2009;301(10):1060–2.

281© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 R. K. Shah, S. A. Godambe (eds.), Patient Safety and Quality Improvement in Healthcare, https://doi.org/10.1007/978-3-030-55829-1_17An Improvement Operating System: A Case for a Digital Infrastructure for Continuous ImprovementDaniel Baily and Kapil Raj Nair Opening QuestionIn 2012, CMS reported on 43 quality measures in their National Impact report [1]. This past year (2018) CMS reported on 762 different measures [2]. That’s an average of ten additional metrics per month, for 6 straight years – just from CMS, let alone the requirements from private payors and watchdogs! As the list of requirements grows and resources to spend on meeting these require-ments shrink, health system leaders are con-stantly challenged to maintain priority and focus. Even with the best intentions and commitment, D. Baily · K. R. Nair (*) Beterra Health, Newnan, GA, USAe-mail: [email protected]; [email protected] Objectives• Share common quality and safety improvement challenges in healthcare systems• Define an improvement operating sys-tem, its aspects, and how it can drive safety and quality improvement• Use patient safety culture improvement case as an exemplar, to demonstrate how an improvement operating system gets operationalized• Share a vision for how an improvement operating system should evolve over time to be used as a broad-spectrum operational infrastructureVignette 17.1A hundred-bed local community hospital has struggled with sustaining patient safety culture improvement. The hospital board is also aware of the sustainability challenges and has asked the VP of Quality to address and build patient safety culture as a founda-tional tool for improving quality and safety. The organization has achieved safety cul-ture improvement in the past but is unable to connect microculture level actions to outcomes. As part of setting the strategy for improvement, the leadership distilled their case into a probing question: “How do organizations leverage the current momen-tum in safety culture improvement while improving or sustaining all other quality and safety metrics they are responsible for?”

282safety and quality leaders have found the ability to create and maintain the needed focus on improvement a significant challenge with little to no capability of systematization, sustainability, or evolution of improvement efforts at scale. In such an era of intense regulatory and perfor-mance requirements combined with dynamic change, how can quality and safety leaders fathom or even envision sustainable improvement?This chapter shares an approach that helps organizations direct improvement programming in a different way. What if, we can do more with less? IntroductionIt’s no secret that healthcare is facing pressure from many sides. In an era of value-based pur-chasing, system consolidation, and changing delivery models, healthcare organizations are mandated to reduce all cause patient harm, increase patient satisfaction scores, engage a changing workforce, and comply with an ever- growing number of reporting requirements from regulators and public advocacy groups. As the list of requirements grows, our resources to spend on meeting these requirements are shrinking. Reimbursements are shrinking – often made even more painful by penalties due to underperform-ing on quality, safety, and satisfaction metrics. We also face severe and growing workforce shortages that are undermining any improvement initiative, leading to disengagement and burnout, and chipping away at our workforce’s most valu-able resource: time. Through all of this, our patients continue to get harmed at unacceptable rates.The organizational stakes are high!!We have to become nimble and resourceful, or yield to disruption, as the healthcare market [3] is evolving at an ever-faster pace. We have to find ways to significantly reduce workforce burnout as we are constantly losing qualified healthcare providers [4]. We have to dramatically strengthen patient safety efforts as we continue to harm our patients at unacceptable rates.Unfortunately, we face some significant inter-nal gaps preventing most organizations from suc-cessfully navigating these tumultuous waters and creating the needed improvement. Common Internal Gaps Leadership GapAlthough highly educated and well-intentioned leaders – especially at the mid-level – struggle to execute on dozens of improvement targets, the successes that do occur are usually the result of heroic efforts that rarely spread across the organi-zations. Most critically, on improvement initia-tives, the perspective of leaders of all levels usually varies dramatically from the frontline caregivers who must ultimately make the needed change. This creates a misunderstanding of enablers and barriers of improvement ultimately stalling improvement actions. A qualifying example would be Agency for Healthcare Research and Quality (AHRQ) Surveys on Patient Safety Culture (SOPS) [5] which is used to measure perceptions of healthcare employees, including frontline staff and organizational lead-ers, on patient safety and quality of care. The SOPS national benchmarks consistently show organizational leaders as having a higher percep-tion of quality and safety than frontline staff. This could be triggered by leaders lacking operational sensitivity. In their famous book, Managing the Unexpected, the authors Karl E. Weick and Kathleen Sutcliffe refer to “sensitivity towards operations” [6] as a core principle to create sus-tained performance. Information GapHealthcare leaders lack real-time, actionable information about the improvement efforts going on across their organization. This problem is exacerbated by lack of strong clinical analytics and compounded by the nonexistence of a digital infrastructure to consistently track improvement actions and progress. This leads to an overwhelm-

283ing number of improvement initiatives whose success can only be evaluated retrospectively. These retrospective analyses are more opinion- based than data driven. This is primarily driven by lack of continuous improvement activity data which typically helps organizations to know what worked and what did not work and why certain activities or tasks were dropped. Therefore, we fail to learn iteratively, or course correct at the rate needed to be successful. Improvement SilosMost improvement efforts are siloed and are addressed as work streams driven by data sets. The experts who are leading these improvement efforts each own these data streams indepen-dently of each other, forcing frontline leaders to operate on the improvement initiatives indepen-dently. For example, a hospital system will have patient experience measures and improvement experts similar to how they have quality and safety improvement experts. This approach sig-nificantly dampens systematic improvement (Fig. 17.1).All of the above lead to a gap in execution – we ultimately fail to see the rates of improvement we both expect and need.SafetyCultureExpertsToolsEmployeeEngagementExpertsClinical EnvironmentPatientExperienceExpertsToolsToolsHACExpertsToolsFig. 17.1 Various siloed data streams applied in a clinical environmentVignette 17.2In context of their probing question, the organization decided to dig further deep into understanding the root causes of their sustainability challenge.They lacked local ownership of improve-ment efforts. They had significant variabil-ity in performance. While the top performers were lifting the overall results above benchmarks, more than half of their units statistically trailed the benchmarks concerning patient safety culture. This problem was compounded by the fact that about one in five units/outpatient areas had new managers. The organization’s improve-ment efforts did not leverage cross cutting strategies such as safety culture to their advantage.They also experienced a significant dif-ference between leadership and frontline’s

284 A Way ForwardTo fix these above challenges, healthcare leaders should acknowledge that their organizations sim-ply lack an improvement operating system. One could think of an improvement operating system as a combinatory set of core values, robust man-agement processes, and a proactive prioritization method through which the organization itera-tively accomplishes its targets while building a strong and vibrant team-based culture. Every organization is unique and therefore has the potential to have its own unique operating sys-tem. But there are certain common attributes especially when considering the core values (Key Points Box 17.1). What Are the Core Values of an Operating System for Improvement?We describe an improvement operating system that will ensure microculture (unit, department, or a clinic)-based improvement, visibility into improvement actions, and leadership engage-ment. These three components will increase vis-ibility, scalability, and sustainability of improvement efforts.• Visibility. An improvement operating system will ensure leaders have clarity around the broad state of improvement efforts across the microculture-based teams, both inpatient and Difference = (Leadership) – (Frontline)Nonpunitive Response to ErrorCommunication OpennessFeedback and Communication About ErrorHospital Management Support for Patient SafetyTeamwork Within UnitsFrequency of Events ReportedTeamwork Across UnitsOverall Perceptions of Patient SafetySupervisor / Manager Expectations and Actions Promoting Patient…StaffingOrganizational Learning / Continuous ImprovementHandoffs and TransitionsLeadership Perceptions Higher than FrontlineDifference between Leadership and Frontline ScoresDifference between Leadership and Frontline(Leadership – Frontline) Benchmark26%21%19%16%10%10%10%9%9%5%3%-17%-20%-15%-10%-5%0%5%10%15%20%25%30%20%16%15%16%8%8%11%13%32%8%12%4%Fig. 17.2 Difference in leadership to frontline perceptions measurementview of safety culture. This gap is depicted using patient safety culture measures through a standardized measurement instrument such as AHRQ Surveys on Patient Safety Culture (SOPS). In the graph shown in Fig. 17.2, we consider the scores of leaderships vs. frontline for the same composite (Fig. 17.2).Key Point Box 17.1Operating System (noun): software that controls the operation of a computer and directs the processing of programs as by assigning storage space in memory and controlling input and output functions [7].First known use: 1961[Source: Merriam-Webster Dictionary]

