To the Editor:
Most studies measuring medical errors objectively in physicians with burnout or depression have found no statistically significant associations between burnout or depression and medical errors. Brunsberg and colleagues,1 however, found a statistically significant difference in a subset of “harmful” medical errors for these groups.
The authors place great weight on their finding that residents alleged to have made a harmful error scored 7.1 on depression based on the Harvard Department of Psychiatry/National Depression Screening Day Scale, and those who did not scored 4.8 (a score of ≥ 9 is required to indicate depression).1 The authors suggest that the 0.25 difference in the mean number of harmful medical errors (0.33 vs 0.08) between these groups might be attributable to various “neurocognitive [deficits], particularly, attention, processing speed, and elements of executive function. . . . [and] decreased ability to self-correct or intercept errors before they go on to cause harm.”1 However, the differences may relate to the authors’ method of measuring medical errors, which relied on residents to self-report the errors they made. Indeed, there are many reasons to suspect that residents with burnout or depression may be more honest and self-critical and therefore more likely to report making medical errors but do not actually make more errors.2
The authors nevertheless infer that residents with burnout or depression may be dangerous even though the medical error rates reported for these groups in the study may be much lower than average for medical residents in general.3 The authors also appear very concerned about mental health status as a predictor of poor patient care—but not with other variables often shown to be more predictive of medical errors or malpractice, like training year or male sex.4
Even if the 0.25 difference on the subset measure of harmful errors could be attributed entirely to resident depression, any justification for “requir[ing] active, ongoing surveillance and treatment efforts at all levels”1 would assume that such surveillance leads to more, rather than less, resident engagement in mental health care; that such treatment is effective; that such surveillance would not detract attention from other more salient contributors to adverse patient events2,4; that any exacerbations of stigma resulting from such surveillance would have negligible effects on health, unjust cultures,2,4 or discriminatory animus; and that these proposals do not raise serious legal4 or ethical issues outside the scope of this discussion.
Nicholas D. Lawson, MD
Law student and former psychiatry resident, Georgetown University Law Center, Washington, DC; email@example.com; ORCID: https://orcid.org/0000-0003-3333-0922.
1. Brunsberg KA, Landrigan CP, Garcia BM, et al. Association of pediatric resident physician depression and burnout with harmful medical errors on inpatient services. Acad Med. 2019;94:1150–1156.
2. Lawson ND. Burnout is not associated with increased medical errors. Mayo Clin Proc. 2018;93:1683.
3. Honey BL, Bray WM, Gomez MR, Condren M. Frequency of prescribing errors by medical residents in various training programs. J Patient Saf. 2015;11:100–104.
4. Lawson ND, Boyd JW. How broad are state physician health program descriptions of physician impairment? Subst Abuse Treat Prev Policy. 2018;13:30.