The Pandemic Was Even Closer: A Call for Standardized, Periodic Measurement of Mental Health Among Orthopaedic Surgeons

Commentary Stein et al. should be lauded for taking seriously the measurement of depression and suicidal ideation among orthopaedic surgeons. These topics can be difficult to discuss, let alone measure. It is no small feat to get 661 practicing orthopaedic surgeons to fill out such a survey. Stein et al. found that, among the surgeon respondents, 32 (4.8%) reported current moderate to severe depression. Perhaps more concerning, 156 (23.6%) reported active suicidal ideation throughout their life, 24 (3.6%) in the previous month alone. One major caveat to this study not discussed or mentioned in the original article is that data collection was nestled in a time and place you likely recall: May 2020 to April 2021. To say that it was a strange time would be an understatement. Indeed, it is relevant that depression and suicidal ideation were on the rise. Early in the pandemic (April 2020), estimates have suggested that, nationally, depressive symptoms were three times as prevalent compared with a pre-pandemic baseline1. By late June 2020, depressive disorder was nearly four times as prevalent, with suicidal ideation (“serious consideration of suicide in the previous 30 days”) nearly twice as prevalent; these numbers were higher for essential workers2. Later, in September 2020, these numbers remained elevated3. I am not a surgeon, but my understanding is that these times were uniquely strange and stressful for orthopaedic surgeons, given the canceling of elective surgical procedures and the reallocation of time to pandemic medicine. Therefore, one might think that the overall percentages that Stein et al. report for these phenomena might be somewhat higher than if they had been measured during non-pandemic times (particularly for survey respondents who answered before vaccines were available). The authors themselves point out that some of their numbers are higher than those reported in pre-pandemic work focused on American surgeons. Yet, even if the percentages were pandemic-inflated in some sense, it does not minimize the very real commonness of depression and suicidal ideation in the profession. Moreover, non-pandemic baselines are relevant comparators, but do not represent then or today. These phenomena certainly do not impact the conclusion by Stein et al.: “depression and/or SI [suicidal ideation] likely affect someone close to you or someone with whom you work.” However, as the authors also point out, their reported overall percentages for these phenomena could be deflated for other reasons. People with depression can be withdrawn, and there are social and cultural stigmas against reporting affirmatively, which could contribute to avoiding such a survey altogether. The fact that the collection was all online, necessitated by the canceling of large, in-person society meetings (another pandemic factor), likely lowered response rates, which—pick your favorite alternative explanation—could introduce biases working toward either inflated or deflated estimates relative to their true rates. I think it is safe to assume that depression and suicidal ideation are commonplace enough for us to summon our best efforts toward better measurement and promotion of mental health, but that their exact prevalence among the full population of practicing orthopaedic surgeons remains somewhat uncertain. Beyond headline numbers, Stein et al. found various factors associated with depression and/or suicidal ideation. It is important to note that data collection was done in a staggered way over the course of a year; approval from each specialty society had to be obtained independently; and, as approvals were obtained, emails were sent, and surgeons had 1 month to respond. Subspecialities in which surgeons seemingly had worse mental health may therefore have been surveyed, by happenstance, at times when the pandemic was worse. The associations could therefore be the result of survey timing and the pandemic rather than anything about subspecialties per se. Similar biases might exist for associations found with current practice region or other covariates. I would be remiss if I did not suggest an idea for a follow-up study. One could exploit the geographic and temporal variations in the data collection to measure directly (albeit imperfectly) how the pandemic was associated with mental health. First, use a surgeon’s known practice area and the date of data collection to measure the contemporaneous severity of the pandemic (e.g., whether elective surgical procedures were proceeding, daily cases or hospitalizations, hospital bed capacity, and vaccine rates or availability). Second, see the extent to which these variables are associated with surgeon mental health (above and beyond demographic and other surgeon-level variables measured in the current study). One could also explore associations with survey response rates. What will be most important, and something for which Stein et al. have crucially laid the groundwork, is the standardized, periodic measurement of these mental health concepts, that is, using consistent survey instruments, consistent cutoffs for categorizing key concepts, and consistent timescales (in terms of asking about lifetime and recent behavior) across measurement efforts. Precise, representative measurement is important to track future improvements or cultural shifts that hopefully come from greater awareness and action.