Thesis Title: Cost-effective Management of Chronic Diseases: Surveillance, Treatment, and Elimination

Advisor: Dr. Turgay Ayer

Committee members: Dr. Jagpreet Chhatwal (Harvard Medical School), Dr. Pinar Keskinocak, Dr. Ton Dieker, and Dr. Paul Griffin

Date and Time: Monday, October 31, 2016, 10:00 AM.

Location: ISyE Groseclose 226A

Summary:

Chronic diseases have become the most common health problems worldwide. Chronic disease management leads to a substantial economic burden to both patients and society, and the cost continues to rise. Mathematical models and operations research methods can serve as useful tools to better inform policy-level decisions on cost-effective management strategies for chronic diseases. This dissertation focuses on three important problems related to the care for chronic diseases---treatment, surveillance, and potential elimination.

In the first part of this dissertation, we study the economic and disease burden of chronic lymphocytic leukemia with the emerging therapeutic options. Oral targeted therapies represent a significant advance for the treatment of CLL patients, but their high cost has raised concerns about affordability and economic impact to the society. We develop a comprehensive simulation model to evaluate the impact of the oral targeted therapies on the economic and disease burden of chronic lymphocytic leukemia (CLL). Our results show that the oral targeted therapies will substantially increase the annual cost of CLL management and per-patient lifetime out-of-pocket cost by more than 5-fold from 2011 to 2025, which far outpaces the rising cost of other cancers. At the current price, the new oral targeted treatment strategy is deemed not cost-effective compared with the old standard-of-care. Our results highlight that such an economic impact could result in financial toxicity, limited access, and lower adherence to the oral therapies, which may undermine their clinical effectiveness. Hence, we conclude that more sustainable pricing strategy for targeted therapies is imperatively needed.

In the second part, our objective is to identify cost-effective strategies for liver cancer surveillance in hepatitis C-infected population. Liver cancer is one of the deadliest cancers and is the fastest growing cause of cancer-related deaths in the United States. Regular surveillance has shown to improve early detection of cancer, but its optimal use remains unclear. We develop a mixed-integer programming (MIP)-based framework to systematically examine the most cost-effective surveillance policies. Our MIP-based framework captures two problem features that make dynamic programming-based formulations computationally intractable. In particular, our proposed framework allows to (1) explicitly formulate M-switch policies that are practical for implementation, and (2) tailor surveillance policies for each subpopulation by stratifying surveillance intervals based on the observable disease states. Our results lead to important policy implications. First, unlike the current one-size-fits-all type policies, the optimal surveillance interval should be stratified based on the stage of hepatitis C infection and age; second, expanding surveillance to patients in earlier stage of hepatitis C infection improves the cost-effectiveness of liver cancer surveillance.

In the last part, we study an optimal resource allocation problem motivated by the ongoing efforts by the World Health Organization to eliminate hepatitis C virus (HCV) by 2030. More than 170 million people are chronically infected with HCV globally. Although new antiviral treatments for HCV offer a hope to eliminate HCV, most countries do not have national programs to screen and treat HCV. We develop two optimal control model formulations considering HCV screening and treatment interventions; analyze the optimal policy structure for each formulation using optimal control theory, and show numerical solutions and policy implications for the case of HCV elimination in India. In the first formulation, we study a cost minimization problem subject to a final target prevalence constraint. We analytically show that the optimal treatment is a pure bang-bang policy for any given screening rate. A simple policy structure with constant screening and treatment rates does not result in a significant loss compared with policies with dynamic rates, which represents a promising practical policy structure for implementation. In the second formulation, we study a budget allocation problem that aims to minimize HCV disease burden subject to a fixed budget. We show that optimal allocation policy follows a simple treatment-first rule via both analytical and numerical results.