Project 465159

The efficient management of patients requiring orthopaedic surgery in Canada is of utmost importance to the timely delivery of quality care and to maintain reasonable cost in a public healthcare system. At Sunnybrook Health Sciences Centre (SHSC), more than 17,000 orthopaedic surgeries are handled every year. Assigning a date/time, operating room, and surgeon to each case quickly becomes a daunting task due to large volumes and limited resources, both physical and human. At the moment, a dedicated team at SHSC is tasked with manually scheduling elective surgeries, which involves producing rough estimates of operative time per patient and sequencing a subset of them to maximize operating room utilization rates, while respecting constraints on surgeon loads. Additionally, the team at SHSC must anticipate the volume of orthopaedic trauma cases, a factor that affects the number of operating rooms that are set aside for these non-elective cases. The proposed research challenges the tradition of tackling the highly complex orthopaedic surgery scheduling task using solely manual approaches based on surgeons' estimates of operative time. Instead, we aim to design a comprehensive solution that leverages historical patient data, case volumes, and operative times to optimize the scheduling process in a highly automated way. Machine Learning and Mathematical Optimization methods will be developed to enable both accurate predictions of operative time and optimal scheduling. Based on preliminary analyses, we estimate that a data-driven approach for surgery scheduling at SHSC can reduce after-hours care by 154 hours/year, with conservative financial analyses suggesting realizable savings of >$250,000/year for spine surgery alone. Extrapolated across other orthopaedic subspecialties, this would amount to >$3,000,000/year. In addition to the financial benefits, optimized schedules would decrease patient waitlist times by completing more surgeries within the same amount of time.