Abstract 136: Enrollment of Black Individuals in Cardiovascular Clinical Trials: A Meta-Analysis

Abstract

Objectives: Lack of racial and ethnic diversity in clinical trials (CT) limits the generalizability of results to individuals who have unique cultural, environmental and physiological factors that affect outcomes. Consequently, the NIH and FDA issued directives to include historically underrepresented racial and ethnic groups in research studies. We sought to determine the enrollment of black relative to non-black populations in cardiovascular clinical trials and whether racial enrollment varied across trial characteristics. Methods: Clinical trials were identified using PubMed, Embase, and Cochrane. Between 2000-2018, studies were included if they were English language, CT of cardiovascular disease or risk factors and race/ethnicity were reported. Trial characteristics were extracted from the primary manuscript or clinical trials.gov and included: study area, funding source, intervention type, location, sample size, and distribution of female participants. Summary statistics for each study included percentage of blacks enrolled with 95% CI. Meta-analysis of the proportion of black enrollment across relevant CT was conducted to analyze overall estimates of the effect compared to non-black enrollment using an inverse variance weighted random effects model. Additional meta-analyses compared candidate variables that could influence the likelihood of black enrollment. All analyses were performed with SAS version 9.4 (SAS Institute Inc., Cary NC). Results: We identified 2501 trials, 85 met inclusion criteria, comprising a total of 159,791 patients: 25,720 black and 134,071 non-black. The percentage of blacks 38.2% (95%CI: 33.0%-43.6%) enrolled was significantly less than non-blacks 61.8% (95%CI: 56.4%-67%). Enrollment differed significantly across study type with peripheral vascular disease enrolling the largest 59.9% (95%CI: 47.3%-71.4%) and cerebrovascular disease the least 25.1% (16.1%-37%). Industry only studies accounted for 41% (35 of 85) of total trials and were less likely 30.3% (95%CI: 22.7%-39.1%), than NIH 40.9% (95%CI: 34%-48.1%), joint NIH/Industry 46.8% (95%CI: 33.3%-60.7%), and non-NIH/Industry 76.6% (95%CI: 38.7%-94.4%) funded studies to enroll black patients. Pharmacotherapy studies 29.3% (23.3%-36.1%) enrolled fewer black patients than behavioral 43.8% (36.9%-51%) and device/other 48.2% (37.5%-59.1%) studies. Black enrollment decreased for non-US based studies, and as a function of decreasing female sex enrollment and increasing sample sizes of studies. Conclusion: CTs of cardiovascular disease and risk factors were significantly less likely to enroll blacks when compared to non-black patients. Trial characteristics such as study area, funding source, intervention type, study site location, sample size, female percentage, and mean age had statistically significant impacts on black patient accrual in cardiovascular CTs.