NUR 282 Research Article Summary 0221

Background: The Joint Commission has identified miscommunication as a leading cause of sentinel events, the most serious adverse events, but it is unclear what role miscommunications play in malpractice claims. We sought to determine the proportion of medical malpractice claims involving communication failure and describe their nature, including providers involved, locations, miscommunications types, costs, and the potential for handoff tools to avert risk and associated costs.

Methods: We retrospectively reviewed a random sample of malpractice claims from 2001 to 2011, collected in CRICO Strategies’ Comparative Benchmarking System, a national claims database. Two researchers reviewed cases to determine if a claim involved communication failure, its type, and potential preventability using a communication tool. Interrater reliability was assessed by dual review of 50 cases (81% agreement, [kappa] = 0.62 for evidence of miscommunication). Claimant demographics, case characteristics, and financial data were analyzed.

Results: Communication failures were identified in 49% of claims. Claims with communication failures were significantly less likely to be dropped, denied, or dismissed than claims without (54% versus 67%, P = 0.015). Fifty-three percent of claims with communication failures involved provider-patient miscommunication, and 47% involved provider-provider miscommunication. The information types most frequently miscommunicated were contingency plans, diagnosis, and illness severity. Forty percent of communication failures involved a failed handoff; the majority could potentially have been averted by using a handoff tool (77%). Mean total costs for cases involving communication failures were higher ($237,600 versus $154,100, P = 0.005).

Conclusions: Communication failures are a significant contributing cause of malpractice claims and impose a substantial financial burden on the healthcare system. Interventions to improve transmission of critical patient information have the potential to substantially reduce malpractice expenditures.

Communication failures are a leading cause of sentinel events, the most serious adverse events in hospitals.1 These incidents result in significant morbidity and mortality for patients, manifesting as falls, delays in treatment and diagnosis, and wrong-site surgeries, among others. A growing body of literature suggests that miscommunications might also be an important and previously underrecognized contributor to medical malpractice claims.1,2

In a study of 43 pediatric claims from a single insurer in the 1990s, Pichert et al 3 found that nearly one-third of medical malpractice claims stemmed from failures of communication and were directly linked to adverse outcomes. More recently, a large multispecialty study by CRICO Strategies evaluated 23,000 malpractice claims and found that 3 of every 10 cases include at least 1 specific breakdown in communication,4 but an in-depth review to determine the precise nature of these communication failures or the potential role of communication improvement strategies in averting these claims was beyond its scope. These studies and others highlight the importance of communication failures as a contributing factor to adverse events and medical malpractice.5-7 Schema that stratify claims and events by the specialty in which they occur or by clinical categories of error may obscure the role that communication plays across clinical categories and settings.8,9

To reduce communication errors, structured communication tools and processes have been developed that promote provider discussion of critical patient care elements that are frequently omitted.10 In one multicenter study, implementation of a bundle of complementary tools organized around I-PASS, a structured communication process, was followed by a 30% reduction in preventable adverse events.11 A second multicenter study that adapted the I-PASS process to structure communication between healthcare providers and patients likewise was associated with a substantial reduction in preventable adverse events.12 Structured communication tools have the potential to improve patient safety and reduce adverse events.13-15 The potential of communication improvement interventions to avert or mitigate the most serious preventable adverse events, however, particularly those that lead to malpractice claims, remains unknown.

To address this knowledge gap, we investigated the role of communication failures in medical malpractice claims with specific attention to provider-provider and provider-patient handoffs of care using a structured retrospective review of 500 closed medical malpractice claims. Our aim was to evaluate the prevalence and characteristics of claims that include communication failures, and the potential of structured handoff tools to prevent them. We also sought to determine the costs associated with communication failure-related claims.

We conducted a retrospective study of a random sample of closed claims filed from 2001 to 2010 from the CRICO Strategies’ Comparative Benchmarking System (CBS). The study was approved by the Institutional Review Board at Boston Children’s Hospital.

