Racial representation in heart failure clinical trials and registries: A systematic review and meta-analysis.

BACKGROUND: Underrepresented racial and ethnic group (UREG) participation in heart failure (HF) clinical trials has not historically represented the disproportionate high burden in those communities. HF registries may include more UREGs based on broader inclusion criteria and observational nature. OBJECTIVES: Study objective was to examine the proportion of UREGs participating in HF trials compared to HF registries using a systematic review and meta-analysis. METHODS: We searched both PubMed and Embase for randomized controlled HF trials and registries completed between 2010-2019 based in the United States. A total of 2124 articles were identified and screened for inclusion. Secondary information was supplemented by Clinicaltrials.gov and/or representative publications. The percentages of different races within included HF trials and registries were calculated and quantitative meta-analyses were performed to compare enrollment percentages between trials and registries. RESULTS: This review included 62 trials and 15 registries of which 45 (72.6%) trials and 10 (66.7%) registries reported granular racial groups rather than aggregated racial categories. Black participants were significantly less represented in HF registries (21.8%, 95% CI: 17.2-27.3%) compared to clinical trials (30.4%, 95% CI: 26.1-35.1%; P=0.016). Registries also demonstrated a lower percentage of non-White participants (18.0%, 95% CI: 11.1-27.7%) compared with trials (24.3%, 95% CI: 18.7-31.0%; P=0.25); however, the difference was not statistically significant. CONCLUSIONS: Black participants were significantly less represented in HF registries than trials despite the observational nature of registries. Contrary to our hypothesis, HF registries included in this systematic review did not yield higher inclusion rates of Black and non-White HF patients compared to clinical trials.

Duke Scholars

Author Kevin Lindsey Thomas Medicine, Cardiology

Author Cynthia Lea Green Biostatistics & Bioinformatics, Division of Biostatistics