Integrating Clinical Research with Electronic Health Records to Improve Diversity in Research: Findings from an Urban Hospital System

BackgroundThe under-representation of racial, ethnic, and other minority populations in clinical research has threatened the validity of novel therapeutic interventions and exacerbated the longstanding inequities in health outcomes. Despite attention and mandates across institutions and federal agencies to address these disparities, effective and actionable strategies have remained a subject of debate in the existing literature. Thus, the need for comprehensive and rigorous methods to assess diversity in study population samples as well as strategies for improved recruitment and retention has been made clear. ObjectiveExamine the distribution of research participant demographics at Norton Healthcare (NHC) and compare to applicable benchmarks from the overall NHC patient population and local census. DesignSuccessive-independent samples comparison of EPIC electronic health records (EHR) and census data from Jefferson County. ParticipantsA total of 2,053 adult research participants enrolled at NHC from 2020 - 2024. Main MeasuresDemographic data were reported as frequency and percentage across the three benchmarks. The z-test for independent proportions was used to compare the research participant demographics to the NHC patients and Jefferson County during a consistent time period. Temporal trends in research participant demographics were also examined. Key ResultsApart from being relatively older, the NHC research participant population closely resembled the NHC patient population. A similar age-related pattern emerged as well as an over-representation of White individuals in the research participant group when compared to the census data. When looking at research participant demographic trends overtime, increases in White, Black and older cohorts were noted while decreases were observed in Hispanic/Latinx and younger cohorts. Trends related to participant sex remained stable. ConclusionsThe findings from this study will inform future strategies for setting enrollment goals and facilitate the creation of tools and metrics to evaluate appropriate standards for diversity in clinical research study population samples.