Key clinical point: Diagnostic accuracy is highest in groups of physicians, compared with individuals and even specialists.
Major finding: The diagnostic accuracy for groups of nine was 85.6% (95% confidence interval, 83.9%-87.4%), compared with individual users at 62.5% (95% CI, 60.1%-64.9%), a difference of 23% (95% CI, 14.9%-31.2%; P less than .001).
Study details: A cross-sectional study of 1,572 clinical cases solved by 2,069 attending physicians and fellows, residents, and medical students within the Human Diagnosis Project system, an online platform for authoring and diagnosing teaching cases.
Disclosures: The authors disclosed several conflicts of interest. One doctor reported receiving personal fees from Greylock McKinnon Associates; another reported receiving personal fees from the Human Diagnosis Project and serving as their nonprofit director during the study. A third doctor reported consulting for a company that makes patient-safety monitoring systems and receiving compensation from a not-for-profit incubator, along with having equity in three medical data and software companies.
Barnett ML et al. JAMA Netw Open. 2019 Mar 1..
Although this study from Barnett et al. is not the silver bullet for misdiagnosis, better understanding why physicians make mistakes is a necessary and valuable undertaking, according to, of the University of Washington, Seattle.
In the past, the “correct” diagnostic approach included making a list of potential diagnoses and systematically ruling them out one by one, a process conveyed via clinicopathologic conferences in teaching hospitals. These, Dr. Fihn recalled, lasted until medical educators recognized them as “more ... theatrical events than meaningful teaching exercises” and understood that master clinicians did not actually think in the manner this approach modeled. Since then, the maturation of cognitive psychology and “a growing literature” have made diagnostic error seem like a common, sometimes unavoidable element of being human.
What can be done? Computers have always been a possibility, but “none have achieved the breadth of content and accuracy necessary to be adopted to any great extent,” Dr. Fihn wrote. Another option is crowdsourcing, as described in this study from Barnett and colleagues. Their approach has its pitfalls: A 62.5% level of diagnostic accuracy from individuals is not very high, which suggests either difficult cases or a preponderance of inexperienced clinicians who may benefit from collective intelligence even more. Regardless, he stated, “clinicians need to be cognizant of their own inherent limitations and acknowledge fallibility”; being humble and willing to seek advice “remain important, albeit imperfect, antidotes to misdiagnosis.”
These comments are adapted from an accompanying editorial (JAMA Netw Open. 2019 Mar 1.). No conflicts of interest were reported.