2D-09 - The Contribution of Racial Disparities in Medication Adherence to Disparities in Hypertension Control

Background: Racial disparities in blood pressure (BP) control are a major contributor to cardiovascular disease inequities in the United States. While medication adherence is a key determinant of BP control, quantifying its role in explaining these disparities remains understudied.

Objectives: To quantify how much racial disparities in medication adherence contribute to disparities in BP control among non-Hispanic Black (NHB) and non-Hispanic White (NHW) adults in a sample of primary care patients.

Methods: We analyzed data from the RICH LIFE Project, a two-arm, cluster-randomized trial in 30 Maryland and Pennsylvania primary care practices, that compared two interventions, Collaborative Care/Stepped Care and Standard of Care Plus (SCP), for reducing hypertension disparities. We included NHB and NHW adults in the SCP arm with uncontrolled hypertension,
≥ 1 cardiovascular risk factor, and on medication at baseline. Participants completed a baseline survey (August 2017-October 2019) on sociodemographic characteristics, patient-centered outcomes, and self-reported health behaviors. Medication adherence (4-item Morisky scale) was assessed at baseline and 12-months, dichotomized as adherent (total score = 0) or non-adherent (total score ≥1). The primary outcome, BP control at 12-months (SBP < 140 mm Hg, diastolic BP < 90 mm Hg), was measured using electronic medical records. A causal decomposition analysis estimated the observed disparity in BP control and the corresponding reduced and residual disparity in BP control upon removing disparity in adherence. A weighting procedure balanced allowable covariates (e.g. age and sex) across racial groups, while shifting the NHB medication adherence distribution to match that of NHW and accounting for confounding by nonallowable covariates (e.g. marital status, patient activation, stress).

Results: Among 757 participants (59% NHB, 41% NHW), NHB were more often younger (mean age: 58 years vs. 65 years), female (63.1%vs. 50.2%), with higher rates of consistent non-adherence at baseline and 12-months (35% vs. 18%) compared to NHW. The observed BP control disparity was -9.7 percentage points (95% CI: -18.0%, 0.6%). The reduced disparity, representing the portion that would be eliminated if NHB had the same adherence distribution as NHW, was -5.9 percentage points (95% CI: -22.7%, 4.5%). The residual disparity, representing the remaining difference unexplained by adherence, was -3.8 percentage points (95% CI: -15.8%, 16.4%).

Conclusions: Racial disparities in medication adherence explained a clinically significant portion of BP control disparities. Addressing adherence disparities could narrow BP control inequities and improve cardiovascular outcomes.