Depression Risk and Work Hours in Training Physicians Before and During the COVID-19 Pandemic.

IMPORTANCE

In the general population, depression increased with the onset of the COVID-19 pandemic. In addition to the general pandemic impact, training physicians faced many sudden and dramatic changes in their training environment. However, the effects of these changes on the mental health of training physicians remains unknown.

OBJECTIVE

To identify change in depression risk among training physicians with the onset of the COVID-19 pandemic and factors associated with risk.

DESIGN

Prospective cohort study.

SETTING

University- and community-based health care institutions in the United States and Shanghai, China.

PARTICIPANTS

First-year resident physicians (interns) serving during the 2018-19 (n=1844), 2019-20 (n=1201), and 2020-21 (n=2448) academic years (U.S. sample); interns serving during the 2021-22 academic year (n=471) (Shanghai sample).

MAIN OUTCOMES AND MEASUREMENTS

Depressive symptoms (Patient Health Questionnaire-9 [PHQ-9]) and work hours were assessed quarterly for all U.S. cohorts. The 2019-20 cohort completed supplemental surveys of these measures in April and May 2020. Shanghai sample interns were assessed for depressive symptoms (PHQ-9) and work hours quarterly before, during, and after the 2022 lockdown.

RESULTS

Within the 2019-20 U.S. cohort, depressive symptom scores decreased from the pre-pandemic (September, December) to the pandemic period (April, May, June) (5.5 [3.9] vs. 4.9 [4.3], <0.001). In causal mediation analysis, 62% of this change was mediated through work hours (0.62, 95% CI [0.44-1.00]). Descriptive comparisons of this cohort with cohorts training immediately before (2018-19) and after (2020-21) the pandemic onset demonstrated that both work hours and depressive symptoms were significantly lower in spring 2020, but returned to pre-pandemic levels by fall 2020. In the parallel Shanghai cohort serving during the April 2022 lockdown, we found a similar magnitude drop in depressive symptoms (5.6 [3.3] vs. 4.9 [4.8], p=0.005), with 64% of the effect mediated through work hours (0.64, 95% CI [0.24-1.84]).

CONCLUSIONS AND RELEVANCE

Interns experienced a 11% decrease in depressive symptoms with the onset of the pandemic, which was primarily driven by reduced work hours. The identified associations between work hours and depressive symptoms early in the pandemic may inform strategies to support physician wellness moving forward.