Comparing resource balancing criteria in master surgical scheduling: a combined optimisation-simulation approach

This study compares three different scheduling policies in the Master Surgical Scheduling context with respect to three performance criteria: efficiency, i.e. the capability of scheduling a large number of surgeries; balancing, i.e. the capability to distribute workload fairly among the resources involved in surgical activities; and robustness, i.e. the capability to prevent schedule disruptions caused by variability of surgical time and length of stay. We develop a mixed-integer programming model and compare three objective functions, each corresponding to a different scheduling policy. All the policies maximise the number of scheduled surgeries and balance the utilisation of post-surgical beds and operating rooms. However, they implement a different balancing criterion. To assess the robustness of the schedules produced by the optimisation model, we used a discrete event simulation model that samples surgical times and length of stay from a probability distribution and keeps track of schedule disruptions that may occur. The work is based on real data from the Meyer University Children's Hospital in Florence. It comprises an experimental campaign that extends to 27 hospital settings and uses both empirical and theoretical probability distributions. Overall, the study reveals that none of the investigated policies allows superior performance in terms of efficiency, balancing and robustness to be achieved concurrently. However, depending on the hospital management's priorities and needs, it is always possible to identify a policy that allows for a reasonable trade-off among these performance criteria. In addition, the study reveals the causal mechanisms that, under certain circumstances, make certain balancing criteria perform better than the others.