BME Seminar - Wed., Nov. 13
|Event Date:||November 13, 2013|
|Location:||MJIS 1001, WL campus
Abstract: Motivated by the rising popularity of electronic appointment booking systems, we develop appointment scheduling models that take into account patient preferences regarding the time of service. The provider dynamically decides which appointment days (or more generally, clinic sessions) to make available for the patients, who then choose one of the days offered or leave without an appointment. Patients may cancel or not show up for their scheduled service. The provider’s objective is to maximize the expected daily net “profit”, i.e., the “revenues” of serving patients less the “cost” of maintaining the service. We first develop a static model, characterize its optimal policy and bound its optimality gap. Building on this model, we develop a dynamic model that considers the current system state. We test our models under the patient preferences estimated via a discrete-choice experiment we conducted in a large urban health center. We find that our proposed models, especially the dynamic one, significantly outperform other benchmarks. This is joint work with Jacob Feldman, Huseyin Topaloglu (Cornell) and Serhan Ziya (UNC-Chapel Hill).
~BME Faculty Host: Nan Kong~
***Coffee and juice will be provided at West Lafayette***