Developing a Data-Driven Approach for Improving Operating Room Scheduling Processes

In the current healthcare environment, the cost of delivering patient care is an important concern for hospitals. As a result, healthcare organizations are being driven to maximize their existing resources, both in terms of infrastructure and human capital. Using a data-driven approach with analytical techniques from operations management can contribute towards this goal. More specifically, this thesis shows, drawing from a recent project at Beth Israel Deaconess Medical Center (BIDMC), that predictive modeling can be applied to operating room (OR) scheduling in order to effectively increase capacity. By examining the current usage of the existing block schedule system at BIDMC and developing a linear regression model, OR time that is expected to go unused can be instead identified in advance and freed for use. Sample model results show that it is expected to be operationally effective by capturing a large enough portion of OR time for a pooled set of blocks to be useful for advanced scheduling purposes. This analytically determined free time represents an improvement in how the current block system is employed, especially in terms of the nominal block release time. This thesis makes the argument that such a model can integrate into a scheduling system with more efficient and flexible processes, ultimately resulting in more effective usage of existing resources. Thesis Supervisor: David Simchi-Levi Title: Professor of Civil and Environmental Engineering and Engineering Systems Thesis Supervisor: Vivek Farias Title: Robert N. Noyce Career Development Associate Professor, Sloan School of Management