285outpatient, working on improvement initia-tives. In contrast to typical, stagnant action plans, improvement team leaders will regu-larly provide simple progress reports that include activities, barriers, and enablers towards improvement. An improvement operating system should also aggregate the activity and performance data to ensure that senior leaders consistently have the pulse of improvement rather than waiting for long improvement cycle feedback. This aggrega-tion of data should also lay the foundation for connecting the improvement actions to out-comes data to create visibility into the effec-tiveness of various efforts.• Scalability. The improvement operating sys-tem will enable leaders to scale improvement efforts without adding additional quality and safety personnel. Furthermore, these efforts will provide leaders with the ability to identify internal best practices and scale those prac-tices across the facility and system. This approach ensures appropriate leaders and champions are given relevant information for the teams they support and the barriers they face in order to work towards removing them, regardless of the number of teams or improve-ment aims.• Sustainability. Within an improvement operat-ing system, improvement plans and priorities are based on leadership’s vision. These are also informed by the current data. This ensures organic, sustainable improvement rather than a reliance on external expertise and resources. This approach also detects when barriers are preventing local teams from making progress on local goals, enabling leaders to intervene far sooner than waiting for lagging metrics to change. As the improvement teams grow their improvement capabilities, they will increas-ingly rely on this operating framework to pro-vide insights and accountability towards their own plans.• Integration. An improvement operating sys-tem should integrate the improvement initia-tives, performance data, and toolkits. This will mitigate the “initiative fatigue” that frontline leaders face. The integrated method could also accommodate new feedback and support mechanisms without draining the organiza-tional capacity. Operationalizing an Improvement Operating SystemPutting an improvement system into action requires careful thought, attention, and execu-tion. Quality and safety improvement is not just a technical business of best practices and advanced data analytics to manage process and quality improvement (PI and QI) projects. It is also an adaptive challenge [8]. It requires behavior and attitude changes at all levels of the organization. Hence an operational plan that does not maintain significant interest into the foundational aspect of organizational culture improvement will not suc-ceed in putting an improvement operating system in place. Hence, we propose a four-staged action framework for iterative and continuous improve-ment for organizations starting with patient safety culture. The four-staged action framework is sim-ilar to the Deming’s cycle (also known PDSA cycle) introduced by Dr. Edwards Deming. This iterative cycle forces a rhythm into an organiza-tion which is really important as healthcare orga-nizations constantly struggle to prioritize (Fig. 17.3).The critical initial stage of the safety culture improvement cycle, PLAN, requires creating clar-ity of purpose and shared language across the core improvement team and executives for systematic improvement of patient safety culture. This stage requires leaders to articulate the strategy, support infrastructure, and timelines for their improve-ment journey. Organizational leaders should clearly understand the current improvement ini-tiative inventory and the organizational approach towards improvement. This will not only paint the picture of the current state and its efficacy but also will provide significant input towards creating the PlanOrganizeUseEvolveFig. 17.3 12-month improvement cycle phases

286vision towards advancing the organizational strat-egy regarding patient safety culture. Leaders should consider this stage as a forming stage [9] and develop messaging that clarifies purpose, timelines, and expectations while creating the alignment among themselves. Most importantly, they should ensure that the latest safety culture data is rolled out to all levels of the organization including frontline staff. One could use a stan-dardized safety culture measurement tool such as prior mentioned AHRQ Surveys on Patient Safety Culture (SOPS) Measurement and associated best practices on distributing results.The objective of the ORGANIZE phase is to prepare for systematic improvement of patient safety culture. This phase of the improvement cycle involves organizing the improvement infra-structure as well as establishing the improvement plans that align with other organizational strat-egy, metrics, and plans. Organizational leaders should create clarity and alignment towards a shared success. Accomplishing a shared success requires a significant amount of attention, focus, and cascading communication.A common set of core activities during this normative stage should include:• Establishing an improvement infrastructure, both globally and locally. Leaders could use models such as Comprehensive Unit-based Safety Program (CUSP) [10] for helping clini-cal teams while establishing an executive engagement network within the organization.• Gathering frontline teams and local leaders input on the set improvement actions is extremely important to create not only buy-in but also continued engagement in the process. Existing organizational engagement modali-ties that could trigger appreciative inquiry [11] such as structured discussions, focus groups, structured leadership rounds, or table top discussions could be proven useful for get-ting the needed bidirectional feedback.• Aligning the safety culture action plans with other organizational strategies and metrics by connecting with the “why” behind our actions can significantly improve the odds of success. It provides the existential connection between organizational core values and the work, increasing employee engagement and satis-faction which is shown to have direct correla-tion to patient safety.• Cascading the targeted and brief action plans throughout the organization creates not only clarity on what needs to be done but also lays the foundation for accountability on the plans.Vignette 17.3In order for them to be successful, the lead-ers of all levels jointly decided a course of action with the following high-level core objectives that aligned with their 5-year plan to be able to provide the best care and health outcomes to the population they served.The leadership team through thoughtful deliberation realized that they have to first tighten their patient care operating model. To achieve this, they have to improve patient experience, patient safety, care quality, and staff engagement at all levels. The leadership team decided to create an infrastructure that is feedback rich so that they can lay the foundation for engaging their mid-level leaders and frontline staff. With commitment from all executive lead-ers, they decide to embark on a safety cul-ture improvement journey. To that end, they articulated the following core objectives: 1. Maintain forward momentum with safety culture improvement 2. Improve and cultivate local ownership of improvement 3. Reduce the leadership gapVignette 17.4The organization established a core team led by the Chief Operating Officer (COO). In this organization, the VP of Quality and

287The most important phase is the USE phase. The purpose of the USE phase is to materially improve the culture of safety. The majority of healthcare organizations fail to see significant improvement in patient safety culture, in part because the prior phases were poorly imple-mented and in part because execution of improvement actions is weak. The few organi-zations that do execute and monitor action plans successfully are able to see significant improvement in patient safety, quality, and staff engagement.This action-focused portion of the improve-ment cycle requires leaders to “walk the talk” by embedding themselves with the improvement teams through rounding, adopting, and coaching an improvement team. This will help maintain high levels of accountability throughout the orga-nization as leaders will be able to monitor prog-ress, remove barriers, encourage enablers, and share internal best practices throughout the improvement cycle. It also helps leaders under-stand the volume of activity, creating an opera-tionally sensitive environment which is really important to create focus and elevate priority. In order to be successful, local leaders should be equipped with executable safety science knowl-edge, skills, and attitudes to lead change.The main purpose of the EVOLVE phase is to course correct or advance current work based on three sets of data (a) performance data based on improvement actions, (b) feedback data on enablers and barriers, and (c) actual outcomes data.While considering safety culture improve-ment, unit-based performance data should include insights into actions. A shared under-standing of organizational performance on the improvement cycle can be created. Momentum can be measured using simple metrics around goal completions such as percentage of goals completed, percentage of goals and actions dropped, etc. Engagement can be measured using number of teams completing improvement actions at various completion rates. Feedback data can be demonstrated by describing top three barriers and enablers, etc. A Kurt Lewin’s force field analysis [12] with appropriate ranking order could be used here to create a consumable view of the data. A sample is given in the case vignette.CNO directly reports to the COO. The core team also included the Chief Medical Officer, a data analyst, and select frontline leaders. This team conducted an initiative inventory along with a key stake holder analysis finally forming an improvement roadmap. The core team also spent signifi-cant amount of energy in crafting the right messaging to the frontline leaders and staff.The core team also articulated the role of the senior leadership team:• Lead improvement effort (quarterly) –Initial input on goals and processes –Quarterly meetings to review overall progress –Sharing experiences and feedback from rounding• Serve as improvement team sponsors (ongoing) –Direct coaching/rounding on improvement teams – (monthly) –Review Improvement Team ProgressThe core team assigned themselves to the following:• Operationalize the improvement operat-ing system (monthly) –Manage and adjust improvement roadmap –Monitor overall progress• Support Improvement Teams (ongoing) –Provide tools and resources (i.e., be the experts)Vignette 17.5As the microcultures started working through their improvement cycle. The microculture leaders started to identity the enablers and barriers which helped the leadership team tremendously.A sample of an organizational force field diagram is shown in Fig. 17.4 based on their feedback from the progress reports (Fig. 17.4).

288Ultimately, as the outcome data becomes available through safety culture measurement at the end of the improvement cycle, organizations will be able to connect actions to outcomes. This marriage of outcomes data with performance and feedback data provides a data-driven contextual report rich with output that should set the new improvement cycles directionality. Though this evolutionary process may be sound in analytical rigor, it should be grounded in process celebra-tions of successes and course corrections on missed opportunities, all the while connecting the work all back to the organization’s vision. ConsiderationsImproving organizational performance is no easy feat. It requires a rare multifactorial combination of vision, leadership, capacity, capability, and a willingness to change. As organizations plan their improvement journey, we recommend a few areas to keep in mind.Driving Forces (58)Barriers (70)Staff/Provider Engagement (11)Mindset/Resistance to Change (21)Competing Priorities (18)Lack of Resources/Staff (10)Poor Execution/Coordination (10)Lack of Leadership Support/Accountability (6)Future Uncertainty (3)Other(2)Leadership Support/Accountability (10)lmprovement Goals/Commitment (9)Teamwork/Collaboration (9)lmprovement Structure/Roadmap (8)Process Improvement (5)Resources (5)Improve Safety CultureFig. 17.4 Organizational force field diagramVignette 17.6The core team measured performance data on improvement actions.Metrics such as goals set versus goals met, number of teams updating progress reports on a regular cadence, percentage of local goals sources to overall, etc. helped the organization to easily communicate organizational performance to senior lead-ers as well as the board.By the end of a 12-month improvement cycle, the organization was able to improve overall safety culture and local ownership of improvement actions. They were unable to close the leadership gap as they desired. In fact, the gap increased.The organization was most excited not by the safety culture improvement but by the power of the improvement infrastruc-ture that they built. They are using the same operational approach, information system, and leadership engagement model to fur-ther advance patient safety culture and improve patient experience in their current improvement cycle.

