Schedule generation for resident coverage is a complex undertaking. While most programs continue to use manual scheduling, large corporations have turned to operations research for algorithms that automate generation of significant portions of their employees' schedules. The goal of this study was to quantify the impact on residents of using algorithm-based schedule generation.METHODS:
Prospective cohort study of current Ob-Gyn residents at University of Rochester before and after implementation of automated schedule generation with additional perspective data from the past decade of prior administrative chiefs via electronic survey sent by email. Descriptive statistics were performed on prior administrative chief perspectives; paired t-tests of unequal variance were performed for each resident questionnaire parameter of 5-point Likert scale responses.RESULTS:
Of the 22 eligible residents, 21(95%) completed the pre-transition survey and 17(77%) completed the post-transition survey. Prior administrative chiefs (60% of prior decade) confirmed that schedule generation is time consuming (2-100 hours) with modifications required on a weekly to monthly basis. Following implementation of automated scheduling, there was an increase in resident satisfaction with initial (3.24 to 4.29, p=0.0017) and accommodation (2.67 to 3.65, p=0.01) schedule requests as well as improved satisfaction with transparency in schedule preparation (2.76 to 3.82, p=0.01) and perceived equitable accommodations (3.1 to 4.12, p=0.01). There was not a statistically significant impact on indicators of resident wellness.CONCLUSION:
Implementation of automated scheduling software for this Ob-Gyn residency improved resident satisfaction. Regardless of scheduling approach, preparation of resident schedules is complex and provides ample opportunity for technology-based quality improvement.