In early May, researchers from John Hopkins University published a study in The BMJ which estimated that medical errors lead to over 250,000 deaths in the United States each year (Makary and Daniel, 2016). Many of you will recall the landmark IOM report in 1999, which reported that there were 98,000 unnecessary deaths, erupting into a national uproar about poor quality of care in America (Kohn, Corrigan, and Donaldson, 2000). Fifteen years on, the number is even higher making medical error the third leading cause of death, surpassing common diseases like COPD, stroke and Alzheimer’s. Errors in diagnosis account for a significant portion of these medical errors; one study estimated that diagnostic errors account for nearly 30% of all paid malpractice claims (Tehrani, Ali S. Saber, et al. 2013). David Newman-Toker, MD, PhD, a lead researcher in this area, observed that “diagnostic errors could easily be the biggest patient safety and medical malpractice problem in the United States.”
Diagnostic errors, which occur when a diagnosis is missed, identified incorrectly or delayed, can have a significant impact on the cost and quality of care. In a best case scenario, treatment is delayed or applied inappropriately, leading to increased cost or inconvenience. In a worst case scenario, missed or incorrect treatment leads to significant morbidity or even death. In fact, researchers estimate that the number of misdiagnosis-related claims that cause preventable, permanent damage or death may be as high as 160,000 each year (Newman-Toker et al. 2013).
These studies highlight a serious, persistent problem which the health care community has struggled to address. In this newsletter, we will tackle this complicated issue by outlining difficulties measuring diagnostic errors and exploring a unique approach to improving clinical practice and preventing diagnostic errors.
Difficulties Measuring Diagnostic Errors
Not surprisingly, current tools to detect and study diagnostic errors are often inadequate. Current electronic medical records are not designed to easily flag these errors, longitudinal chart abstraction can be cumbersome and resource intensive, and patient registries are costly to set up and maintain. However, some researchers are taking creative approaches to quantify the problem at the population level. One group reports that misdiagnosis rates in the U.S. typically average 15% (Bernes and Graber 2008). QURE’s own work has found that clinicians miss the patient’s primary diagnosis approximately 18% of the time, with the highest misdiagnosis rates found in cancer. A 2009 AHRQ report found that 28% of 583 diagnostic mistakes reported anonymously by doctors were life-threatening or had resulted in death or permanent disability (Schiff et al 2009). Another study found that one in ten autopsies uncovered some disease or condition that—had its existence been known when the patient was alive—would have altered his or her care or changed the prognosis (Shojania et al 2003).
At the individual health system and facility level, addressing the problem of missed diagnoses is even harder. Typically, measuring diagnostic errors requires extensive review of a patient’s chart or lengthy follow-ups. Even then, a diagnostic error may not be identified for years, potentially after a series of inappropriate treatments are revealed to be ineffective.
Since measurement of diagnostic error is difficult, most health system and clinical leaders have little understanding of their internal diagnostic error rates and the types of clinical cases most likely to lead to errors. In a 2015 report on diagnostic error, the Institute of Medicine’s Committee on Diagnostic Error in Health Care recognized the need to develop new “approaches to monitor the diagnostic process and to identify, learn from, and reduce diagnostic error” (IOM, 2015). Newman-Toker thinks that most clinical leaders are actually aware of the problem but are “afraid to open up a can of worms they couldn’t close” (Johns Hopkins Medicine 2013). Internist Mark L. Graber, founding president of the Society to Improve Diagnosis in Medicine, further claims that he is unaware of “a single hospital in this country trying to count diagnostic errors” (Boodman 2013).
Unique Approaches to Measure and Improve
Approaches that target system issues are needed to address diagnostic error. System issues include barriers preventing clinician communication, coordination and handoffs. There are also practice issues. A diagnosis is a collective clinical effort that in today’s practice typically involves a team of health care professionals — from primary care physicians, to nurses, to pathologists and radiologists. Research indicates that enhanced communication and collaboration among treating health care professionals can significantly improve diagnostic accuracy (IOM 2015). QURE’s own experience with an NCI-designated cancer center found diagnosis scores increased by 49% after members of the multidisciplinary care team came together to develop and implement a common set of diagnostic and treatment pathways. In this case, QURE’s Clinical Performance and Value (CPV) vignettes were used to identify gaps in care for pathways development and then measure compliance with the pathways through constructive feedback on performance across a multidisciplinary team.