289 1. Maturity levels: It is imperative to consider the maturity level of an organization’s capa-bility for each improvement data set as an organization embarks on operationalizing integrated improvement. After working with myriad hospitals systems across the world, we developed the framework (Fig. 17.5) to con-vey organizational maturity. 2. Broad-spectrum approach: It is important to build and flex the organizational muscle of integrated improvement programming with a couple of data sets before using it as a broad- spectrum approach throughout the organiza-tion (Fig. 17.6). 3. Digital environment: A digital improvement operating system unleashes the power of data science. Digitization of improvement pro-gramming and tracking will help organiza-tions maintain data integrity and data deliverability, freeing up subject matter experts to facilitate improvement rather than spending their time and resources on simply collecting and analyzing data. 4. Feedback-rich environment: A fundamental assumption for a successful long-term deployment of an integrated improvement envi-ronment is the organization’s commitment towards creating a feedback-rich “just culture.”JourneyCrawlWalkRunRole of Subject matter orImprovement ExpertsRole of Senior LeadersRole of Local LeaderAreas of FocusOutcomesAct as expert resources,guides, and consultantsSet strategic direction anddirectly support team basedimprovementLead the local improvementeffortOwnerDriven by strategic focus.Metrics are integrated andlocal (safety, quality,satisfaction, etc.)Year after year of methodicalimprovementCollect and share data,support senior leadershipunderstanding andengagementDevelop org. – wide actionplan and implement “adopt ateam”OwnerLeading local team anditerating approach as neededCascading areas of focusbased on safety culture andother metrics.Pockets of ImprovementDrive the improvementprogram, engage andeducate leadersOwnerBegin rounding on andcommunicating aboutimprovement initiativeFollow simple outlinedstructures for team basedbased action plansOrganization wide focus,some local team baseddifferences (e.g. safetyculture survey)Limited SuccessFig. 17.5 Journey to organizational maturityImprovementCycleTeamsSponsorsInitiativesOutcomesSummaryCycle 1:FoundationCycle 2:EngagementCycle 3:HardwiringLimited to select,organizational teams(usually well formedteams such as nursing toaux units)More teams included,starting to see “ad hoc”improvement teamsAll improvement teamsengage through theimprovement systemSmall group of highlyengaged senior leadersExpanded group ofleaders. Begins to betterreflect natural hierarchyAll leaders have somerole as sponsor/teamleaderFoundation laid forsystematic improvementinfrastructureLimited to a key area ofquality or patient safetyperformanceFocus is on process morethan outcomesExpanded to multiple keyareas {such as patientexperience or employeeengagement)Begin to see consistentconnections betweenactions and outcomesEngagement increasedand lessons applied tobetter configureinfrastructureHardwired improvement infrastructureImprovement Systemused to organize andsimplify expansive list ofinitiativesOngoing connection of actions and outcomesFig. 17.6 Broad spectrum approach to improvement cycle

290 ConclusionThe organizational impact of an improvement operating system can be very positive. With an improvement operating system, organizations have the ability to set a direction and follow through with intended actions. Senior leaders have meaningful input from local leaders and frontline staff about what type of support is needed in order to achieve the stated goals (cas-caded from the organizational vision). Middle managers confidently lead their teams forwards – directly with the support of senior leaders – and generate new ideas and approaches. Frontline staff of these organizations are engaged, they speak the language of high performance, and they work together as highly performing teams. New efforts to improve are not shunned; they are adapted with constant feedback from the front-line. This not only creates a psychologically safe environment for iterative learning but also paves the way for a generative culture.An intentionally crafted, supported, and exe-cuted improvement operating system will ulti-mately help an organization break down silos between various improvement efforts and teams, promoting rapid learning across various levels of the organization and creating a data-driven accountability system. Chapter Review Questions 1. Describe common internal gaps faced by healthcare leaders.Answer: Common internal gaps include:Editors’ CommentsWhat is an improvement operating system? This term is certainly new to the Editors and we believe to many in healthcare. The authors of this innovative chapter are co- founders of a company that creates Improvement Operating Systems for healthcare. We could not think of a more forward-reaching topic that has the poten-tial to transform healthcare improvement science than this chapter. The authors describe the tenets of operating systems and the current state of continuous quality improvement. The logical leap, which admittedly has taken the Editors some sig-nificant time and reflection to fully under-stand and appreciate, is that the authors then use the fundamentals of operating sys-tems and apply these to healthcare. Once the reader understands how to bring together a traditional operating system with the values of healthcare into a continuous quality improvement glide-path, the next step is to make it happen. They adroitly have created a four-step pathway that orga-nizations can actually follow prescriptively (Fig. 17.3) to embark upon an improve-ment cycle with an improvement operating system. The steps of plan, organize, use, and evolve are well described in the middle of the chapter. Once the data (i.e., results) are obtained from an improvement operat-ing cycle, the authors use an organizational force field model (Fig. 17.4) to demonstrate who to use the feedback to demonstrate driving forces compared to barriers. As co- founders of a digital improvement science company, the authors bring unique per-spective. The figures that describe the jour-ney to organizational maturity and a broad-spectrum approach to an improve-ment cycle are crucial messages for the reader to understand and potentially use to communicate the value proposition of an improvement operating system within their organization. The Editors suggest the reader pausing to ask themselves, “Where is my organization in their maturity”? Using this as a starting point, the reader can then consider if an improvement operating system is within their scope and possibili-ties. The question should not be “if,” but rather “how soon.” Fortunately, this chapter provides a framework to start and acceler-ate the readers’ organizational journey to an improvement operating system.

291Leadership Gap: A gap in perception of organizational culture around improvement efforts or organizational culture between lead-ership and frontlineInformation Gap: A gap between real-time actionable information on improvement efforts 2. How will an improvement operating system create visibility within an organization?Answer: An improvement operating sys-tem will create visibility through simple prog-ress reports that contain organizational activities as well as drivers of change around these improvement activities. 3. True or False: Solving for patient safety is just a technical challenge.Answer: False. A strong culture of engage-ment and commitment for improvement of patient safety at all levels of the organization is also needed. 4. What are the core value elements of an operat-ing system?Answer: Visibility, scalability, sustainabil-ity, and integration of improvement efforts are the core value elements of an operating system. 5. What are the different stages of a safety cul-ture improvement cycle? How does it relate to Deming’s cycle?Answer: The various stages are Plan, Organize, Use and Evolve. It is a variant of the Deming’s cycle or Plan, Do, Check/Study, Act (PDCA or PDSA) cycle. 6. Provide a set of metrics that you could use to measure engagement of improvement teams during an improvement cycle.Answer: Metrics that could measure team engagement rates based on number or per-centages of teams regularly updating progress reports, percentage of goals that are set at grass root level vs. organizational level, etc. 7. What are certain considerations that you should make while designing an improvement operating system?Answer: Organizational maturity in terms of improvement infrastructure and organiza-tional culture should be considered. Also, when an organization is willing to introduce digital transformation towards improvement, programming needs to be considered as well.References 1. Centers for Medicare & Medicaid Services. 2012. National impact assessment of the Centers of Medicare & Medicaid Services (CMS) quality measures reports: 2012 impact report. https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/QualityMeasures/National-Impact-Assessment-of-the-Centers-for-Medicare-and-Medicaid-Services-CMS-Quality-Measures-Reports.html. Accessed 10 May 2019. 2. Centers for Medicare & Medicaid Services. 2018. National impact assessment of the Centers of Medicare & Medicaid Services (CMS) quality measures reports: 2018 impact assessment report and appendices. https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/QualityMeasures/National-Impact-Assessment-of-the-Centers-for-Medicare-and-Medicaid-Services-CMS-Quality-Measures-Reports.html. Accessed 10 May 2019. 3. Forbes 2018. Bain insights. https://www.forbes.com/sites/baininsights/2018/05/01/riding-the-disruption-wave-in-healthcare/#73beee872846. Accessed 10 May 2019. 4. Medscape. National physician burnout & depres-sion report 2018. https://www.medscape.com/sites/public/lifestyle/2018. AHRQ Patient Safety Network (PSNet). Perspectives on safety: burnout among health professionals and its effect on patient safety. https://psnet.ahrq.gov/perspectives/perspective/190/burnout-among-health-professionals-and-its-effect-on-patient-safety. Accessed 10 May 2019. 5. AHRQ surveys on patient safety culture. Hospital Survey on Patient Safety Culture (HSOPS). https://www.ahrq.gov/sops/surveys/hospital/update/index.html. Corresponding 2018 databases. https://www.ahrq.gov/sops/databases/hospital/index.html. Accessed 15 May 2019. 6. Weick KE, Sutcliffe KM. Managing the unexpected: sustained performance in a complex world. 3rd ed. Hoboken: John Wiley and Sons; 2015. 7. Merriam -Webster’s Collegiate Dictionary. Page 869. 11th edition. Merriam-Webster; 2013. 8. Pronovost PJ. Navigating adaptive challenges in qual-ity improvement. BMJ Qual Saf. https://qualitysafety.bmj.com/content/20/7/560.full. Accessed 31 July 2019. 9. Tuckman B. Tuckman’s stages of group develop-ment. Wikipedia. https://en.wikipedia.org/wiki/Tuckman%27s_stages_of_group_development. Accessed 31 July 2019.

