Impact of Cost-sharing on Utilization of Medications for Secondary Prevention of Cardiovascular Morbidity and Mortality in Medicare Beneficiaries

Abstract

Purpose: The purpose of this study was to determine the influence of out-of-pocket prescription and healthcare costs on adherence to guideline recommended statins, angiotensin converting enzyme (ACE) inhibitors/angiotensin receptor blockers (ARB), and beta-blockers (BB) used for secondary prevention of coronary heart disease and the associations of adherence with cardiovascular mortality in community-dwelling Medicare beneficiaries ≥ 65 years. Methods: Data from the 2004, 2005, and 2006 Medicare Current Beneficiary Survey (MCBS) was utilized to conduct a retrospective, cross-sectional (i.e., multiple cohort) study. Dependent variables of interest included adherence to statins, ACE/ARBs or BBs, and all-cause mortality, with out-of-pocket (OOP) costs, and adherence to these medications the primary independent variables of interest in these models. Adherence was analyzed as a binary variable with ≥ 80 percent annual adherence the threshold utilized in primary analyses. Total OOP prescription costs for all medications and total OOP healthcare costs borne by the beneficiary were reported. Complex survey design-specified logistic regression with sampling weights was the main statistical analysis used. Sensitivity analyses on adherence thresholds and subgroups were additionally conducted. Results: A significant positive relationship between total OOP prescription costs and statin adherence was identified across observation years in the primary models. Similar relationships were noted for ACE/ARBs and BB in 2004, and ACE/ARBs in 2005. No significant association between adherence and total OOP healthcare costs was indicated in the primary models. Mortality could not be used as a clinical outcome of interest due to limitations with the data. Thus, acute coronary syndrome (ACS) events were used as the clinical outcome. At the ≥ 80 percent threshold, no significant reductions in ACS events were reported. However, various sensitivity analyses did suggest significant reductions in ACS events with ACE/ARBs. Additionally, significantly higher risk of ACS was noted when BB adherence thresholds were reduced to ≥ 60 percent. Conclusions: OOP prescription costs are a significant factor influencing adherence to these medications used for secondary prevention of CAD/MI in Medicare beneficiaries. Continuing to monitor how these costs impact adherence and ultimately outcomes will be critical, particularly given policy changes such as Medicare Part-D.