CBS Data Set and Claims Sample

Data were extracted from the CRICO Strategies’ CBS. The CBS database includes more than 300,0000 medical malpractice cases from nearly 550 hospitals and healthcare entities, representing nearly 30% of all malpractice cases in the United States. The hospitals represented in the data set include institutions from across the spectrum of healthcare organizations, including academic and community health care systems. Self-insured and commercially insured hospitals are included. Comparative Benchmarking System cases have data elements coded in detail using the CRICO Coding Taxonomy and a summative loss abstract of the medical and claims chart review. Of note, some insurance companies submit cases without using the coding taxonomy; these cases were excluded from the current study if limited data were available. The CRICO Coding Taxonomy uses structured criteria to codify medical malpractice data elements. Data elements investigated in this study include demographic data, financial data, basic case data, and loss and clinical information (Supplementary Table 1, http://links.lww.com/JPS/A442). Claims are defined as cases in which there is a formal notice of intent to seek compensation for bodily injury that occurred because of negligence of the insured. Suits are defined as claims that result in filing of a lawsuit in a court of law. Case severity is determined during CBS assessment in alignment with the National Association of Insurance Commissioners severity scale with low severity (0-2), medium (3-5), and high (6-9), with low scores indicating a legal issue only and high scores consistent with death. To be included in the current study, cases must have undergone review and coding through use of the CRICO Coding Taxonomy and must have been a closed case.

In this study, we sought to review a random sample of 500 cases. The data set included 125,390 closed cases, of which 39,670 cases included loss abstracts. From those cases with loss abstracts, we randomly selected cases until we reached our goal of 500 cases (1000 cases were randomized, from which 627 cases underwent initial review to reach our target number of cases; cases were excluded from among these 627 if the loss abstract was incomplete). Two cases were later excluded from our list of 500 because of limited data, leaving a final data set of 498 cases that were included in the analysis (Fig. 1).

Development of Review Instrument and Review of Claims

To systematically evaluate claims in the database, we developed a study instrument through an iterative process of question development, case review, and subsequent question clarification. The final instrument contained 15 questions developed to evaluate claims for evidence of communication failure and to characterize the type of failure, where the error occurred, who the communication failure involved, the type of information not communicated, whether a handoff of care was involved, and the potential of a handoff tool to prevent the communication failure (Supplementary Table 2, http://links.lww.com/JPS/A442). A handoff of care was defined as a transition in the responsibility and accountability for patient care, for example, when a patient was transferred from one clinical location to another. If the patient harm described in the loss abstract was determined by the reviewers to be preventable and a miscommunication occurred involving a transition of care, the reviewers determined whether the use of a structured handoff of care tool could have mitigated or averted the risk of harm to the patient. The study instrument was then programmed into Research Electronic Data Capture (REDCap), a secure web-based software platform designed to support data capture for research studies that is hosted at Boston Children’s Hospital.16,17 As a final step in development, 2 investigators (K.E.H. and M.S.) jointly applied the study instrument to 50 cases to develop a shared understanding of how to complete each item.

To test the reliability of case classification, dual independent review of an additional 50 cases was conducted after the data instrument was finalized. Interrater reliability was good (81% agreement, [kappa] = 0.62 for evidence of miscommunication).

Financial Data

Financial data including indemnity and expense payments were included in this analysis. Indemnity payments include the money paid to the claimant for their damages. Expense payments are the defense costs incurred while adjudicating a claim, and no money is paid to the claimant. All financial data were adjusted for inflation to 2020 dollars using the U.S. Bureau of Labor Statistics Consumer Price Index (https://www.blas.gov/cpi).

Statistical Analysis

Patient demographics and case characteristics were reported using frequencies (percent) for categorical variables and means (SD) or medians (interquartile range) for continuous variables. Differences in claimant demographics and case-related factors (e.g., allegation type, severity, disposition) between cases involving communication errors and those that did not were compared using [chi]2 tests for categorical variables, and 2-sample t tests or Kruskal-Wallis tests for continuous variables.

Financial data analyses including total amount paid were calculated as the sum of expenses to defend the case and total gross indemnity paid (if applicable). Cumulative totals and means are reported for financial data. Because of large outliers, financial data were log10 transformed for bivariate analysis. Cumulative totals were compared using gamma regression analysis, whereas comparisons between means were made using 2-sample t tests. All analyses were performed in SAS (v9.4; Cary, North Carolina) at an [alpha] level of 0.05.

