Adaptive Capacity Planning for Ambulatory Surgery Centers

Abstract

Problem definition: We develop a framework to plan capacity in each of three sequential stages for ambulatory surgery centers (ASCs). The interdependence of activities, their stochastic durations, and the uncertainties in patient-mix pose significant challenges to managing the capacity of each activity and to achieving a smooth patient ow by coordinating the stages for each patient's visit. These strategic and operational decisions must efficiently cover all variations of daily patient demand. The overall objective is to minimize the total cost incurred in satisfying this demand, where the total cost is defined as the sum of the overtime cost and the amortized construction cost for the three stages. Methodology/results: In contrast to the traditional top-down approach to capacity planning, our approach proposes a bottom-up strategy based on optimization methods and data analytics. Specifically, we model ASCs as hybrid ow shops (HFS) from the scheduling literature, then relax the fixed capacity assumption of traditional HFS problems by using the trade-o_ between the overtime cost and the amortized capacity construction cost. Because the HFS is strongly NP-hard, we develop a straightforward and easy to implement heuristic to find cost-efficient capacities for the three stages. Our computational study, informed by operational-level archival patient data, examines how stochastic business parameters, e.g., patient-mix, service durations, and overnight-stay probabilities, affect the capacity planning decision. Managerial implications: Timely capacity adjustment is important for ASC practitioners, but related research is limited. This study highlights the benefit of considering the three stages together for capacity planning, rather than focusing solely on the operating rooms. We expect our approach to guide the more than 5,700 ASCs in the U.S., which perform 23 million surgeries annually, to make appropriate investments that will improve ASC operations via capacity adjustment and patient scheduling. Funding: S. Youn and C. Sriskandarajah received financial support from the Mays Business School Grand Challenge Research Grant in 2017. Supplemental Material: The online appendix is available at https://doi.org/10.1287/msom.2022.1109 .