Qin, Xuanzi2024-01-192024-01-192020-11https://hdl.handle.net/11299/260140University of Minnesota Ph.D. dissertation.November 2020. Major: Health Services Research, Policy and Administration. Advisors: Beth Virnig, Peter Huckfeldt. 1 computer file (PDF); xi, 127 pages.Out-of-pocket (OOP) costs can affect patients’ access to care and clinical outcomes. This dissertation takes advantage of the natural experiment provided by the combination of the introduction of generic aromatase inhibitors (AIs) and Medicare Part D low-income subsidy (LIS) policy to understand the role of OOP costs and subsidies in the delivery of breast cancer care. Guidelines suggest postmenopausal women diagnosed with hormone receptor-positive (HR+) breast cancer initiate adjuvant hormonal therapy with either AIs or tamoxifen. Switching to another therapy drug is an important strategy to manage treatment related side effects. Those diagnosed at early stages should also receive surgery and/or radiation before or after the initiation of hormonal therapy. The three AIs, anastrozole, exemestane and letrozole, went off patent sequentially in 2010 and 2011. Generic entry lowered OOP costs for AIs. Medicare Part D beneficiaries receiving LIS (subsidized) have substantial lower OOP costs for prescription drugs than those without LIS (unsubsidized), and thus are unlikely to be affected by changes in OOP costs due to generic entry. Medicare and Medicaid dually eligible beneficiaries (duals) receive LIS and Medicaid support for other medical services. This dissertation uses the Surveillance, Epidemiology and End Results (SEER)-Medicare linked database to identify the main study cohort that consists of women first diagnosed with HR+ breast cancer at age 65 years and older between 2007 and 2013 (N=93,650). I find the introduction of generic AIs was associated with improved pharmaceutical access indicated by increased probability of initiating any hormonal therapy drugs, increased timeliness of initiation and increased probability of receiving AIs over tamoxifen. Despite the minimal reduction in OOP costs for AIs after generic entry among the subsidized group, the subsidized experienced similar changes in outcomes like the unsubsidized group who had large reductions in OOP costs after generic entry. Thus, reduced OOP costs due to generic entry can only partially explain the improved access after generic entry. I also find that generic entry increased therapy adherence and reduced early discontinuation both directly through reduced OOP costs and through intermediary changes in switching behaviors resulted from reduced OOP costs. Generic entry affected therapy adherence and continuation without drug switches directly through reduced OOP costs, and thus increased adherence and continuation without drug switches after generic entry were only observed in the unsubsidized group. Generic entry might affect adherence and continuation with drug switches through both reduced OOP costs and improved management of side effects, and thus increased adherence and continuation with drug switches were observed in both the subsidized and unsubsidized groups. Finally, I find that although duals have their medical service use and prescription drug use subsidized, duals were less likely to be diagnosed at early stages, receive guideline-compliant treatment and initiate hormonal therapy than non-duals. Among duals, those with Medicaid coverage gaps or coverage changes were less likely to be diagnosed at early stages and more likely to discontinue their hormonal therapy than duals without any coverage changes. These results demonstrate that interventions simply reducing OOP costs may be not enough for improving access to care. The inconsistent relationships between costs and health care use and outcomes among the subsidized group highlight the importance to understand non-financial barriers to access, such as health insurance literacy, physician and pharmacist behaviors.enBreast cancerMedicareOut-of-pocket costsPharmaceutical outcomesPrescription drugsSubsidyThesis or Dissertation