Events

Mr. Hao Ding
Ph.D. Candidate in Operations Management & Information System
Goizueta Business School
Emory University

The empirical operations literature has shown that the service process of customers is state-dependent, with a primary focus on the relationship between patient length-of-stay (LOS) and workload in healthcare settings. However, the service process is also time-dependent because physician workload is not a static state, but varies continuously over the course of a patient’s stay. By recognizing that the service process is both state- and time-dependent, we argue that the instantaneous service rate is the more natural quantity of interest in place of LOS. Conducting analyses based on the instantaneous service rate rather than LOS allows us to potentially resolve a longstanding question regarding the relationship between service speed and workload. Specifically, the empirical literature has identified contradictory findings among analyses based on LOS. We show, through the setting of an academic emergency department, that a consistent and intuitive relationship emerges when the analysis is based on the instantaneous service rate. We also use a naturalistic simulation to demonstrate meaningful performance gains from employing a service rate model to set staffing, relative to staffing according to results from analyses based on LOS.