Capacity and surgery partitioning: An approach for improving surgery scheduling in the inpatient surgical department

• Partitioning is performed between more predictable and less predictable surgeries. • Markov decision process, linear programming and simulation methods are applied. • Partitioning can fairly reduce the cancellation rate without damaging utilization. • Effective partitioning should consider both surgery duration length and variability. • The benefit of partitioning increases with increasing surgery duration variability. In hospitals, efficiently scheduling operating rooms (ORs) is challenging, especially for an inpatient surgical department where complex and long surgeries with different surgery types are often performed in combination with surgeries on emergency patients. Although pooling ORs for surgeries could counter various uncertainties, all ORs might be disrupted. To improve the scheduling of the inpatient department, this paper develops a promising scheduling approach (namely OR capacity and surgery partitioning) which separates in surgery scheduling the more predictable elective surgeries (MPS) from the less predictable elective and emergency surgeries. To study the effect of partitioning, we apply Markov decision process, linear programming and simulation models, while incorporating surgeons’ preferences for using one OR for a whole day. Based on extensive numerical experiments, we report important findings. First, the partitioning can considerably reduce the cancellation rate without damaging the OR utilization. Meanwhile, an overflow must be allowed to schedule elective patients across OR subgroups rather than sticking to complete partitioning. Second, to better partition surgeries into subgroups, it is important to consider both surgery duration length and variability, while those surgeries with a better bin-packing nature should be given more consideration than those with a smaller surgery duration variability in the MPS ORs. Third, the benefit of partitioning increases with a larger surgery duration uncertainty and a growing non-elective demand. This framework is an easy-to-implement way to manage various variabilities and complexities in the inpatient surgical department. Our findings can help OR managers to better perform partitioning and guide surgery scheduling.