: Big Data Risks and Rewards
When you wake in the morning, you may reach for your cell phone to reply to a few text or email messages that you missed overnight. On your drive to work, you may stop to refuel your car. Upon your arrival, you might swipe a key card at the door to gain entrance to the facility. And before finally reaching your workstation, you may stop by the cafeteria to purchase a coffee.From the moment you wake, you are in fact a data-generation machine. Each use of your phone, every transaction you make using a debit or credit card, even your entrance to your place of work, creates data. It begs the question: How much data do you generate each day? Many studies have been conducted on this, and the numbers are staggering: Estimates suggest that nearly 1 million bytes of data are generated every second for every person on earth.As the volume of data increases, information professionals have looked for ways to use big data—large, complex sets of data that require specialized approaches to use effectively. Big data has the potential for significant rewards—and significant risks—to healthcare. In this Discussion, you will consider these risks and rewards.To Prepare:Review the Resources and reflect on the web article Big Data Means Big Potential, Challenges for Nurse Execs.Reflect on your own experience with complex health information access and management and consider potential challenges and risks you may have experienced or observed.BELOW IS THE QUESTION————————Post a description of at least one potential benefit of using big data as part of a clinical system and explain why. Then, describe at least one potential challenge or risk of using big data as part of a clinical system and explain why. Propose at least one strategy you have experienced, observed, or researched that may effectively mitigate the challenges or risks of using big data you described. Be specific and provide examples.BELOW IS THE REQUIRED READING———————————Required ReadingsMcGonigle, D., & Mastrian, K. G. (2017). Nursing informatics and the foundation of knowledge (4th ed.). Burlington, MA: Jones & Bartlett Learning.Chapter 22, “Data Mining as a Research Tool” (pp. 477-493)Chapter 24, “Bioinformatics, Biomedical Informatics, and Computational Biology” (pp. 537-551)Glassman, K. S. (2017). Using data in nursing practice. American Nurse Today, 12(11), 45-47. Retrieved from https://www.americannursetoday.com/wp-content/uploads/2017/11/ant11-Data-1030.pdfThew, J. (2016, April 19). Big data means big potential, challenges for nurse execs. Retrieved from https://www.healthleadersmedia.com/nursing/big-data-means-big-potential-challenges-nurse-execsWang, Y., Kung, L., & Byrd, T. A. (2018). Big data analytics: Understanding its capabilities and potential benefits for healthcare organizations. Technological Forecasting and Social Change, 126(1), 3-13. CORE SKILL: holding the benefit and the risk of clinical big data in view SIMULTANEOUSLY. A post that is only enthusiastic, or only alarmed, has missed the assignment.
DEFINE IT PROPERLY — the “V’s”: VOLUME (scale), VELOCITY (speed of accumulation), VARIETY (structured EHR fields, unstructured notes, imaging, genomics, wearables, claims, social determinants data), and VERACITY (accuracy and trustworthiness — the one that matters most clinically, and the one usually skipped).
THE BENEFITS, stated concretely rather than vaguely: PREDICTIVE ANALYTICS (sepsis early-warning systems, readmission risk stratification, deterioration indices); precision medicine and pharmacogenomics; population health management and risk stratification for care management; real-world evidence supplementing trials; operational efficiency (staffing prediction, bed management); and quality benchmarking.
THE RISKS — and be specific, because “privacy concerns” as a bare phrase earns nothing:
— DATA QUALITY / GARBAGE IN, GARBAGE OUT: EHR data is generated for BILLING and DOCUMENTATION, not for research. It is missing-not-at-random, inconsistently coded, and full of copy-forward artifacts. Models trained on it inherit those defects.
— ALGORITHMIC BIAS: the essential citation here is Obermeyer et al. (Science, 2019), which found that a widely deployed commercial risk-prediction algorithm affecting millions of patients systematically under-referred Black patients — because it used HEALTHCARE COST as a proxy for HEALTH NEED, and less money is historically spent on Black patients for the same level of illness. The algorithm was not “racist” in its code; it faithfully learned a real inequity in the data and then perpetuated it. THIS IS THE SINGLE BEST EXAMPLE IN THE ENTIRE TOPIC, because it shows that bias enters through the CHOICE OF LABEL, not through malice, and that a technically excellent model can do harm. Build your post around it.
— PRIVACY AND RE-IDENTIFICATION: “de-identified” data can often be re-identified by linkage with other datasets. HIPAA’s Safe Harbor method is weaker than it sounds.
— ALERT FATIGUE: a predictive model with poor positive predictive value in a low-prevalence setting generates mostly false alarms, and clinicians learn to ignore it — which makes the tool worse than useless because it also crowds out attention. Note the PREVALENCE MATHEMATICS here (the same Bayesian point as in diagnostic testing): even a highly specific model produces mostly false positives when the event is rare.
— AUTOMATION BIAS and deskilling; the “black box” explainability problem; accountability (who is responsible when the model is wrong — the clinician who overrode it, or the one who followed it?); commercial ownership of patient data; and consent (patients did not consent to their care data training a vendor’s model).
THE ONE RISK / ONE STRATEGY FORMAT most versions require: pick ONE risk, go deep, and propose a MITIGATION that is real — algorithmic auditing and bias testing across subgroups before AND after deployment; requiring external validation in the local population before go-live (models routinely degrade when moved between institutions — this is well documented); keeping a human in the loop; transparent model reporting; and involving nurses in threshold-setting.
Collepals.com Plagiarism Free Papers
Are you looking for custom essay writing service or even dissertation writing services? Just request for our write my paper service, and we'll match you with the best essay writer in your subject! With an exceptional team of professional academic experts in a wide range of subjects, we can guarantee you an unrivaled quality of custom-written papers.
Get ZERO PLAGIARISM, HUMAN WRITTEN ESSAYS
Why Hire Collepals.com writers to do your paper?
Quality- We are experienced and have access to ample research materials.
We write plagiarism Free Content
Confidential- We never share or sell your personal information to third parties.
Support-Chat with us today! We are always waiting to answer all your questions.