292 10. The Comprehensive Unit-based Safety Program (CUSP). https://www.ahrq.gov/professionals/education/curriculum-tools/cusptoolkit/index.html/. Accessed 15 May 2019. 11. Appreciative inquiry. Wikipedia. https://en.wikipedia.org/wiki/Appreciative_inquiry. Accessed 15 May 2019. 12. Force field analysis. Originating from the seminal works of Kurt Lewin recognized as the “founder of social psychology”. Practical Application of Lewin’s Force Field Analysis. https://www.latestquality.com/lewins-force-field-analysis/. Accessed 15 May 2019.

293© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 R. K. Shah, S. A. Godambe (eds.), Patient Safety and Quality Improvement in Healthcare, https://doi.org/10.1007/978-3-030-55829-1_18Patient Flow in Healthcare: A Key to QualityKaren Murrell IntroductionIn the current healthcare system, inherent delays and frustration have come to be expected. These same delays impact the quality of care through-out the system [1]. In the white paper “Achieving Hospital-Wide Patient Flow,” Pat Rutherford from IHI discusses the impact of hospital flow on quality and patient care [2]. She describes the need for interdependent, interconnected systems to improve patient outcomes but also describes the challenges of such a system. With leadership and specific strategies, smooth patient flow is possible. Imagine the hospital of the future where diseases are identified at their earliest stages in the primary care office, and when evaluation or treatment is needed, it is carried out seamlessly in the hospital without time delays and interruptions between each step. A hospital where specialists are available at a moment’s notice and all care-givers are focused on getting the patient well and home to family. This can be created – but will require a shift in how system-wide operational strategies are designed. In this chapter, principles will be illustrated with actual case studies about journeys to improve patient flow and discuss how system-wide spread of best practices can occur.K. Murrell (*) Performance Improvement, TeamHealth, Knoxville, TN, USAe-mail: [email protected] Objectives• To introduce the concept of “flow” in healthcare as a quality measure• To describe concrete methods to create operational improvement using Lean methodology• To discuss the role of leadership in improving a system and setting a vision to jumpstart any project• To use a case-based approach to illus-trate the role of flow to either impede or improve patient careVignette 18.1In 2010, there was a case that spread across the news in Northern California. It describes the case of a little girl who required multiple amputations after waiting 4 hours to see a physician in the ED. Her father describes his despair as he waited to see a physician with his sick little girl. One cannot know if the outcome would have been different if she had been seen quickly, but this delay could not have helped. At another Northern California hospital in 2007, there was an impending crisis. A per-

294 BackgroundThere is much data that supports the adverse health outcomes associated with poor patient flow [3]. Studies show that ED boarding increases both mortality and patient length of stay once they are admitted – further compounding the problem [4]. On the inpatient side, studies show that for every added patient with heart failure, pneumonia, or heart attack given to an over-worked nurse, odds of readmission increase from 6% to 9%. For the ICU, every patient discharged early because of overcrowding has double the chances of being readmitted to the ICU. While this is a well-known phenomenon, very few hos-pital systems have been able to strategically address solutions to the problem. In the USA, the most common strategy is building additional space for care without looking at waste and poor flow as an etiology for problems [5].Historically, the US healthcare system pay-ment model has not rewarded patient flow. Hospitals depended on hospital admissions for reimbursement, and this translates to the more procedures the better for the hospitals’ financial bottom line. Physicians in many areas have been compensated per patient, leading to multiple con-sultations per admission. The idea that hospital admission equals quality was widespread in the culture of patients. This system resulted in many unnecessary admissions to the hospital, with multiple consultations and procedures that could have been safely delayed but were often per-formed. The new current bundled payment model and payment for quality means hospitals now have a new focus on decreasing length of stay and avoiding readmissions. If designed well, this can improve patient care and encourage interde-pendence across the system from outpatient through the hospital. It also means that now a few smart, strategic, and unique hospitals are consid-fect storm was developing that started in the ED and continued on the inpatient side. There were long delays for patients that often led to quality issues because of poor flow. At the same time, the hospital was seeking trauma center designation, vol-umes were increasing at double digit rates annually, and the county psychiatric unit closed their crisis stabilization unit and half of the inpatient beds. Boarding in the ED was a frequent problem. New leadership in the ED did not know what to do to improve care and patient flow. Leaders in the ED and hospital read about a course in “Lean Healthcare” and attended and began to spread the basic principles throughout the hospital. They methodically set about changing how healthcare was delivered and transformed healthcare delivery. They found that excellent flow dramatically improves quality for patients. These gen-eral principles apply across the health sys-tem both in the USA and around the world. The general flow principles learned can be replicated in any healthcare system but will require process changes and cultural transformation.In 2011, another little girl came into the transformed emergency department with her father. She had a complaint of nausea and vomiting with normal vital signs. Previously, she easily could have waited 4 or 5 hours, but with improved flow, she is immediately seen by a doctor. He decides she is ill, does a spinal tap, and discovers she has meningococcal meningitis – a very time sensitive disease. She is immediately treated and admitted and makes a full recovery. For many years, the ED physician gets updates from her mother about how well she is doing, how she has started school on time, and most importantly she says: “without you my life would have been very different.” This is the impact of flow in a hospital, consistent, reliable care that allows doctors to treat and save patients. This chapter will discuss general principles to improve operations and flow as well as change management and lean principles learned.

295ering flow a top priority. There is hope that this will increase even more in the future. LeadershipWithout top-level hospital leadership support, system-wide change is difficult [6]. The first step in any change management strategy is setting a clear vision. This vision must be simple and clear and able to be articulated by all employees. No change is easy, and without this leadership sup-port and direction, new processes that are devel-oped will quickly regress back to the status quo. Employees must know the “why” behind deci-sions that are made in order to stand behind them. Intermountain Healthcare is an example of one organization where visible leadership has helped to transform the organization. Leaders have set a clear vision that the goal is to “be a model health system by providing extraordinary care and supe-rior service at an affordable cost.” All employees understand this goal, and each department has goals that align with this vision. These goals are discussed at rounds where daily metrics are dis-cussed. Frontline staff can escalate concerns to their manager. There are cascading reports up the chain to a daily report out at the highest leader-ship level, and each employee feels that their concerns are addressed in a timely manner. Successes on the department level are recognized by top leadership. This high-level leadership sup-port is critical for system-wide flow improve-ment. Successful leaders are visible, break down barriers for the team, and create the inspiration necessary for success [7]. Flow PrinciplesIt is well-known that having satisfied, engaged employees leads to optimal healthcare for patients. Job satisfaction in healthcare is related to many factors: optimized workflows, autonomy and the ability to participate in decisions that are made, and excellent communications and the ability to have a voice without risk of repercus-sion. Job burnout in healthcare is rampant, and healthcare workers mention that the hours of data entry required in the era of electronic medical records lead to stress and anxiety. While these factors are well-known, it is also very common in healthcare to have a reactive approach to prob-lems instead of really considering the impact of decisions on workers. When a sentinel event occurs, it is common that new procedures are lay-ered on a process and on workers who are already living at the high end of the utilization curve. (Fig. 18.1)As illustrated by Dr. Chuck Noon in his book The Definitive Guide to Emergency Department Operational Improvement, the curve shows that when any system is utilized over 80%, wait times increase exponentially [8]. There is no doubt that procedures and standard work are necessary to Vignette 18.1 (Continued)Flow improvements at the California Hospital described earlier started in the emergency department. After coming back from Lean Training, a multidisciplinary “flow team” was established in the ED. This team had representatives from physician, nursing, tech, and clerical staff. They did an extensive redesign of the process for low-acuity patients. This project was picked because all care was under the con-trol of the ED. Operational improvements for this cohort of patients immediately decreased “left without being seen” rates and markedly improved patient satisfac-tion. What was most important was that the hospital CEO was very supportive of the work and came to the ED and recognized each of the employees involved and the department. This began the cultural trans-formation and increased support for the next flow projects. Other employees were eager to participate and join further proj-ects. Without this executive level support, sustained change would have been difficult.

