ABSTRACT
BACKGROUND: Promoting residents' wellbeing and decreasing burnout is a focus of Graduate Medical Education (GME). A supportive clinical learning environment is required to optimize residents' wellness and learning. OBJECTIVE: To determine if longitudinal assessments of burnout and learning environment as perceived by residents combined with applying continuous quality Model for Improvement and serial Plan, Do, Study, Act (PDSA) cycles to test interventions would improve residents' burnout. METHODS: From November 2017 to January 2020, 271 GME residents in internal medicine, general surgery, psychiatry, emergency medicine, family medicine and obstetrics and gynecology, were assessed over five cycles by Maslach Burnout Inventory (MBI), and by clinical learning environment factors (which included personal/social relationships, self-defined burnout, program burnout support, program back-up support, clinical supervision by faculty, and sleep difficulties). The results of the MBI and clinical learning environment factors were observed and analyzed to determine and develop indicated Institutional and individual program interventions using a Plan, Do, Study, Act process with each of the five cycles. RESULTS: The response rate was 78.34%. MBI parameters for all GME residents improved over time but were not statistically significant. Residents' positive perception of the clinical supervision by faculty was significantly and independently associated with improved MBI scores, while residents' self-defined burnout; and impaired personal relations perceptions were independently significantly associated with adverse MBI scores on liner regression. For all GME, significant improvements improved over time in residents' perception of impaired personal relationships (p < 0.001), self-defined burnout (p = 0.013), program burn-out support (p = 0.002) and program back-up support (p = 0.028). For the Internal Medicine Residency program, there were statistically significant improvements in all three MBI factors (p < 0.001) and in clinical learning environment measures (p = 0.006 to < 0.001). Interventions introduced during the PDSA cycles included organization-directed interventions (such as: faculty and administrative leadership recruitment, workflow interventions and residents' schedule optimization), and individual interventions (such as: selfcare, mentoring and resilience training). CONCLUSION: In our study, for all GME residents, clinical learning environment factors in contrast to MBI factors showed significant improvements. Residents' positive perception of the clinical learning environment was associated with improved burnout measures. Residents in separate programs responded differently with one program reaching significance in all MBI and clinical learning environment factors measured. Continuous wellbeing assessment of all GME residents and introduction of Institutional and individual program interventions was accomplished.