Abstract P007: Atrial Fibrillation, Black or White Race, and Research Study Participation in a National Cohort Study: The REasons for Geographic And Racial Differences in Stroke (REGARDS) Study

INTRODUCTION. Atrial fibrillation (AF) has been consistently observed in epidemiologic studies to be less prevalent in blacks than in whites, despite higher prevalence of AF risk factors in blacks. A research methods question related to this paradox is whether AF and race jointly affect participation in research studies that require in-person examinations, leading to biased estimates of the racial distribution of AF. The unique two-stage data collection design of REGARDS allowed us to study this issue. HYPOTHESES. We hypothesized that people who self-report AF by telephone are less likely to participate in in-person research studies than people who do not self-report AF, and that this difference is more pronounced in blacks than in whites. We also hypothesized that the inverse association of black race with AF is more pronounced in people who participate in in-person studies than in people who do not participate. METHODS. In REGARDS, 42,018 people completed a preliminary telephone interview in which they reported race (black or white) and answered the question “has a doctor or other health professional ever told you that you had atrial fibrillation?” Subsequently, 29,403 (70%) participated in an in-person examination, where AF was determined by electrocardiogram (ECG). We used multiple imputation to estimate prevalence of ECG-detected AF in people who did not participate in the in-person examination. RESULTS. Participation in the in-person examination was 77.0% in whites without self-reported AF, 76.5% in whites with self-reported AF, 62.0% in blacks without self-reported AF, and 61.6% in blacks with self-reported AF. Racial distributions of self-reported AF and ECG-detected AF were similar in people who did or did not participate in the in-person examination (Table). CONCLUSIONS. Self-reported AF status was not related to participation in an in-person examination following a telephone interview. We found no evidence that estimates of racial distributions of AF in REGARDS were biased by differential study participation.