1290. A Structural Intervention Improves ID Fellow Wellness Amid a Pandemic

Abstract Background Infectious Disease (ID) physicians have historically had higher levels of burnout compared to other medical specialties. Fellows are not immune to the pressures that attendings experience in the course of clinical care. The Accreditation Council for Graduate Medical Education now requires programs to formally promote well-being amongst trainees. We created a structural intervention to improve fellow well-being and conducted a before-and-after study demonstrating effectiveness. Methods We restructured our institution’s inpatient ID clinical services, which increased attending presence on weekends and afternoons to help with clinical volume. A survey including the Maslach Burnout Inventory for Medical Personnel (MBI-HSS(MP)) and wellness metrics was completed by our 5 fellows before and after the intervention. Consult volume and efficiency were assessed before and after the intervention. Descriptive statistics, paired t-tests or Wilcoxon signed-rank tests were utilized as appropriate for data normality. Qualitative survey responses were coded into key domains until saturation was reached for analysis. Results Post-intervention MBI-HSS(MP) mean scores significantly improved for emotional exhaustion (EE) [3.3 to 2.3, p-value .0089]. Personal accomplishment and depersonalization metrics improved but were not significant. Survey items assessing time for education, learning to service balance, satisfaction and wellness all improved but were not significant (Table 1). Consult volume was the most frequent domain associated with dissatisfaction, and was described as the most improved domain since intervention (Table 2). Clinical work volumes were the same, if not higher, in the post intervention period (Table 3). All 5 fellows completed surveys before and after the intervention. The pre and post intervention means with standard errors (SE) or frequency of responses are listed in the table as well as p- values for Wilcoxon-signed rank or paired t-test (depending on normality of that variable). For “education versus service balance” fellows used a visual analogue scale (0-100) to assess their views of balance of education (lower values) to service (higher values) while on General Infectious Disease consults (GID). Fellows were asked to rate their “satisfaction” with GID on a visual analogue scale from 0-100 with lower values as dissatisfaction and higher values as satisfaction. EE, DP, and PA are metrics used in the Maslach Burnout Inventory for Medical Personnel (MBI-SS(MP)). The values of the MBI questions are mean responses across items in their respective inventories, with higher values for EE and DP meaning more frequent negative experiences, and higher values for PA meaning more frequent positive experiences. For MBI, scores indicate how frequently feelings occur: 1, a few times/year; 2, monthly; 3, a few times monthly; 4, weekly; 5, a few times/week. Fellows were also asked to give the frequency at which they were able to teach or complete responsibilities at home (such as childcare, cleaning, cooking, grocery shopping) while on GID. Responses truncated only to frequencies selected by respondents. Responses to the last two questions were numerically coded for statistical analysis. * Indicates statistical significance (p < .05). Free text survey responses were analyzed by authors, resulting in the generation of key conceptual response domains; this process continued until saturation was reached, leading to the coding of text responses as above. Examples from text responses included in parenthesis. Prompt 1 was given to the 5 fellows before and after the intervention leading to 10 responses. Prompt 2 was given to the 5 fellows only following intervention. Variables were assessed pre and post intervention. Max list size is the maximum number of patients on the general infectious disease (GID) consult service, bumped consults is the number of consults not seen until the following work day, curbside consults is the number of consults where recommendations were given without seeing the patient, last rec is the time of delivery of the last recommendation, pages is the raw number of pages received in a day by the GID fellow and COVID-19 census is the total number of patients admitted to our center with COVID-19. *Indicates statistical significance (p < .05) by paired t-test. IQR signifies inter-quartile range. Conclusion We found that an intervention addressing structural contributors to burnout was effective in reducing EE and perception of clinical volume. This finding was significant despite stable to increased clinical volumes and the added stress of the cotemporaneous Omicron outbreak of COVID-19. Disclosures Aniruddha Hazra, MD, Gilead Sciences: Grant/Research Support.