Background: Large scale data on atrial fibrillation (AF) prevalence in acute ischemic stroke (AIS) is sparse since approval of dabigatran for non-valvular AF in 2010. We studied recent trends in prevalence of AF in AIS and transient ischemic attack (TIA) in the United States (US) and association of AF with in-hospital mortality, cost and length of stay (LOS) in AIS.
Methods: Adults admitted to US hospitals from 2007-2012 with diagnosis of AIS (n=3,427,806) and TIA (n=502,820) were identified from the Nationwide Inpatient Sample. Weighted prevalence of AF in AIS and TIA by demographics and region was computed. Multivariate logistic regression was used to evaluate association of AF with other clinical factors and mortality in AIS. Association of AF with LOS and cost was assessed using generalized linear models.
Results: AF prevalence increased by 11.5% in AIS (22%-24.5%, p<0.001) and by 29% in TIA (13.5%-17.4%, p<0.001) from 2007-2012. AF prevalence varied by age (AIS: 7% in 50-59yo vs 38% in >80yo; TIA: 5% in 50-59yo vs 27% in >80yo), sex (AIS: 20% in M vs 25% in F); TIA: 16% in M vs 15% in F), race (AIS: whites 26% vs blacks 12%) and region (AIS: Northeast 25% vs South 21%). AF prevalence increased in all subgroups over time (p<0.001) except AIS <40yo and TIA<50yo (Figure 1). Advancing age, female sex, white race, high income, Medicare insurance, CHA2DS2-VASc score and large hospital size were associated with increased odds of AF in AIS. AF was positively associated with death (OR=1.60, 95%CI 1.56-1.64) but mortality in AIS with AF decreased from 13.2% in 2007 to 10.7% in 2012 (p<0.001). AF was associated with increased cost of $2,631 and LOS 1.1 days in AIS.
Conclusion: Prevalence of AF in AIS and TIA has continued to increase. Disparity in AF prevalence in AIS and TIA exists by patient and hospital factors. AF is associated with increased mortality, LOS and hospital cost in AIS but mortality in AIS with AF is decreasing. More AIS preventive efforts are needed in AF patients especially in the elderly.