Health Disparities
DARE: A Health Disparities Assessment Rubric for Epidemiologic Data Analysis Yulin Hswen* Maggie Hurley Hurley Hurley Hurley Hurley Hurley Hurley Hurley Hurley Hurley Boston University, School of Public Health
Addressing disparities is often essential to maximize the population health impact of epidemiologic research results. However, manuscript planning rarely systematically considers disparities. We developed the DARE (Health Disparities Assessment Rubric for Epidemiologic Data Analysis), a structured tool designed to integrate such considerations throughout paper development. The rubric covers 4 domains, scored from 1 (weakest) to 5 (strongest integration of items): Populations (e.g., does the sample include populations disproportionately affected by the outcome?), Analysis (e.g., are systemic drivers considered? Are covariates appropriately selected considering plausible mechanisms?), Subgroups & Intersectionality, and Triangulation (e.g., efforts to ensure robustness of results such as multiple data sources and/or methods and transparency in limitations). To pilot DARE in our research project, 2 reviewers independently applied the rubric to 6 paper proposals across the domains. Reviewer scores were collapsed into low (1 to 2), moderate (3), and high (4 to 5). Observed agreement between reviewers was highest for the Populations domain (100 %), followed by Analysis (83.3%), Subgroup & Intersectionality (66.7 %), and Triangulation (33.3 %) (Figure 1). Dark green cells indicate exact agreement on the original 5-point score, while light green cells reflect agreement within collapsed integration categories. Light red cells denote scores falling into adjacent collapsed categories +/- 1-point difference on the original scale, and dark red cells indicate disagreement across non-adjacent categories with a difference > 1 point. DARE may assist research teams that are committed to maximizing population health impact identify weak areas in their approach, integrate health disparities considerations effectively into their methodology, and strengthen the relevance and interpretability of epidemiologic findings for population health.