296avoid patient harm but should be done thought-fully with consideration of what the new proce-dure does to workers and after a deep dive into the root cause of the problem. It is not uncommon for an unusual and non-preventable patient case to generate new procedures for every other patient in a system. This work may be non-value- added and take away from high-value procedures. When just a few minutes are added on to each patient in a busy system, wait times can increase exponentially. If a new policy or procedure must be incorporated by a busy server, it is critical to look at all steps in the process and either take away another task or redesign the system. Involving the frontline workers in decisions con-tributes to feelings of autonomy and improved patient satisfaction.How does patient flow improve quality for patients and worker satisfaction? There are two critical principles of redesign: (1) do “this hour’s work this hour” and (2) create systems that are better for patients but easier for people doing the work.As stated in the IHI white paper on patient flow, the goal is to give patients “the right care, in the right place, at the right time” [2]. Healthcare is full of faulty systems that are done because of legacy designs. An optimal system allows the patient to seamlessly arrive, tell their story once to the healthcare team, get treatments and testing without delay, and go home or be admitted to the hospital. A system like this promotes quality healthcare and also allows healthcare workers to focus on doing what they like best: caring for patients. On the emergency medicine side, this means an even length of stay for patients across the 24 hours of arrivals. On the hospital side, thinking about a 24/7, 7-day a week hospital is ideal. Obviously, this cannot be attained over-night but should be the goal of anyone looking to improve flow in healthcare.The second principle is creating systems that are better for patients but easier for people doing the work. There are only three ways to create capacity in a healthcare system: decrease length of stay, decrease arrivals to a system, or build more beds. Building beds is not an optimal plan in today’s financial climate, so the ideal place to start is thinking about how to decrease length of stay. Decreasing length of stay does not mean decreasing patient care time. Each patient should receive compassionate, customized care with ample time with the healthcare team. This is con-sidered “value-added” work in Lean Healthcare 3002401801206000 (0%)0.4 (8%)0.8 (16%)1.2 (24%)1.6 (32%)2.0 (40%)2.4 (48%)2.8 (56%)3.2 (64%)3.6 (72%)4.0 (80%)4.4 (88%)4.8 (96%)Arrival Rate (Percent Utilization)Average Waiting Time (minutes)Fig. 18.1 Utilization curve. (Image courtesy of Charles E. Noon)

297[9]. Where length of stay can be decreased is in all of the non-value-added activities that occur in medicine. Examples are waiting for care, excess movement and transportation, equipment not available, poor communication and defects, and rework. Improving flow in the ED is a “war won in minutes,” and on the hospital side, it is a “war won in hours.” When looking at ED flow proj-ects, the goal should be to decrease length of stay by minutes at a time. On the hospital side, the goal should be to decrease total length of stay by several hours at a time. Surprisingly, these small incremental improvements can drive bed utiliza-tion down and eliminate waits for patients. This defined means to look at every step in a process critically and redesign patient-centered flow sys-tems that decrease length of stay. (Key Points Box 18.1)Key Points Box 18.1Lean Healthcare uses some of the same operational principles originally developed by the Toyota Production Company. Lean Healthcare is a set of operating philoso-phies and methods designed to create max-imum value for patients. It uses basic tools and methods to systematically reduce waste and therefore waits for patients. It emphasizes what is value-added from the patient perspective, employee involvement, and continuous improvement. Waste in the system is considered and removed when-ever possible. Waste in Lean Healthcare includes excess transportation, inventory, motion, waiting, overproduction, overpro-cessing, and defects. Examples include waits for transport, blood hemolysis, car-diac monitoring when not required, and hospital readmission.Vignette 18.2Imagine a patient arriving in an emergency department with a wrist injury after snow-boarding. The patient signs in at registra-tion and her verbatim chief complaint is entered. She is immediately brought into a treatment room chair and assessed by a physician assistant and a nurse together. The history is taken just once, and all nec-essary information is gathered. Vital signs are obtained, and pain is treated on the way to radiology. After the x-ray is taken, the physician assistant immediately shows the x-ray to the patient who is reassured there is no fracture. A splint is applied, and dis-charge instructions are discussed by the team, and all questions are answered. Within 30 minutes the patient is heading back home to their family. The patient can-not stop talking about the excellent care, and the healthcare team has the ability to spend even more time with patients because much of the waste is removed from the system.There is an old saying “there is only one way to eat an elephant: a bite at a time.” The above described care pathway was not developed overnight. It occurred after a systematic redesign of each aspect of patient care. First, triage was eliminated when possible or markedly shortened for all patients. Then, care teams were devel-oped for low-acuity flow. Initially there was a separate office for providers and nurses, but the team recognized this created waste for the patient since the history was repeated. Every step in the process was carefully considered and redesigned after multiple trials with input from all workers and patients. Every step was measured, equipment was standardized and opti-mized, and care teams were developed. No detail was too small. An example was treat-ing patients in chairs instead of lying on a gurney. A gurney was available if needed, but if not, a chair could be cleaned in 30 seconds, while a gurney took 3 minutes to turnover. This small step multiplied over hundreds and thousands of patients makes a huge difference. The new seamless

298 Segmentation for QualityWe can infer in medicine that waiting and board-ing lead to adverse outcomes for patients simply due to the lack of immediate attention and care. Mathematically, it would seem that pooling of resources to care for patients would optimize care and flow through a hospital system. Considering and optimizing workflows with the main idea of preserving high-acuity beds in both the ED and the hospital can jumpstart flow through the entire system. This segmentation allows care providers to standardize care based on evidence-based best practices. Starting in the ED, the most common type of segmentation is a fast-track area for low- acuity patients. One excellent model locates the fast track near the front door. The goal of design is to minimize movement of the patient and the care team and be sure all equipment needed for the team is at their fingertips. This optimized model has the physician or advanced care pro-vider and the nurse sitting together. The patient tells their story one time to the care team, and all necessary information and data (vital signs, exam) are collected in the one room. If no testing is necessary, discharge instructions and prescrip-tions are printed in the room and given to the patient. If studies are necessary, a clearly defined area in view of the care team is established as a visual signal that the x-ray is completed. With this model, only two rooms are necessary for care: the assessment room and a procedure room. This system maximizes team work and elimi-nates most of the motion waste and poor com-munication of other systems. Improving flow is a war won in minutes, and this method eliminates much of the unnecessary flow. The team can see three to four patients per hour but usually only have one to two active patients because it is a flow- based model. When care is done, the patient is discharged home quickly and efficiently. All members of the care team have more time to spend with patients and less on wasted motion.Another idea utilizing operational redesign is a vertical treatment area in the emergency depart-ment. Traditionally, most patients are treated on a gurney in the ED. These beds become the scarc-est resource, and lack of beds creates long waits and delays in care. Almost every ED has over 50% of patients who are categorized as “mid- acuity” and are not differentiated more than that. Many of these patients are well-appearing patients who need more testing to determine the diagnosis. These are the patients that can rapidly occupy all the high-acuity beds and create long delays for patients. There are many workflows done in emergency medicine that continue just because historically they have always been done a specific way. Starting IVs on all patients getting a blood draw is one such procedure. When inter-viewed, most patients would prefer to just know what is wrong and not have an IV unless medica-tions are needed. There is also much evidence that many medications are safer and as effective when given orally. If a patient presents with a complaint such as abdominal pain and looks well, a vertical treatment space may be ideal. In this system, the patient would again be seen by the care team of a provider and nurse together. An assessment is done including history and physi-cal exam and tests are ordered. A nurse or phle-botomist draws blood, all radiology studies are ordered, and oral medications are given. Instead of lying on an uncomfortable gurney, the patient waits in a “results waiting room.” When all tests are completed, the provider brings the patient back into the assessment room to discuss the results and a plan for care. If results reveal some-thing that requires admission, then at that time, the patient can be transferred to the main ED for further treatment. This is a very small percentage of patients. This requires system redesign and process created high-quality care, the high-est patient satisfaction, and job satisfaction that was unmatched. Because the team had more time and care was standardized, patient complaints were essentially elimi-nated. With the new system, the teams were much more productive but still had time to talk with their paired colleague and “make a new friend” as a department leader often stated. These changes were both better for patients in both quality and experience and easier for people doing the work with higher job satisfaction.