Of the 498 cases examined, the majority of claimants were female (58%), with nearly 60% of the cases occurring between the ages of 30 and 70 years with an average claimant age of 45 years. The average time from the event to the claim filing was close to 1.2 years, with the majority of cases resulting in medium (50%) or high severity (40%) patient harm. Most cases within the study sample proceeded to suit (70%), with the majority of cases being dropped, denied, or dismissed (60%); 38% resulted in settlement, and relatively few (<3%) went to trial. The majority of the cases occurred in inpatient units (45%), outpatient areas (30%), and the emergency department (10%); the most common responsible services caring for the patients were medicine (21%), surgery (18%), and obstetrics and gynecology (14%; Table 1). Allegations related to errors in diagnosis (25%), surgical treatment (25%), and medical treatment (22%) were most common. The total costs for all 498 cases in the data set were $97.1 million, with $20.8 million in expenses and $76.3 million in gross indemnity paid. The average cost per case was $195,000, with an average of $153,200 in indemnity costs and $41,800 in expenses paid. The average total expense in defending cases that were dropped, denied, or dismissed was $21,200 versus $460,600 for those that settled or went to court (P < 0.001).

Characteristics of Cases That Involved Communication Errors

Communication errors were identified in 49% of cases. Cases with communication failures were significantly less likely to be dropped, denied, or dismissed than were cases that did not involve communication errors (54% versus 67%, P = 0.015; Table 1). Forty-four percent of cases involving communication failures were settled. Two percent went to trial, all with a defense verdict. Cases with communication errors most frequently had allegations of errors in diagnosis (25%). Cases with an allegation category of surgical treatment, other, or anesthesia-related treatment were less likely to have communication errors as compared with all other allegation categories (Supplementary Fig. 1, http://links.lww.com/JPS/A442). The majority of cases where the primary responsible service was identified as nursing (70%), medicine (55%), oral surgery (55%), and emergency medicine (54%) were identified to have communication errors (Supplementary Fig. 2, http://links.lww.com/JPS/A442).

Overall, cases involving communication errors were more expensive to defend, with a cumulative total amount paid for all cases involving communication errors of $58.0 million versus $39.1 million for cases that did not involve communication errors (P = 0.03; Table 2). Average total costs (indemnity and expenses combined) for cases involving communication errors were higher at $237,600 versus $154,100 for cases without communication errors (P = 0.005). This difference was driven by the fact that cases with communication error were more likely to be settled or go to court; there was no association between communication error and average cost per case within dropped cases (indemnity and expenses combined; P = 0.25) or within settled cases at $485,900 for those with communication errors versus $426,500 for those without (P = 0.72). Furthermore, cases with communication failures were more likely to have indemnity payments; 108 (44%) had any indemnity payment as compared with 81 (32%) of those without communication failures (P = 0.005). Overall, 5% of those with communication errors resulted in an indemnity payment of >$1 million compared with 2% for those without communication errors (P = 0.04; Supplementary Table 3, http://links.lww.com/JPS/A442).

Type of Information Miscommunicated and People Involved With Communication Failures

Communication failures were most frequently errors of omission (79%) and included verbal transmission of information (54%) or verbal and written communication (35%; Table 3). These most often occurred in the inpatient (47%) and outpatient settings (40%). Fifty-three percent of communication failures occurred between the medical staff and the family, primarily between the physician and the family (75%). The remaining 47% of communication errors occurred among medical staff, most often between the attending physician and the nursing staff (37%), attending physicians between specialties (30%) and within a specialty (19%). The type of information most frequently not communicated or miscommunicated included contingency planning for the patient (50%), the patient’s diagnosis (33%), and the patient’s severity of illness (32%).

Communication Failures Between Medical Staff and Family Versus Among Medical Staff

The percentage of communication failures involving patient contingency planning or patient diagnosis did not differ depending on whether the failure was among the medical staff or between the medical staff and family (P = 0.12 and P = 0.75, respectively; Table 4). However, patient severity of illness was more common in communication errors among medical staff than between the medical staff and family (50% versus 14%, P < 0.001). Radiologic results (13% versus 6%, P = 0.04) and specialist recommendations (7% versus 2%, P = 0.02) were also more common among medical staff than between medical staff and families.

Communication failures occurring among medical staff were more likely to be settled or go to court compared with failures between medical staff and family (54% versus 39%; P = 0.02). Total costs were higher for failures occurring among medical staff with average total costs of $359,400 versus $130,700 for failures between medical staff and family (P < 0.001; Table 5).

Approximately, 40% of the cases that involved communication failures included a handoff of care, 77% of which were likely preventable with a handoff tool (Table 3). Among errors between medical staff, 53% involved handoff errors versus 30% among errors between medical staff and family (P < 0.001). Average total costs among cases involving a handoff of care were $348,900 versus $160,600 among cases that did not involve handoffs (P = 0.003; Table 5). Cases involving handoff errors that went to court or were settled were particularly expensive, averaging $643,100 compared with $348,500 for non-handoff communication cases that went to court or were settled (P = 0.01).

