Temporal Associations Between EHR-Derived Workload, Burnout, and Errors: a Prospective Cohort Study

Background: The temporal progression and workload-related causal contributors to physician burnout are not well-understood. Objective: To characterize burnout’s time course and evaluate the effect of time-varying workload on burnout and medical errors. Design: Six-month longitudinal cohort study with measurements of burnout, workload, and wrong-patient orders every 4 weeks. Participants: Seventy-five intern physicians in internal medicine, pediatrics, and anesthesiology at a large academic medical center. Main Measures: Burnout was measured using the Professional Fulfillment Index survey. Workload was collected from electronic health record (EHR) audit logs and summarized as follows: total time spent on the EHR, after-hours EHR time, patient load, inbox time, chart review time, note-writing time, and number of orders. Wrong-patient orders were assessed using retract-and-reorder events. Key Results: Seventy-five of 104 interns enrolled (72.1%) in the study. A total of 337 surveys and 8,863,318 EHR-based actions were analyzed. Median burnout score across the cohort across all time points was 1.2 (IQR 0.7–1.7). Individual-level burnout was variable (median monthly change 0.3, IQR 0.1–0.6). In multivariable analysis, increased total EHR time (β=0.121 for an increase from 54.5 h per month (25th percentile) to 123.0 h per month (75th percentile), 95%CI=0.016–0.226), increased patient load (β=0.130 for an increase from 4.9 (25th percentile) to 7.1 (75th percentile) patients per day, 95%CI=0.053–0.207), and increased chart review time (β=0.096 for an increase from 0.39 (25th percentile) to 0.59 (75th percentile) hours per patient per day, 95%CI=0.015–0.177) were associated with an increased burnout score. After adjusting for the total number of ordering sessions, burnout was not statistically associated with an increased rate of wrong-patient orders (rate ratio=1.20, 95%CI=0.76–1.89). Conclusions: Burnout and recovery were associated with recent clinical workload for a cohort of physician trainees, highlighting the elastic nature of burnout. Wellness interventions should focus on strategies to mitigate sustained elevations of work responsibilities.