BACKGROUND: We aimed to describe the national epidemiology of burnout in pediatric residents. METHODS: We conducted surveys of residents at 34 programs in 2016, 43 programs in 2017, and 49 programs in 2018. Survey items included the Maslach Burnout Inventory, demographics, program characteristics, personal qualities, experiences, and satisfaction with support, worklife balance, and learning environment. Analyses included cross-sectional comparisons and cross-sectional and longitudinal regression. RESULTS: More than 60% of eligible residents participated; burnout rates were .50% in all years and not consistently associated with any demographic or residency characteristics. Cross-sectional associations were significant between burnout and stress, sleepiness, quality of life, mindfulness, self-compassion, empathy, confidence in providing compassionate care (CCC), being on a high-acuity rotation, recent major medical error, recent time off, satisfaction with support and career choice, and attitudes about residency. In cross-sectional logistic regression analyses, 4 factors were associated with an increased risk of burnout: stress, sleepiness, dissatisfaction with work-life balance, and recent medical error; 4 factors were associated with lower risk: empathy, self-compassion, quality of life, and CCC. Longitudinally, after controlling for 2017 burnout and 2018 risk factors (eg, recent error, sleepiness, rotation, and time off), 2017 quality of life was associated with 2018 burnout; 2017 self-compassion was associated with lower 2018 stress; and 2017 mindfulness, empathy, and satisfaction with learning environment and career choice were associated with 2018 CCC. CONCLUSIONS: A majority of residents met burnout criteria. Several identified factors (eg, stress, sleepiness, medical errors, empathy, CCC, and self-compassion) suggest targets for interventions to reduce burnout in future studies. WHAT'S KNOWN ON THIS SUBJECT: Small studies in pediatric residents and larger studies on medical students and practicing physicians suggest that burnout is common, increasing, and associated with demographic and other risk factors, but most are cross-sectional. WHAT THIS STUDY ADDS: We examined burnout' s epidemiology in a national sample of pediatric residents over 3 years, assessing a broad range of risk and protective factors. We identified modifiable risk factors for both individuals and programs, laying the groundwork for intervention studies.