In a national database of malpractice claims, we found that communication failures pose a significant threat to patient safety and exert a substantial financial burden on the healthcare system. Nearly 50% of medical malpractice cases involve a breakdown in communication either between clinical care providers or between clinicians, patients, and families. Failure to clearly communicate contingency plans, patient diagnosis, and severity of illness were particularly common sources of malpractice claims. On average, claims involving communication failures cost 1.5 times as much as claims without communication failures and were over twice as likely to result in a payout of more than $1 million. Handoff failures were particularly costly, with those that settled or went to court averaging more than $600,000 per claim. Most handoff failures could potentially have been averted with the use of a structured handoff tool.

Prior research has identified communication failures in approximately one-third of medical malpractice cases.3,4 We found that miscommunications played an even larger role in malpractice cases than previously recognized. Our work reflects attention to communication as having any contribution to the event with more detailed analysis specific to the nature of communication failures as compared with prior studies. In addition, the CBS database we used for our study allowed for a detailed breakdown of cases and an analysis of the costs of different types of claims. In 2018, Painter et al 18 reported that, on average, a large health system experienced 0.82 malpractice claims per 1000 inpatient admissions (outpatient claims were excluded from analysis). This rate, combined with the cost data generated in our study, suggests that malpractice payouts in this health system would be estimated at approximately $15 million annually (excluding outpatient payouts), of which more than $9 million per year would be spent on cases to which communication failures contributed. The degree to which this may be generalizable across systems is unknown, as rates of malpractice claims vary widely across the United States. Our findings validate and build on previous studies that have suggested the enormous financial toll of miscommunications in hospitals. The CRICO Benchmarking report found that in the CBS database alone, communication errors were a root cause of malpractice claims totaling $1.7 billion over a 5-year period. Studdert et al 19 found that all medical malpractice cases involving medical errors were more costly than claims without errors. Understanding key areas of vulnerability can help organizations develop targeted, intervention-based solutions to improve communication and reduce medical malpractice expenditures.

A growing body of literature has demonstrated that communication failures are complex and occur at multiple points in the delivery of patient care, but are especially common during handoffs of care.7,10,20-22 Implementation of the I-PASS handoff program has been associated with significantly reduced miscommunications, medical errors, and injuries due to medical errors for physicians in diverse specialties, for nurses, and for a variety of hospital types.11,23-28 Provider-provider miscommunications represent only about half of all communication failures that occur in hospitals, however. In our study, as in CRICO Strategies’ prior analysis, provider-patient miscommunications are also a major source of malpractice claims.4 Consequently, it is also extremely important to engage patients and families in collaborative, shared care models and use highly reliable processes in provider-patient communications.29,30 Khan et al 12 demonstrated a 37.9% reduction in harmful adverse events after an adaptation of the I-PASS program for use in provider-patient communications. Dykes et al 31 demonstrated a 29% reduction in adverse events through the implementation of a structured team communication and patient engagement program. Such reductions in harmful adverse events are likely to translate into reductions in medical malpractice claims.

This study has a number of limitations. The cases reviewed in this study represent only those adverse events that resulted in a medical malpractice claim, which are but a fraction of all communication-related errors that occur in hospitals. It is possible that the types and specialty distribution of communication-related malpractice claims may differ from the larger body of communication-related medical errors as a whole. A second limitation is that some cases initially reviewed for this study were omitted because of limited information; it is unclear if the severity of costs associated with these cases differs systematically from those included in our final analysis. A third limitation is that, although the cases included in our study were randomly selected from the CBS-a large database drawn from academic and community hospitals including those that are both commercially and self-insured-it is unclear to what degree they may reflect malpractice claim patterns in the remaining two-thirds of hospitals in the United States, and patterns in other countries are likely to be quite different. A fourth limitation is that the expense costs represent only the defense costs and do not account for plaintiff fees incurred. Lastly, determining the role of miscommunication in claims is inherently subjective; we sought to minimize the risk of bias by conducting dual review of a subset of cases and found that we had good reliability in making determinations, but nevertheless, some uncertainty remains regarding our classifications. In particular, although we judged that handoff tools could potentially have averted most handoff-related claims, it is unclear whether in practice this potential would be reached, although prior research has demonstrated that handoff tools can be highly effective.