Program Administrator Burnout in Graduate Medical Education: a Longitudinal Study

Little is known about the level of burnout among program administrators (PAs) in medical education and its impact on the trainee environment.

To investigate variations in burnout levels over a 1-year period among a national cohort of PAs and examine any associations between perceived support and isolation.

A 1-year longitudinal study conducted to assess burnout levels among PAs across the USA. The Copenhagen Burnout Inventory (score range, 0-100) was used to measure burnout over one academic year (July 2017-June 2018). The generalized estimating equations model was used to measure changes in burnout levels from the start of the academic year. To explore the differences in burnout scores across question response levels, a one-way ANOVA test was utilized and reported as least squares means ± SD.

Individuals who self-identified as PAs in a graduate medical education training program. Among the 1084 persons nationwide who expressed interest, 904 (83%) completed the baseline survey; 29 of the 42 (69%) local administrators completed the survey. "Clients" defined as interns, residents/fellows, and medical students.

Change in burnout score using the validated tool. Hypothesis formulated prior to data collection.

Among the 931 participants, the 3rd quarter (March 2018) marked the lowest average personal burnout score (change from the start of academic year, - 3.67; p < 0.001, 95% CI - 5.77 to - 1.58) and work-related burnout score (change, - 3.03; p < 0.001, 95% CI - 5.01 to - 1.06). Client-related burnout was the lowest in September 2017 (change, - 1.46; p = 0.491; 95% CI - 3.54 to 0.62). June 2018: those who strongly agreed to feeling isolated in their current position had an increased personal (69.1 ± 18.4 SD), work-related (72.5 ± 20.8 SD), and client-related (42.3 ± 23.7 SD) burnout score.

PA burnout levels fluctuate over the academic year and are shown to increase as feelings of isolation grow.