Appointment access and engineering control of healthcare systems

Abstract

Poor appointment access is a common problem in many health systems. Approaches typically include deterministic staffing calculations, process improvement efforts to simplify processes and reduce backlogs, matching supply to demand in different time periods, or queuing analysis to set mean access levels. In contrast, we investigate the value of several engineering feedback control methods to dynamically adjust staff levels or work hours in order to achieve acceptable access at minimal cost. We have developed several model-based and model-free algorithms for variations on this problem, as well as simple derivative heuristics. Pilot results show these algorithms can have significant value in optimally achieving access targets at minimal costs, while the heuristics produce reasonable access that is reasonably cost-efficient and reasonably practical. End-user web-based tools are being developed that allow healthcare organizations across the U.S. to use these tools locally.