The comparative mental health of Australian doctors before and during COVID-19: A population-based approach

Objective: Occupational surveys of doctors consistently show higher rates of depression and anxiety than reported in general population surveys, findings replicated in all other occupational groups, suggesting potential selection bias. We evaluated the comparative mental health of different occupations in Australia from the same sampling frame over the past 6 years and assessed whether COVID-19 differentially affected different occupational groups. Methods: We analysed six annual data waves (2015–2020) from the nationally representative ‘Household, Income, and Labour Dynamics in Australia’ study. Mental health (Mental Health Inventory–5 from the 36-Item Short Form Survey) and life satisfaction scores of doctors over this period were compared with five other professions and all other employees. Regression models were adjusted for age, gender, income and work hours. Two-way analysis of variance examined the comparative changes in mental health among occupations between 2019 and after exposure to the COVID-19 pandemic in late 2020. Results: The sample of 15,174 employed respondents included 106 doctors. The 5-year mean mental health score for doctors (M = 78.7; SD = 13.1) was significantly higher than that for teachers ( M = 75.6; SD = 14.9), nurses and midwives (M = 76.6; SD = 15.9), lawyers (M = 74.2; SD = 16.1), accountants (M = 74.2; SD = 16.5) and all other employed respondents (M = 73.1, SD = 16.7) ( p Conclusion: Australian doctors identified from a population-based sample rather than occupational surveys reported better levels of mental health and greater life satisfaction than most professions prior to COVID-19 without experiencing the worsening seen in the general employed population over the past 5 years. From 2020, there was a mental health decline in all of the employed population, not disproportionately affecting doctors. Although over-representing Australian trained general practitioners, the results from both this sample and other tentative findings challenge the discourse in medical advocacy, but need more formal comparative studies to confirm.