299thinking in a different way and considering every step in a process. Each time an unnecessary IV is established, it adds to the nursing workload. Every IV that is inserted must be removed as well. IVs are generally safe but are not com-pletely without risk as well. Peripheral IV infil-trates and extravasations, site hematomas, phlebitis, and air embolism are all known rare complications. Care is improved when the care team is trained to think differently – if a patient is dehydrated and needs fluids, an IV is established, but if it is not needed avoiding this procedure can decrease workload, preserve high-acuity beds, and improve patient quality and satisfaction.On the hospital side, these same principles hold true. Creation of a low and mid-acuity observation unit will preserve high-acuity beds and prevent boarding for patients. Many hospitals cohort observation patients based on CMS rules that were developed and state that any patient who does not stay “two midnights” should be put into observation status. This has broadened the number of patients on observation status. Cohorting of these patients is faulty and will not improve hospital flow. Observation should be created around standard diagnosis-based care plans with the idea for 24-hour and 48-hour care plans. There is much evidence available about optimal care and quality for various diagnoses. A perfect example is transient ischemic attack (TIA) or stroke without deficits. Often these patients are in the hospital for several days. With careful planning all evidence-based care can be completed within 24 hours. Often, patients are observed on a cardiac monitor for long periods of time even when they do not require it. Patients can have all imaging done, see a consultant, and get an evidence-based treatment plan in under 24 hours. Again, a system that is better for patients, delivers higher quality care, and is better for the healthcare system. There are many diag-noses that are amenable to this kind of care, and a system can be created to optimize both quality of care and flow. Chest pain, syncope, asthma, head injury, and TIA as noted above are all diag-noses that are amenable to a 24-hour observation unit, but any diagnosis that can have a protocol established with clear pathways can be placed in observation. An example of a new process to improve flow is a procedure room co-located within the observation unit. This can allow GI and pulmonary specialists to do procedures eas-ily and allows many more patients to be placed in the unit. Most 24-hour observation units are staffed by emergency physicians or supervised advanced practitioners and nursing staff with hospitalist consultation as needed.Several other segmented units are possible in the hospital to improve flow. For example, a rapid surgical unit with agreed-upon protocols can pro-vide high-quality, patient-centered care. Patients can understand what needs to happen postop to progress to discharging home before the surgery even happens. Technology can be used to record walking, PO intake, and pain control. These rapid surgical units can decrease length of stay by 24 hours or more – especially if patients are given the discharge information pre-op. Another seg-mented unit is a medical “48-hour” unit staffed by hospitalists. This is ideal for diagnoses like congestive heart failure and pneumonia that require therapies but are well-defined and have outcome measures for discharge.There are two primary goals of segmentation: provide protocol driven, patient-centered high- quality care and preserve high-acuity beds in the ED and hospital for admitted patients. There is risk with segmentation, however. Queueing the-ory can help to illustrate this problem. Queuing theory is the mathematical study of waiting lines or queues. It is considered a branch of operations research and originated with research by Agner Erlang when he created models to describe the Copenhagen telephone exchange [10].Using this mathematical model, in general systems should pool resources, which means sharing patient load among all physicians and nurses to prevent one server sitting idle while the other is overwhelmed. In the ED, an example would be when physicians pick up patients when they feel they are ready in any area and work with all nurses. Despite this general principle, seg-mented care improves quality and flow when there is an assigned team and clear workflows [11]. The power of each of these segmented care pathways is in the teamwork and how the well-defined workflows are designed to decrease the length of stay. These areas must be designed to

300not become the safety net when things are going wrong and prior to initiating volumes must be assessed so each area is highly productive. Using data to design workflows will determine the hours of operation for each segment and help to decide if segmented areas should be combined (combination low and mid- acuity patients in one area) and also the number of observation unit beds in the hospital. Well- defined segmentation can jumpstart flow of the entire hospital system. Teamwork and CommunicationIn the book Team of Teams, General Stanley McChrystal [12] describes two case scenarios: the first in which a plane crew with an hour of fuel, no incapacitating technical issues and clear protocols in place to deal with small technical issues crashed in 1978, and the second and very famous scenario where Captain Sullenberger landed a plane in the Hudson River in 2009 after complete engine failure 2000 feet above the ground. General McChrystal describes how a pri-mary difference in the two scenarios was how the airline crews were structured. In 1978, airline crews were structured as a command. All team members looked to the captain for instructions, orders, and guidance. By 2009, airline crews were structured as teams, and each crew member knew their role and had the autonomy to make decisions under pressure.Atul Gawande describes this same concept in The Checklist Manifesto [13]. He describes the importance of standard work in the form of checklists while still having communication and teamwork to allow each team member to feel comfortable identifying and solving identified problems. In order to improve operations, each team member must understand their individual role in the patient care system. In our current healthcare system, this lack of standard work and role clarity can lead to delays and quality issues throughout the system.A clear recommendation to improve flow and quality in healthcare is the establishment of care teams. In the ED, the pairing of low and mid- acuity providers with a nurse while minimizing movement improves flow, communication, and quality. For high-acuity ED patients, establishing care teams in the main ED will improve all met-rics. When the high-acuity teams are sitting in close proximity and processes are implemented where all team members greet patients on arrival, there are clear communication and expectations set with the care team. Patients are also aware immediately about the treatment team and can be involved in decisions at the outset. This close proximity allows the team to round on patients together multiple times in a shift. This improves communication and balances the workload for nursing as well. Close communication prevents missed orders and clarifies the care plan for the team. On the hospital side, a similar situation is the geographic assignment of hospitalists on the floor. When this is possible, communication is markedly improved and nurses can prioritize work after rounding with the physician. Whether implemented or not, daily multidisciplinary rounding as standard work involving the patient and family has been shown to improve communi-cation, assure all care activities are completed Vignette 18.3A 75-year-old woman comes in with a lower GI bleed at 10 pm. She is rapidly seen by an ED physician. An IV is estab-lished, baseline labs are drawn, and her vital signs and hemoglobin are stable. The observation unit is contacted and care is transferred. A standardized bowel prep is started, serial hemoglobin levels are drawn, and a message is left for the GI physician. At 7 am, the GI physician checks the phone and prioritizes patients in the observation unit. The procedure is done in the unit, and bleeding has stopped. Results are discussed with the patient, and she is very happy to go home 14 hours after admission to the ED. The GI doctor states “he can do twice as many procedures easily because he is not waiting for patients to be transported.” The patient raves to her neighbors about the timely high-quality care she received.

301(checklist), and involve patients and families in decision-making. Performance Improvement in Healthcare(Key Points Box 18.2)Creating a healthcare system with optimal flow and quality starts with a passion for excel-lence initiated by a determined leader. This leader creates the vision for either the organization or department and shares it on a daily basis with the healthcare team. Servant leadership means that the professional development of the team is a high priority. The vision of the leader is consid-ered the “true north” and guides project develop-ment. The basic principles of Lean Healthcare mean that leaders “go to the Gemba” to see the work, and workflows are not top-down driven but involve frontline workers and a “Kaizen” mindset of continuous improvement. “Kaizen” is a Japanese term that means continuous improve-ment. This leadership vision combined with a Kaizen mindset leads to a culture of respect that runs through the entire organization.This leadership and vision can jumpstart a process improvement project. It lets frontline staff know that leadership considers consistent improvement a priority. Leadership commitment to the monetary costs of training staff on basic principles of process improvement and the tools needed for a disciplined approach emphasizes this organizational prioritization. This training allows staff to look at processes with an engineer-ing mindset and a critical thinking approach. The goal of any process improvement project is to create systems that are better for patients but eas-ier for people doing the work. This essential prin-ciple guides the selection of frontline staff involved in each project. Process improvement is Vignette 18.4In a high-volume ED with a footprint as large as a football field, prior to any opera-tional changes, physicians cared for patients in any area of the department. It was very common for physicians to have patients on each end of the department. It was difficult for nurses to know which phy-sicians were treating patients, and no com-munication system was in place. One day, the treating physician was sitting at one end of the department, while a patient located on the opposite side had a systolic blood pressure of 70. The nurse did not want to leave the critical patient but was unsure how to find the physician.The ED leadership team recognized that the system that was in place impeded com-munication and created uneven workflows. They created care teams of one MD and sev-eral nurses who sat together and cared for patients in a pod. A patient came in with a fever and low blood pressure. The team rec-ognized that the patient likely had sepsis, and timely antibiotics and fluids were important to decrease mortality. The team had standard work where the entire team met the patient on arrival. While the physi-cian spoke to the patient and family and got history and placed orders, one nurse com-pleted all tasks and started IV fluids and obtained labs. The other nurse documented the care and obtained timely antibiotics. With timely care, the patient’s blood pres-sure stabilized, and the patient received anti-biotics within 1 hour of arrival. The patient was quickly admitted to the hospital.Key Points Box 18.2Servant leadership is a leadership philoso-phy in which the leader embraces care of employees. This varies from traditional leadership where the leader’s focus is orga-nizationally driven. A servant leader shares power, puts the needs of the employees first, and helps people develop and perform as highly as possible. The focus is on employee personal growth. The benefit is in increased employee engagement and commitment to the organization [14].

302a long-term strategy. The goal is continuous improvement with clearly measurable metrics determined at the outset that are visible to all employees. A several day Kaizen event has been shown to change faulty processes that have been in place for years, and the rewards will be in improved quality care, engaged staff with higher satisfaction, and improved organizational finan-cial performance. The staff becomes a “commu-nity of scientists.”One option to start the journey is at the depart-ment level. After the vision is set by senior lead-ership, focused Lean training can begin on the unit level. A multidisciplinary team is identified to work on a predetermined problem that creates a bottleneck in the department. Pre-work is done to determine the scope of the project and to develop metrics that will show improvement. During the event, the current process is mapped out with each step written down from the patient perspective. This opens the eyes of the team to problems. Each step is determined to be “value- added” or “non-value-added” from the patient perspective. This creates a framework for the development of a new improved process that is then trialed and a system is put in place to trial along with a plan for complete implementation. At the end of the event, the team presents to upper leadership who relate the project to the overall organizational vision. This validation by leader-ship begins to create a flow-based culture focused on patient-centered continuous improvement. This step-by-step approach is repeated over and over and when combined with daily operational boards with escalation to senior leadership results in a recipe for high-quality, cost-effective patient care without delays. There are several healthcare organizations in the USA that are well-known for their implementation of Lean Healthcare systems. Virginia MasonVirginia Mason, a healthcare system based in Seattle, is known throughout the world as a leader in patient safety and quality by using the princi-ples described above. The journey started in 2002. In 2000, Dr. Gary Kaplan took over as CEO. The company had suffered financial losses that threated long-term survival. He was a vision-ary leader who told his team: “we change or we die.” The entire executive team flew to Japan to observe Lean management techniques and the Toyota Production System and then developed a new strategic plan that was patient-centered and focused on four pillars: people, quality, service, and innovation. They created a program of con-tinuous improvement that also involved patient input. This vision has not varied, and now the medical group has integrated the philosophy throughout the entire system. They now teach process improvement and training to healthcare leaders and are known throughout the world as a patient-centered quality leader because of their search for a perfect patient experience that elimi-nates errors and defects [15]. Intermountain HealthcareAnother well-known health system that has applied process improvement principles to healthcare is Intermountain Healthcare. They are based in Salt Lake City, Utah, and have visitors from around the world who come to learn how they have developed and sustained improvement. Intermountain started their improvement journey with their frontline team members. They recog-nized that there was no standard work for front-line managers and employees. They redeveloped charge nurse positions with coaching and clear expectations. Charge nurses were required to round on frontline staff to see if standard tasks were completed. Starting here allowed the orga-nization to spread the culture of improvement both up and down the chain of command. Daily rounds were initiated on each unit with an escala-tion up the leadership chain including to the CEO, allowing upper leadership to understand problems at the frontline daily [16].Each of these health systems has made the commitment to strive for perfection using pro-cess improvement methodology.

