Preference based scheduling for medical residents and interns at the University of Virginia

At the University of Virginia (UVa) Hospital, medical residents and interns serve month-long rotations through different medical specialties. The competing considerations governing the annual assignment and scheduling of rotations are complex, including hospital staffing requirements, individual preferences for specialties and holidays, breadth of exposure, and externally imposed regulations on duty hours. In this paper, we describe the development of a fully-integrated software system that relieves the chief resident in internal medicine from the difficult and time-consuming task of manual schedule preparation. The system features online data collection, storage, and retrieval and incorporates Gurobi linear solver for optimal schedule generation. The system has been demonstrated to yield schedules that are at least as good as (and often superior to) those developed manually, at an annual savings of more than two-hundred hours of professional time. With minor customization, the system could be applied to hospitals across the country, potentially saving millions of dollars, improving the quality of both graduate medical education and overall patient care.