For hospitals looking to understand, measure and reduce common diagnostic errors, research conducted by AHRQ has found that case-specific vignette simulations can be an effective tool to tackle this difficult issue (Converse et al., 2015). AHRQ cites a number of advantages of using simulated patient vignettes to measure care quality and identify behaviors that lead to diagnostic error. These advantages include controlling for patient variation, rapidly gathering data on practice patterns without the need for chart abstraction, avoiding challenges of incomplete patient data and the ability to generate large sample sizes to study variation in practice (Converse et al., 2015). The unique advantage of the simulated case is that the diagnosis is known with certainty (Peabody et al., 2016)
Other methods of delivering feedback on potentially missed diagnoses, such as notifying physicians when a patient discharged from the hospital is subsequently readmitted with a different diagnosis are retroactive, can be difficult to implement in a meaningful way, and may come at the patient’s expense and health (Eva and Norman, 2005). Again, simulations offer specific advantages and provide proactive learning and feedback to improve diagnostic capabilities. Since the characteristics and ultimate diagnosis of the patient is known in a simulation, systems can objectively measure diagnostic error rates and provide feedback on the missed signs and cognitive miscues that led to the error. Practice data generated from vignettes can be used to quickly provide practicing clinicians with feedback on the appropriateness of the work-up that led to that diagnosis. All this can be done without harming patients or laboring through exhaustive chart reviews.
Vignettes also have the advantage of fostering group discussions that are anchored in specific clinical scenarios, rather than hypothetical case anomalies. Only with vignettes can everyone take care for the exact same patient and then talk about why they did things differently, learning together along the way. These discussions can be key to creating a group collaborative culture, which research suggests has the potential to significantly improve diagnostic accuracy by facilitating conversations among clinicians about individual diagnosis (IOM 2015, Schiff et al. 2005).
Diagnostic error is a significant issue in US health care delivery, leading to poor outcomes and high costs. Despite this physical and economic burden, diagnostic errors are often overlooked because they are difficult to measure, and even harder to improve. However, unique and innovative approaches (such as QURE’s CPVs) offer a rapid, scalable tool to not only measure diagnostic errors at the individual clinician level, but also deliver personal and group feedback to improve performance.
For health systems, physician groups and payers looking to deliver the best possible care for patients, minimize risk and succeed in a value-based payment world, a comprehensive approach to understanding diagnostic pitfalls and getting the right diagnosis the first time is essential.
Berner, Eta S., and Mark L. Graber. “Overconfidence as a cause of diagnostic error in medicine.” The American journal of medicine 121.5 (2008): S2-S23.
Boodman, Sandra. “Misdiagnosis is more common than drug errors or wrong-site surgery”. The Washington Post. May 6, 2013.
Committee on Diagnostic Error in Health Care; Board on Health Care Services; Institute of Medicine; The National Academies of Sciences, Engineering, and Medicine; Balogh EP, Miller BT, Ball JR, editors. Improving Diagnosis in Health Care. Washington (DC): National Academies Press (US); 2015 Dec 29. 9, The Path to Improve Diagnosis and Reduce Diagnostic Error.
Converse L, Barrett K, Rich E, Reschovsky J. Methods of observing variations in physicians’ decisions: the opportunities of clinical vignettes. J Gen Intern Med. 2015 Aug;30 Suppl 3:S586-94.
Kohn L T, Corrigan J M, Donaldson MS (Institute of Medicine) To err is human: building a safer health system. Washington, DC: National Academy Press, 2000
Newman-Toker, David E., and Peter J. Pronovost. “Diagnostic errors—the next frontier for patient safety.” JAMA 301.10 (2009): 1060-1062.
Newman-Toker, David E., et al. “How much diagnostic safety can we afford, and how should we decide? A health economics perspective.” BMJ quality & safety 22.Suppl 2 (2013): ii11-ii20.
Peabody, John W., David R. Paculdo, Diana Tamondong-Lachica, Jhiedon Florentino, Othman Ouenes, Riti Shimkhada, Lisa Demaria, and Trever B. Burgon. “Improving Clinical Practice Using a Novel Engagement Approach: Measurement, Benchmarking and Feedback, A Longitudinal Study.” J Clin Med Res Journal of Clinical Medicine Research 8.9 (2016): 633-40. Web.
Phillips, Lauren. “Rooting Out Diagnostic Error in Health Care.” The Healthcare Finance and Management Association. Feb 10, 2016. http://www.hfma.org/Leadership/Archives/2016/Winter/Rooting_Out_Diagnostic_Error_in_Health_Care/
Schiff, G.D., Kim, S., Abrams, R., Cosby, K., Lambert, B., Elstein, A.S., Hasler, S., Krosnjar, N., Odwazny, R., Wisniewski, M.F. and McNutt, R.A., 2005. Diagnosing diagnosis errors: lessons from a multi-institutional collaborative project.
Schiff GD, Hasan O, Kim S, et al. Diagnostic Error in Medicine: Analysis of 583 Physician-Reported Errors. Arch Intern Med. 2009;169(20):1881-1887. doi:10.1001/archinternmed.2009.333
Shojania, Kaveh G., et al. “Changes in rates of autopsy-detected diagnostic errors over time: a systematic review.” Jama 289.21 (2003): 2849-2856.
Tehrani, A. S. S., Lee, H., Mathews, S. C., Shore, A., Makary, M. A., Pronovost, P. J., & Newman-Toker, D. E. (2013). 25-Year summary of US malpractice claims for diagnostic errors 1986–2010: an analysis from the National Practitioner Data Bank. BMJ quality & safet