Abstract
RATIONALE: The differential attrition of racial/ethnic minority participants in clinical research is a major threat to advancing medical and behavioral science. OBJECTIVE: Our aim was to examine the influence of racial/ethnic concordance between participants and research staff on study attrition. METHOD: Data were pooled from participants and clinical research coordinators (CRCs) in six longitudinal studies of respiratory illness. Dyads were classified as concordant if the patient and CRC were of the same racial/ethnic group. Multilevel modeling examined the effect of racial/ethnic concordance on attrition at the first and one-year follow-ups. RESULTS: Spanish language, lower education, and greater depressive symptoms predicted greater attrition, but these effects disappeared in adjusted models. Race/ethnicity, age, gender and health literacy did not predict attrition. Contrary to hypotheses, attrition was greater among concordant than discordant dyads: Attrition was almost five times greater at first follow-up for Black and Hispanic participants in concordant dyads, and almost four times greater at one year. CONCLUSIONS: Racial/ethnic concordance between participant and CRCs was related to greater attrition in a highly diverse sample of adults with respiratory illness. Differential attrition of racial/ethnic minorities is a major threat to advancing public health. Interactions with research staff may be critical to bridging the disparities gap.