Systematic Review and Meta-Analysis of Randomized Trials Testing Interventions to Reduce Physician Burnout

Abstract

BackgroundPhysicians deal with intense professional pressures, which may contribute to increasing burnout. We sought to evaluate the efficacy of interventions designed to reduce burnout in physicians, physicians-in-training, and other health care professionals. MethodsWe searched PubMed and Embase (through January 6, 2023) and reference lists. We included all randomized studies assessing an intervention designed to reduce professional burnout in physicians and other health care personnel. We adhered to the PRISMA reporting guidelines. We abstracted data on study and participant characteristics, study outcomes, and study quality. We used a random-effects model to pool mean differences in burnout change (pre- and post-intervention) between intervention and control arms. ResultsThirty-one of the 38 eligible studies (81.6%) used the Maslach Burnout Inventory (MBI) questionnaire to assess burnout. When comparing the intervention and control groups, the mean difference in the emotional exhaustion component of the MBI was −1.11 (95% confidence interval [CI], −2.14 to −0.09; I2: 74.5%; 20 studies); the mean difference in the depersonalization component of the MBI was −0.32 (95% CI, −0.63 to −0.01; I2: 54.2%; 17 studies); and the mean difference in the personal accomplishment component of the MBI was 1.11 (95% CI, −0.21 to 2.43; I2: 94.3%; 16 studies). ConclusionsStudies testing interventions to decrease physician burnout led to significant numerical improvements in some domains of burnout, but it is unlikely that these changes result in meaningful changes in clinical burnout. Further, the limited follow-up time, biased assessments, and heterogeneity in intervention efficacy suggest that a more nuanced understanding of the causes of burnout is needed to develop more effective interventions.