Document Type



Doctor of Philosophy


Industrial Engineering

First Adviser

Thiele, Aurelie C.

Other advisers/committee members

Defourny, Boris; Duzgun, Ruken; Paxton, Stuart; Zuluaga, Luis


This dissertation provides robust, quantitative models in healthcare finance to aid decision-makers with rigorous, analytical tools that capture high complexity and high uncertainty of problem. The first chapter investigates the impact of parameter uncertainty on risk scoring, and presents an approach to obtain robust risk scores to address uncertainty in risk adjustment, which is used to quantify payment transfers across health plans under the Affordable Care Act. We provide a tractable methodology to incorporate uncertainty in the risk factor weights via linear programming to improve risk adjustment among payers and discuss the impact of uncertainty on the risk scores. In the second chapter, we provide an analytical methodology to help individuals narrow down plan choices in the Health Insurance Exchanges (HIX) by identifying plans that are dominated by competitors' offerings in terms of premium, metal level, maximum out of pocket payment and plan type. We further quantify the amount by which the premium of a dominated plan should be reduced to make it competitive in our framework. This part of our work provides important quantitative tools to guide the discussions between payers and policy-makers regarding HIX. Our approach also provides payers with a novel way to analyze their own plans in the HIX landscape. The third chapter identifies the key factors that drive enrollment rates of the two major types of Medicare Part D prescription drug plans: MAPD and PDP, to assist policy makers in better promoting their plans to Medicare beneficiaries. The fourth chapter investigates trends in physician services usage and Medicare reimbursement rate from CMS public files. We analyze the HCPCS (Healthcare Common Procedure Coding System) codes and investigate the validity of the concern that doctors tend to upcode on purpose for more reimbursement. We also utilize time series analysis to predict Medicare spending in ten years. In the last chapter, we survey and propose robust optimization models in healthcare systems engineering, particularly in the applications of healthcare costs prediction, disease management, IMRT fluence map optimization, and operating room planning, among others.