303 Open Data to Drive PerformancePhysicians are often skeptical about data, particu-larly unblinded data. This is related to a number of concerns. Physicians question the accuracy and relevance of data. Most importantly physi-cians are skeptical about the intent behind the release of unblinded data. There is general con-cern that data will be used punitively or that data will be released to the public. Usually, the goal of data release is to influence physician behavior or increase physician productivity.When implementing multiple changes to improve flow, it is critical to have well-defined metrics to show that the system is improving. Having a discussion with the staff involved in the process about which metrics will be tracked and addressing concerns about data validity at the onset will improve buy-in and avoid contro-versies later. When the data is used in context with the vision for high-quality and improved patient care, physicians are less likely to feel individually attacked.Many organizations make the mistake of shar-ing multiple metrics without context with work-ers. Most workers will not look closely at the metrics, and even more importantly many times, this will create a defensive culture and not lead to any organizational improvement. Instead of this, a few well-targeted metrics shared in the context of the vision for improved patient care and flow can engage the entire team. When these metrics are shared with advice on best practices from peers, data can be used to drive performance. Physicians and other healthcare professionals are much more likely to respond to active sharing and a collaborative approach in contrast to pas-sive sharing [17].Vignette 18.5A study was done looking at two individual EDs. One ED shared blinded data with the group. The second ED shared unblinded data in a different way. The data was com-bined with identification of high perform-ers and discussion of best practices that were validated with the group. The study found that the first ED had no significant improvement in physician metrics, while the second ED that combined sharing data with education around best practices had a 10.9% improvement in physician produc-tivity with significant reduction in variation across providers. This improvement was not associated with any declines in quality or service scores. The data that was shared was done in a nonpunitive way and was combined with a discussion around the vision for improved patient care and elimi-nating waits.Vignette 18.6At a monthly staff meeting, unblinded data was shared by the radiology cham-pion on CT scan utilization for all provid-ers. Prior to sharing, there was a discussion about possible concerns of the physician group and because of this the data excluded trauma patients where ordering was not under the ED control. When the data was shared, lower utilizers gave their own tips and strategies to the group in a fun and entertaining way. The newest research was also shared including the PECARN (Pediatric Emergency Care Applied Research Network) study with best practices for children with head inju-ries. One physician reported to the group “I had no idea I was the highest utilizer of head CT’s in the department!” She reported later that the positive attitude of the group and the education she received allowed her to comfortably change her practice. The next month, her CT utiliza-tion was at the 50th percentile with no quality issues. This improvement was sus-tained the following year.

304 Technology to Improve FlowIn the book Punish the Machine [18], Dr. Uli Chettipally states that 50% of what physicians do in American healthcare is unnecessary, ineffec-tive, or dangerous. This waste can be categorized as excess movement, waiting, unnecessary steps, unnecessary procedures, and errors. To improve medical care and reduce cost, Lean operations’ management with a critical eye and embracing technology are necessary. Currently the elec-tronic medical record is not being used to its full-est potential. At the present time, the computerized medical record is a crude instrument used for documentation, reimbursement, and regulatory compliance. There are two areas in particular that technology could improve flow and quality: oper-ational flow and clinical care.For operational flow, there is a wealth of opportunity. In Silicon Valley, there are a group of engineers engaged in improving healthcare operations. This markedly different approach uses predictive analytics and machine learning to predict when crowding and surge are about to occur and offers possible solutions. In real time, the analytics can help with flow by deciding which patients should get testing first when a test is the only barrier to discharge. It can schedule the hospital stay so patients know what to expect, and it can notify the physician when all testing is completed and even predicts the probability the patient will be admitted and request a bed much earlier in the process. On the retrospective side, software allows a manager to perform root cause analysis of problems quickly and efficiently. An example is when it is noted that the length of stay was longer during a particular part of the day. The manager can simply click on that period and find out the cause. Perhaps the lab was slower than normal, or radiology turnaround time was not optimal, or there was an influx of patients that overwhelmed the treatment team. This allows tar-geted improvement to occur. These examples are only the tip of the iceberg for what artificial intel-ligence can do for operations. Imagine if the sec-ond patient walks into the ED, it is predicted that the patient is at risk of a serious outcome. The patient is pulled in front of other less serious patients and immediately seen by a treatment team and care is initiated.On the clinical side, artificial intelligence is even more promising. New medical research is released daily, and it is impossible for the indi-vidual physician to keep abreast of all new trends. There is a widely quoted article that it takes 17 years for a new best practice to be imple-mented broadly [19]. Programs are being devel-oped to put best practice healthcare recommendations in front of the physician while treating the patient. This allows patients and their families to be involved in the clinical decision- making. While this has much promise and is markedly better than what is available now, the future holds even more promise. Imagine if after seeing a patient, the physician is able to see what the outcome was for the last thousand patients with the same diagnosis and if treatments could be targeted to patient’s specific characteristics.The future is bright for the use of technology to improve both patient clinical care and flow through the hospital system. It will require a new mindset and a partnership between clinicians and engineers to put this in practice. The emphasis must be on helping the clinician deliver high- quality clinical care while eliminating waste and unnecessary cost in the system. The case study below illustrates how data can be used with patients for shared decision-making utilizing best practices.Vignette 18.7A 7-year-old girl is brought into the ED after a head injury while playing soccer. She hit her head on another player and fell to the ground. There was no loss of con-sciousness, but she has a mild headache and nausea. The mother is very concerned and wants a cat scan. As the physician opens the chart, the PECARN rules open, and the physician explains the significance of the study and goes through each of the questions together. At the end, the com-puter states the risk for serious head injury is under 1/2000, and the physician and

305GCS=14 or other signs of altered mental statustor palpable skull fracturetOccipital or parietal or temporal scalp haematoma,or history of LOC ≥5 s, or severe mechanism ofinjuryt‡, or not acting normally per parent53·2% of population<0·02% risk of ciTBI32·9% of population0·9% risk of ciTBI• Physician experience• Multiple versus isolated§ findings• Worsening symptoms or signs after emergency department observation• Age <3 months• Parental preferenceObservation versus CT on the basisof other clinical factors including:13-9% of population4.4% risk of ciTBICT recommendedCT recommendedCT not recommended¶abCT not recommended¶YesYesYesYesNoNoNoNo14·0% of population4.3% risk of ciTBI28·8% of population0·8% risk of ciTBI57·2% of population<0·05% risk of ciTBIHistory of LOC, or history of vomiting, or severemechanism of injury‡, or severe headacheGCS=14 or other signs of altered mental statust,or signs of basilar skull fracture• Physician experience• Multiple versus isolated§ findings• Worsening symptoms or signs after emergency department observation• Parental preferenceObservation versus CT on the basisof other clinical factors including: Fig. 18.2 a children <2 years of age and b is for children ≥2 years of age [20]. Reprinted from The Lancet, Vol. 374/Edition Number 9696, Kuppermann N, Holmes JF, Dayan PS, et al., Pediatric Emergency Care Applied Research Network (PECARN). Identification of children at very low risk of clinically-important brain injuries after head trauma: a prospective cohort study, p. 1160–11770, Copyright (2009), with permission from Elsevier)mother have an informed discussion about the risks and benefits of CT. They decide to forgo the CT for now, and head injury instructions are given. They are discharged within 30 minutes of arrival, costs are mini-mal, and the child has an excellent outcome without the radiation risk associated with an unnecessary CT scan usage. Both physi-cian and mother are very happy with the clinical interaction. (Fig. 18.2)

