|| Checking for direct PDF access through Ovid
Background: Atrial fibrillation (AF) is the most common dysrhythmia in clinical practice, and is a significant contributor to morbidity and mortality. Prior reports have projected a large increase in AF burden over time. A contemporary assessment of epidemiology is needed to assess if an emphasis of prevention strategies over the last decade has been effective in alleviating this risk.Methods: We used a 5% national sample of all Medicare beneficiaries in the US from 2002 through 2013 to construct a longitudinal cohort of 2.3 million fee-for-service Medicare beneficiaries administratively followed for ≥2 years using claims data. Trends in incident and prevalent AF were assessed for 2004 through 2013. Using ICD-9 codes, encounters with AF were identified from inpatient, outpatient, and physician claims. AF during the first 2 years of entry into the cohort was defined as pre-existing AF. Incident AF was defined as having either 1 inpatient claim with a diagnosis of AF or 2 outpatient or physician claims with AF. Calendar-year prevalence comprised pre-existing and incident AF for the respective years as well as those with incident AF in preceding years. Age-adjusted time trends were assessed using Poisson regression.Results: Between 2002 and 2013, 219,570 patients had incident AF. At incidence, mean age was 79 years, 55% were women, and 92% and 5% were white and black, respectively. Age-adjusted AF incidence decreased by 0.4/1000 per year between 2004 (20/1000) and 2013 (17/1000). While incidence declined for white men and women (P<.05), it has remained unchanged for black men and women (Figure). Proportion of incident events in the outpatient setting increased from 26% to 40%. One-year mortality was 9%, and remained unchanged throughout the study period. Over this period, the overall prevalence of AF decreased by 0.9/1000 per year (p<.05), however, there was a relative increase in AF prevalence among black men.Conclusions: Between years 2004 and 2013, the overall incidence and prevalence of AF among a 5% sample of Medicare beneficiaries stabilized. There were, however, differences across racial groups, with a slight decline in incidence among white men and women, which was not observed in black men and women.