2087 Evaluating Ethnicity, Gender, And Race Representation in Landmark Stroke Clinical Trials Relative to Real-World Incidence and Prevalence Patterns

INTRODUCTION: The Binomial Distribution Diversity Disparity Index (BD3I) is a new conceptual framework, not based on a specific pre-existing index. The specific application, combination of methodologies, and focus on binomial distribution parameters are tailored to the context of comparing disease prevalence/incidence in patient demographics with those in clinical trial demographics. This index can be implemented using statistical software by inputting the demographic data for both the disease-affected population and the clinical trial participants. METHODS: The demographic profile of the Disease Population (DP) and Clinical Trial Population (CTP) are independently categorized by ethnicity/race. For example, categories could include White, Black, Hispanic, Asian, and Other, depending on the region and the disease's prevalence. For each ethnic/racial category, the proportion of individuals are calculated within the respective total DP (P[DP i]) and CTP (P[CTP i]). The proportions define the parameters of a binomial distribution for each ethnic/racial category. The Disparity Index (D i), calculated by the difference in binomial distributions |P(DP_i) - P(CTP_i)| for each ethnic/racial group, measures the absolute disparity between the disease population and the clinical trial population for each category. The weighted aggregate of (D i), based on the number within the respective population, constitutes the BD3I. Transforming the measure (100*BD3I/2) generates a number ranging from 0-100. Closer to 100, the clinical trial participant demographic profile aligns with the CDC “typical” demographic patient profile. RESULTS: Higher BD3I indicates a greater disparity, suggesting that the clinical trial does not adequately represent the demographic profile of the disease-affected population. CONCLUSIONS: Validating the index by applying it to different diseases and clinical trials ensures meaningful and consistent results. Adjusting the categorization, weights, and calculation method as necessary is based on feedback and observed disparities.