Comparison of Frequency of Atrial Fibrillation in Blacks Versus Whites and the Utilization of Race in a Novel Risk Score

Abstract

Blacks have a lower prevalence of atrial fibrillation (AF) compared with Whites. We sought to confirm previously reported ethnic trends in AF in Blacks and Whites in a large database, and develop a prediction score for AF. Over 330 million hospital discharges between the years 2003 to 2013 from the National Inpatient Sample database were analyzed. All hospitalizations with a diagnosis of AF formed the study cohort. Traditional risk factors for the development of AF were compared between Blacks and Whites. Univariate and multiple logistic regression analyses were used to formulate a risk score to predict AF-CHADSAVES (Congestive heart failure, Hypertension, Age>65 years, Diabetes Mellitus, prior Stroke, Age>75 years, Vascular disease, White Ethnicity, and previous cardiothoracic Surgery). AF prevalence in Whites was 11.3% vs 4.6% in Blacks (p < 0.001). Blacks were younger (33.8% vs 14.4% patients <65 years, p < 0.01) and had less males (46.3% vs 49.4%, p < 0.01). Blacks had more hypertension (71.3% vs 64.1%, p < 0.01), congestive heart failure (24.8% vs 22.6%, p < 0.01), diabetes mellitus with (7.5% vs 4.7%, p < 0.01) or without complications (30.3% vs 23.1%, p < 0.01), renal failure (29.7% vs 17.1%, p < 0.01), and obesity (13.1% vs 8.7%, p < 0.01). CHADSAVES predicted AF in the study population (NIS 2003 to 2013) with an AUC of 0.82 and verified in a validation cohort (NIS 2014) with an AUC of 0.85. In conclusion, our data confirm a significant AF ethnicity paradox. Despite a higher prevalence of traditional risk factors for AF, Blacks had >2-fold lower prevalence of AF compared with Whites. CHADSAVES can be used effectively to predict AF in inpatients.