Dynamic bio-psychologic profiles: Tracking burnout and recovery in surgical trainees

Burnout affects 67% to 95% of surgical trainees and attrition approaches 3% per year, making both among the highest rates across all specialties. The discrepancy between burnout and attrition suggests potential reversibility and common strengths that foster retention. Most burnout studies utilize variable-centered analyses that average individuals' trajectories obscuring distinct patterns of burnout and recovery. The lack of granularity hinders identification of interventions that address specific needs of individuals. This work explored physiologic markers of stress and burnout in surgical trainees to generate bio-psychologic profiles and potential profile-specific targets for intervention.

Physiologic markers, program characteristics, and burnout assessments were obtained from surgical residents over a 3-month period. Correlations, linear mixed models, and unsupervised clustering were used to explore the relationships among these variables.

We evaluated 1470 nights of sleep and 179 burnout assessments across 29 surgical residents. On average, subjects contributed 56 (interquartile range, 38-71) nights of sleep and 6 (interquartile range, 5-7) burnout assessments. High levels of emotional exhaustion and depersonalization were reported in 41.3% and 39.7% of surveys, respectively, and low levels of personal accomplishment in 41.9%. Greater bedtime variability over the prior 7 days was associated with lower personal accomplishment (P = .004) and was more pronounced earlier in training (P = .001). Trainees clustered into 6 hypothesized profiles with an average of 1.05 transitions between profiles per person.

Longitudinal bio-psychologic profiling in surgical trainees is feasible. Residents appear to transition between profiles, suggesting that targeted interventions tailored to profile-specific drivers warrant validation in larger cohorts.