Dissertation Title:  "Risk-based Strategies for Population Screening and Disease Management"The objective of this dissertation is to explore the tradeoffs of risk-based strategies to screen populations for disease and manage patients with abnormal test outcomes. Many guidelines on screening or management of a patient population are based on a “one-size-fits-all” approach. However, tailoring guidelines to sub-groups based on risk has the potential to improve efficiency in health care spending and can result in adoption of strategies which decrease health care costs and improve patient outcomes

Chapter 1 introduces a risk-stratified approach to screening for cervical and colorectal cancer. Published literature indicates that early negative screens may predict decreased future risk of cancer incidence. This suggests that individuals with negative screens early on may not be harmed by extending their screening interval moving forward, potentially presenting a significant cost-savings to the health care system. In this chapter, we use simulation modeling to explore the cost-effectiveness of strategies to widen the screening interval for women with early single or serial negative screens for both cervical and colorectal cancer. For cervical cancer, we find that an adaptive strategy to extend the screening interval from five years up to 15 years after just 1 negative screen is cost-effective compared to current guideline screening. On the other hand, guideline screening using fecal immunochemical testing (FIT) for colorectal cancer is cost-effective compared to any adaptive strategy.

Chapter 2 evaluates the cost-effectiveness of updated guidelines by the American Society for Colposcopy and Cervical Pathology (ASCCP) on the management of abnormal cervical cancer screening results. While previous guidelines in 2012 were based on specific actions for specific test results, updated guidelines in 2019 have shifted to a risk-based approach with the goal of applying “equal management of equal risks.” We modified a microsimulation model of cervical cancer to assess the cost-effectiveness and resource utilization associated with newer guidelines compared to previous guidelines. We find that, under current screening practices, the 2019 guidelines are cost-effective and cost-saving relative to previous guidelines, indicating that a risk-based approach improves the efficiency of cervical cancer screening and management.

Chapter 3 explores diagnosis and screening for post-transplantation diabetes mellitus (PTDM), a complication of solid organ transplantation. There is sparse and inconsistent literature on PTDM, leading to inconsistent estimates of incidence and no differentiating guidelines on how to manage transplantation patients who may be at risk for PTDM. We analyzed data sets from kidney, liver, and heart transplantation patients at the Mayo Clinic, and used data imputation and simulation modeling to evaluate the potential impact of a screening program to collect hemoglobin A1c (HbA1c) and fasting blood glucose (FBG) from all patients in the immediate post-transplantation setting. We find that poor and inconsistent collection of HbA1c and FBG results in underestimating of PTDM incidence and that better screening of these measures may be a cost-effective intervention to improve long-term patient outcomes.