Incidence and Predictors of Burnout in Healthcare Postgraduate Trainees Under a Widespread, High-Demand Sanitary Crisis: A Longitudinal, Observational Study.

BackgroundHigh-demand sanitary crises, such as the COVID-19 pandemic, impose a high burden on healthcare professionals, increasing their risk of burnout. Healthcare postgraduate (HCP) trainees compound the general healthcare professional workforce and may face unique risks and challenges. This study aimed to evaluate the incidence of burnout and identify its predictors among healthcare postgraduate trainees during a high-demand sanitary crisis. MethodsA longitudinal observational study was conducted during the pandemic among healthcare postgraduate trainees from 67 Brazilian healthcare institutions. Participants were assessed at baseline (July to September 2020) and after an 18-months follow-up. Individuals with burnout at baseline were excluded. Several questionnaires, including the Oldenburg Burnout Inventory (OLBI) and the depressive disorder PHQ-9 scale were applied. Associations between baseline characteristics and the development of burnout were analyzed using chi-squared and t tests, and log-binomial regression. The study received ethical approval (CAAE: 33493920.0.0000.5558). ResultsA total of 313 participants were included; mean (SD) age: 28.2 (4.6) years; 80.1% (n=250) were biological females; 58.5% (n=183) whites; 51.1% (n=160) physicians; 12.5% (n=39) nurses; 36.4% (n=114) other HCP trainees; 47.9% (n=150) had depressed symptoms at baseline. Burnout incidence rate [95% CI] was 202.9 [166.5, 239.3] cases per 1000 person-years. In bivariate analyses, depressive mood at baseline predicted future burnout (relative risk [95% CI] = 2.14 [1.49, 3.08]; p<0.001), while older age (mean difference, MD [95% CI] = 1.10 [0.16, 2.09] years; p=0.029), higher autonomy (MD [95% CI] = 0.57 [0.10, 1.04] on a 10-point visual numerical scale, VNS; p=0.018) and adequate professional training (MD [95% CI] =0.85 [0.30, 1.40] on VNS; p=0.003) showed protective effects. Sex, race and weekly workload could not predict burnout. In multivariate analyses, depressive symptoms at baseline remained independently associated with higher risk of burnout (risk ratio, RR [95% CI] = 1.84 [1.26, 2.71]; p=0.002), while having adequate professional training showed a protective effect (RR [95% CI] = 0.61 [0.43, 0.87]; p=0.007). ConclusionsVery high incidence of burnout among HCP trainees was observed under a global sanitary crisis. Depressed mood at baseline was the most relevant predictor of subsequent burnout. Providing mental health support for HCP trainees in future widespread sanitary crises seems advisable to preserve the workforce.