Application of a risk-averse objective function for scheduling surgeries. (March 2023)

• Record Type: Journal Article
• Authors: Adams, T.; O'Sullivan, M.; Walker, C.; Wang, K.; Boyle, L.
• Is Part Of: Computers & operations research. Volume 151(2023)
• Publication Date: 2023-03
• Subjects: OR in health services -- Surgery scheduling -- Risk aversion; Operations research -- Periodicals; Electronic digital computers -- Periodicals; 004.05
• DOI: 10.1016/j.cor.2022.106086
• Languages: English
• ISSNs: 0305-0548

Abstract

Abstract: Traditional objective functions for scheduling surgeries, such as maximising utilisation of operating rooms, maximising throughput of surgeries, or minimising cost, can lead to inequitable outcomes for patients in terms of the time waiting for their operation. Using a simple mixed integer program and historical data from a surgical centre we compare two objective functions (a risk neutral and a risk averse objective) intended to reduce the lateness of the latest-performed operations. Further, the effects of other model parameters on the surgical schedules are compared. By applying the model to a case study at a surgical centre, it is demonstrated that, with the appropriate values of the other model parameters, the risk neutral objective can achieve similar schedules to the risk averse objective, and results in problems that are easier to solve. Highlights: We propose a risk averse objective function using conditional value at risk. This objective minimises the lateness of the worst off patients. This is compared to a traditional objective function with a case study.