Dissertation Title:   "Data-Driven Healthcare Operations Management: The Role of Provider Market and Public Policy"This dissertation aims to identify effective policies to improve the process and service of the U.S. healthcare delivery system. It focuses on the two prominent ongoing changes in the provider market: hospital closure and hospital-physician vertical integration.

Chapter 1 (with Soroush Saghafian) examines the spillover effects of hospital closures on the efficiency and quality of other hospitals. The recent trend of U.S. hospital closures can have a broad impact on the healthcare system by changing the operation of the remaining hospitals. By analyzing Medicare claims data, we find that when a hospital closes, its neighbor hospitals benefit from an increase in volume and operational efficiency. However, we find evidence that the neighbor hospitals do so via a speed-up response (i.e., by reducing their service durations to accommodate the increased demand), instead of improving their bed utilization rate. We further show that such speed-up behavior negatively affects some aspects of the care quality, including an increase in the 30-day mortality rate.

Chapter 2 (with Soroush Saghafian) identifies an optimal financial allocation decision for U.S. hospitals under access-quality trade-off. Supporting rural (as opposed to urban) hospitals can improve access at the risk of reducing care quality. We study allocation decisions for the two hospital support programs currently under debate—the Disproportional Share Hospital (DSH) payment and the Critical Access Hospital (CAH) program---where one primarily targets urban hospitals and the other rural hospitals. Using a self-selection model that incorporates patients’ preferences for geography and quality, we analyze the characteristics of an optimal allocation that improves social aggregate utility under budget constraints. Based on the result, we suggest a threshold-based decision rule and evaluate the performance of our policy against the current payment policies through simulation analyses.

Chapter 3 (with Soroush Saghafian, Joseph Newhouse, Mary Beth Landrum, and John Hsu) studies how the financial integration between physicians and hospitals (“vertical integration”) affects the care quality of a medical procedure (colonoscopy for colorectal cancer screening and diagnosis) for which the patient outcome is highly sensitive to the quality of the procedure. By examining the recent acquisitions of specialist practices by hospitals and the FFS Medicare patients, we find that integration significantly reduces outcome-related quality, such as post-procedure complications. We further show that such changes are due to the resource constraints that integrated physicians face that drive them to reduce some recommended process-based care. We suggest that carefully designing the incentive structures of the administered payment system for integrated providers can mitigate adverse consequences.