306 ConclusionIn the USA, about 17% of GDP is currently spent on healthcare. On average, other wealthy coun-tries spend about half as much per person on healthcare. Even more importantly, the quality of care delivered is poor in many instances. A 2014 report from the Commonwealth Fund stated that the USA “ranked last overall among 11 industri-alized countries on measures of health system quality, efficiency, access to care, equity and healthy lives” [21]. While the USA has the high-est costs, it also has the lowest performance.Changing this alarming pattern will require a paradigm shift in how medical care is delivered across the system. Excellent patient flow across a system utilizing technology and best practices will deliver the highest quality care. There are bright spots across the country. These must be embraced and used as a stepping stone to even further improvement.To make this fundamental shift, leadership will be key. These leaders will set the vision for delivering high-quality, cost-effective care while recognizing how important timely care is for both patient and caregivers. This respect for time and elimination of waste will improve patient and family satisfaction and prevent physician burn-out. Thinking about how every decision impacts flow is fundamental. Real change will come when the frontline staff doing the work under-stands the vision and are engaged to solve prob-lems. High-quality care and operations will always be intertwined. While this is not a small undertaking, this vision combined with a mindset of continuous improvement will assure improved high-quality patient care [22].Editors’ CommentsThe most important facet impacting the business of healthcare is probably patient flow; having patients move throughout the system in a safe, expeditious manner is cru-cial for optimizing operations while simul-taneously enhancing the finances of the organization. As such, this chapter presents excellent strategies on how to ensure the patient flow in your healthcare system drives quality.The concept of patient flow is relatively new and may not be fully understood by all of our readers. The author begins the chap-ter with striking case studies that demon-strate through the vignettes the impact of patient flow on quality, safety, and out-comes; indeed, those of us in healthcare settings have seen this replayed – some-times on a daily basis. Once the reader understands how patient flow contributes to overall quality and safety not only for that one patient, but for patients within the entire system, then we have reached com-mon ground.The author builds on the burning plat-form of patient flow by delving into the sci-ence of flow. For the sophisticated reader, the chapter explores how improvement sci-ence can lead and support patient flow efforts and how to best consider patient flow from a scientific approach. Through the vignettes, the author highlights specific scenarios such as “low-acuity flow” and “segmentation for quality.” Both are explained thoroughly by the author that it is not unreasonable for the reader to pilot some of the strategies that may apply in their healthcare system.As seen throughout the prior chapters in this textbook, success in managing and optimizing patient flow comes down to teamwork and communication. Teamwork and communication are the keystones for any successfully change endeavor that is of the magnitude and significance as patient flow. The teamwork and communication can be microunit based or can be organiza-tional; depending on the scope of the patient flow initiative, the improvement scientist will have to build the right team and utilize the proper communication channels.

307 Chapter Review Questions 1. What is the definition of Lean Healthcare?Answer: Lean Healthcare is a set of oper-ating philosophies and methods based on the Toyota Production System principles designed to create maximum value for patients. It uses basic tools and methods to systematically reduce waste and therefore wait times for patients. It emphasizes what is value-added from the patient perspective, employee involvement, and continuous improvement. 2. What are the two known impacts on patient care and hospital operations found with board-ing patients in the emergency department?Answer: Increased mortality and longer length of stay after admission. 3. Describe two examples of segmentation in the emergency department and one example for the inpatient units.Answer: (a) In the ED, a streamlined low- acuity treatment area and a vertical treatment area for mid-acuity, well-appearing patient, and (b) on the inpatient unit, an observation unit with defined patient pathways. 4. True or False: The role of a leader is to direct frontline staff on their daily work.Answer: False. The role of the leader is to create a vision for the staff and support the process improvement work in process. 5. Does transparent data improve physician performance?Answer: It depends. If transparent data is paired with sharing of best practices, physi-cian performance can improve. Sharing of transparent data is not a panacea, but there is broad subjective sentiment that transparency drives improvement. 6. What is the definition of Kaizen?Answer: Continuous improvement. A Kaizen event brings a group of people together in a structured way to solve a well-defined problem. 7. Describe two ways that analytics will help improve flow and patient care in the future.Answer: Analytics can help with retrospec-tive review of issues, real-time “pushes” to staff to assist with workflow, and prospective predictions of high volume with recommen-dations to plan for it. On the clinical side, ana-lytics can bring best practices up to the clinician and allow shared decision-making with patients and family. 8. Describe a basic framework for process improvement in healthcareAnswer: Process improvement is best when leadership sets the vision. Lean education of frontline staff ensures everyone uses the same methodology for projects. This combined with multidisciplinary teams for process improvement, Kaizen events, and clear met-rics using technology create a robust frame-work for improvement.References 1. Pedroja AT. The tipping point: the relationship between volume and patient harm. Am J Med Quality. 2008;23(5):336–41. 2. Rutherford PA, Provost LP, Kotagal UR, Luther K, Anderson A. Achieving hospital-wide patient flow. IHI while paper. Cambridge, MA: Institute for Healthcare Improvement; 2017. 3. Singer AJ, Thode HC Jr, Viccellio P, Pines JM. The association between length of emergency depart-ment boarding and mortality. Acad Emerg Med. 2011;12:1324–9. 4. Liu SW, Chang Y, Weissman JS, Griffey RT, Thomas J, Neugui S, Hamedani AG, Carmago CA Jr, Singer S. An empirical assessment of boarding and qual-ity of care: delays in care among chest pain, pneu-monia, and cellulitis patients. Acad Emerg Med. 2011;18:1339–48. 5. Litvak E. Managing patient flow in hospitals: strategies and solutions. Oakbrook Terrace: Joint Commission Resources; 2010.Ultimately, success in managing patient flow will come down to data and technol-ogy, the two final areas discussed in the chapter. This chapter serves as an excellent primer to introduce readers to patient flow and to simultaneously guide the advanced improvement scientists on the best strate-gies to incorporate improvement method-ologies to patient flow.

308 6. Liker JK. Developing lean leaders at all levels: a prac-tical guide. San Bernadino: Lean Leadership Institute Publications; 2014. 7. Wince R Top 5 leadership characteristics for process improvement. Retrieved from https://www.govloop.com/community/blog/top-5-leadership-characteris-tics-for-process-improvement-success/ 8. Crane J, Noon C. The definitive guide to emergency department operational improvement. New York: Productivity Press; 2011. 9. Chalice R. Improving healthcare using Toyota lean production methods. Milwaukee: ASQ Quality Press; 2007. 10. Sundarapandian V. 7. queueing theory. Probability, statistics and queueing theory. PHI Learning; 2009. ISBN 978-8120338449. 11. Song H, Tucker A, Murrell K. The diseconomies of queue pooling: an empirical investigation of emergency department length of stay. Manag Sci. 2015;61(12):3032–53. 12. McChrystal S. Team of teams: new rules of engage-ment for a complex world. New York: Penguin Publishing Group; 2015. 13. Gawande A. The checklist manifesto: how to get things right. New York: Metropolitan Books; 2010. 14. Sen S, Sarros JC. Servant leadership: its origin, development, and application in organizations. Organizational Studies. 2002;9(2):57–64. 15. Kenney C. Transforming healthcare: Virginia Mason medical Center’s pursuit of the perfect patient experi-ence. New York: Productivity Press; 2011. 16. Baker G, MacIntosh-Murray A, Porcellato C. High performing healthcare systems. Toronto: Longwoods Publishing Corporation; 2008. 17. Song H, Tucker A, Murrell K, Vinson D. Closing the productivity gap: improving worker productiv-ity through public relative performance feedback and validation of best practices. Management Sci. 2018;64(6):2473–972. 18. Chettipally UK. Punish the machine: the promise of artificial intelligence in health care. Charleston: Advantage Media Group; 2018. 19. Morris Z, Wooding S, Grant J. The answer is 17 years: what is the question: understanding time lags in translational research. J Royal Soc Med. 2011;04(12):510–20. 20. Kuppermann N, Holmes JF, Dayan PS, et al. Pediatric emergency care applied research net-work (PECARN). Identification of children at very low risk of clinically- important brain injuries after head trauma: a prospective cohort study. Lancet. 2009;374(9696):1160–70. 21. Potraj J. The quality of US healthcare compared with the world. 2015. Retrieved from: https://www.ajmc.com/contributor/julie-potyraj/2016/02/the-quality-of-us-healthcare-compared-with-the-world 22. Scoville R, Little K, Rakover J, Luther K, Mate K. Sustaining improvement. IHI white paper. Cambridge, MA: Institute for Healthcare Improvement; 2016. (Available at ihi.org).

309© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 R. K. Shah, S. A. Godambe (eds.), Patient Safety and Quality Improvement in Healthcare, https://doi.org/10.1007/978-3-030-55829-1_19It Takes Teamwork: Consideration of Difficult Hospital-Acquired ConditionsJ. Wesley Diddle, Christine M. Riley, and Darren KlugmanJ. W. Diddle Pediatric Cardiac Intensivist, Children’s National Hospital, Cardiac Critical Care Medicine, Washington, DC, USAe-mail: [email protected] C. M. Riley (*) Cardiac Intensive Care Unit, Children’s National Hospital, Washington, DC, USAe-mail: [email protected] D. Klugman Cardiac Intensive Care Unit, Children’s National Heart Institute, Children’s National Hospital, Cardiac Critical Care Medicine, Washington, DC, USAe-mail: [email protected] Objectives• To explain the high risk ICU environ-ment and the associated impact on patient safety• To define hospital-acquired conditions, using unintended extubations as an exemplar• To highlight the importance of interpro-fessional collaboration and system redundancy in quality improvement processes• To understand the impact proactive pro-ces