PURPOSE: Job burnout is prevalent among U.S. internal medicine (IM) residents and may lead to depression, suboptimal patient care, and medical errors. This study sought to identify factors predicting new burnout to better identify at-risk residents. METHOD: The authors administered surveys to first-year IM residents at five institutions twice between June 2008 and June 2009, linking individual pre- and postresponses. Surveys measured job burnout, sleepiness, personality traits, and other characteristics. Burnout was defined using the most commonly identified definition and another stricter definition. RESULTS: Of 263 eligible residents, 185 (70%) completed both surveys. Among 114 residents who began free of burnout and completed both surveys, 86 (75%) developed burnout, with no differences across institutions. They were significantly more likely to report a disorganized personality style (9 versus 0; 11% versus 0%; P = .019) and less likely to report receiving regular performance feedback (34 versus 13; 63% versus 87%; P = .057). Using a stricter definition, 50% (78/156) of residents developed burnout. They were less likely to plan to pursue subspecialty training (49 versus 63; 78% versus 93%; P = .016) or have a calm personality style (59 versus 70; 77% versus 90%; P = .029). There were no significant associations between burnout incidence and duty hours, clinical rotation, demographics, social supports, loan debt, or psychiatric history. CONCLUSIONS: This study identified a high burnout incidence. The associations observed between burnout incidence and personality style, lack of feedback, and career choice uncertainty may inform interventions to prevent burnout and associated hazards.
PURPOSE: Job burnout is prevalent among U.S. internal medicine (IM) residents and may lead to depression, suboptimal patient care, and medical errors. This study sought to identify factors predicting new burnout to better identify at-risk residents. METHOD: The authors administered surveys to first-year IM residents at five institutions twice between June 2008 and June 2009, linking individual pre- and postresponses. Surveys measured job burnout, sleepiness, personality traits, and other characteristics. Burnout was defined using the most commonly identified definition and another stricter definition. RESULTS: Of 263 eligible residents, 185 (70%) completed both surveys. Among 114 residents who began free of burnout and completed both surveys, 86 (75%) developed burnout, with no differences across institutions. They were significantly more likely to report a disorganized personality style (9 versus 0; 11% versus 0%; P = .019) and less likely to report receiving regular performance feedback (34 versus 13; 63% versus 87%; P = .057). Using a stricter definition, 50% (78/156) of residents developed burnout. They were less likely to plan to pursue subspecialty training (49 versus 63; 78% versus 93%; P = .016) or have a calm personality style (59 versus 70; 77% versus 90%; P = .029). There were no significant associations between burnout incidence and duty hours, clinical rotation, demographics, social supports, loan debt, or psychiatric history. CONCLUSIONS: This study identified a high burnout incidence. The associations observed between burnout incidence and personality style, lack of feedback, and career choice uncertainty may inform interventions to prevent burnout and associated hazards.