A Methodological Framework for Investigating Diversity in Clinical Research

Abstract

AbstractBackgroundMany reports and reviews have been published on best practices in research recruitment, engagement, and retention (RER), yet clinical research broadly lacks identifiable mechanisms to translate these qualitative reports into quantifiable, testable hypotheses on the science of RER. A science of diversity requires precise mechanisms for true efficacy, but before this is possible, we must develop key methodological and ontological frameworks. Thus, we are in dire need a basic sciences approach to identify mechanisms that underpin RER research, and disparities in Alzheimer’s disease and related dementias (ADRD) clinical research more broadly.MethodA science of diversity in clinical trials must ensure research conducted under the umbrella of this discipline is not methodologically confounded. Although researchers studying health inequities have begun to parse health disparitiesfrom health inequities, recruitment into research may still conflate foundational causal concepts, obscuring researchers’ ability to truly determine the nature of group differences. An ordered methodological framework was proposed, formalizing key biases that may confound efforts to diversity clinical research samples, and to determine the nature of health inequities and disparities among marginalized and minoritized populations.ResultIn addition to aspects of biological differences and structural inequities that may inform group differences in both RER and ADRD clinical research, three additional sources of bias were identified: biases of selection and inclusion into research, error and biases of measurement, and biases of implementation of research protocols. Notably, existing tools to measure and mitigate these sources of noise are available, ensuring that detected group differences are not confounded by shortcomings in scientific design.ConclusionI present this novel framework to support a rigorous methodological approach to the study of diversity in ADRD clinical trials, anchoring a formalized science of RER and ensuring that estimates of health disparities and inequities in ADRD research are free of biases introduced by well‐intentioned research teams. Implementing this methodological approach is expected to improve accrual of diverse and representative populations into ADRD research, and to improve the quality and generalizability of ADRD clinical research itself.