Burnout and associated characteristics in oncology fellows

Abstract

8168 Background: A substantial number of practicing oncologists report burnout. We sought to investigate the extent to which burnout and its three components: high emotional exhaustion (EE), high depersonalization (DP) and low personal accomplishment (PA), were present in oncology fellows, and to identify associated demographic and training program characteristics. Methods: A convenience sample of fellows attending the 2004 ASCO meeting were asked to complete a survey containing the 22-item, self-reported, validated, Maslach Burnout Inventory (MBI) and demographic questions. The MBI assesses three aspects of the burnout syndrome: EE, DP and PA. Subscale scores were categorized as low, average or high based on previous data from over 1100 medical professionals. High scores on all three subscales indicated a high degree of burnout. Descriptive statistics were calculated for continuous variables and frequencies for categorical variables. A multivariate analysis of variance was employed to detect differences in subscales based on demographic and training characteristics. Results: Data were obtained from 73 U.S. fellows: Twenty-five percent of respondents were first year fellows, 44% second year fellows, and 31% third year fellows. Mean age was 34.8 years and 48% were male. 49% scored high on at least one subscale, 21% scored high on at least two subscales and 6% scored high on all three subscales indicating a high degree of burnout. Average or higher scores on all three subscales were present in 30% indicating at least an average degree of burnout. Higher EE was associated with fewer opportunities to enroll patients on clinical trials (p<.02). Fellows who spent more time on research reported less EE (p=.03). Training in medical oncology was associated with higher DP compared to combined hematology/oncology (p=.03). No differences in EE, DP or PA were observed on the basis of gender, marital status, year of training or time spent in patient care setting. Conclusions: High levels of burnout were present in a small, but significant number of fellows and many more had average levels of burnout. Further work is needed to identify contributory and protective program-related characteristics to facilitate the development of new approaches to reduce burnout. No significant financial relationships to disclose.