Abstract
OBJECTIVE: To examine the association between socioeconomic status (SES) and racial and ethnic disparities in medication adherence for three widely prescribed therapeutic classes. METHODS: We linked longitudinal claims data from a large US-based insurance provider (2011-2013) to detailed SES information to identify patients treated with oral antidiabetic (N = 56,720), antihypertensive (N = 156,468) or antihyperlipidemic (N = 144,673) medications. We measured adherence and discontinuation by therapeutic class, and conducted regression analysis to quantify the contributions of different factors in the association between race/ethnicity and medication adherence. RESULTS: During an average follow-up period of 2.5 years, average adherence rates of Blacks and Hispanics were at least 7.5 percentage points lower than those of Whites. Controlling for demographics, health status, out-of-pocket costs, convenience of refilling prescriptions and SES attenuated the association by 30 to 50 percent, nonetheless substantial racial disparities persisted (4.1-5.8 percentage points), particularly for asymptomatic conditions. Separating adherence among existing users from those that discontinued therapies indicates that racial/ethnic disparities in adherence reflect inconsistent pill-taking rather than differential rates of discontinuation. CONCLUSIONS: Racial/ethnic disparities in adherence are mitigated, but persist after controlling for detailed socioeconomic measures. Interventions should focus more on improving medication adherence of existing users, particularly in treating asymptomatic conditions.