Abstract TP159: Risk of Ischemic Stroke in Patients with Atrial Flutter

Abstract

Introduction: Guidelines advise anticoagulation for patients with both atrial fibrillation (AF) and flutter (AFL). However, the risk of stroke associated with these two conditions has not been robustly compared. Methods: Using inpatient and outpatient Medicare administrative claims data from 2008-2014 for a 5% sample of all beneficiaries ≥66 years of age, we identified all patients diagnosed with AFL or AF. Patients with prior stroke were excluded. The primary outcome was ischemic stroke. Predictors and the outcome were ascertained by validated ICD-9-CM diagnosis codes. Patients were censored at the time of ischemic stroke, death, or last available follow-up. In the primary analysis, patients with AFL were censored upon diagnosis of AF. Survival statistics were used to compare stroke incidence in patients with AF versus AFL. We used Cox proportional hazards analysis to compare the associations of AFL and AF with ischemic stroke after adjustment for demographics and risk factors. In a secondary analysis, patients with AFL were not censored upon diagnosis of AF because the natural history of AFL frequently features the development of AF. Results: We identified 16,441 patients with AFL and 338,726 with AF. Patients with AFL were less often female (42.4% versus 53.0%), but other baseline characteristics, including mean CHA 2 DS 2 -VASc scores (3.9), were similar between groups. Over 2.4 (±2.0) years, 16,451 ischemic strokes were identified. The annual incidence of ischemic stroke in patients with AFL was 0.60% (95% confidence interval [CI], 0.48-0.75%), whereas it was 2.00% (95% CI, 1.96-2.02%) in patients with AF. After adjustment for demographics and vascular risk factors, AFL was associated with a lower risk of ischemic stroke than AF (hazard ratio [HR], 0.30; 95% CI, 0.24-0.37). Within 1 year, 64.9% (95% CI, 64.1-65.6%) of patients with AFL received a diagnosis of AF. The difference between stroke risk in AFL compared to AF was attenuated (HR, 0.85; 95% CI, 0.79-0.92) when patients with AFL were not censored upon diagnosis of AF. Conclusions: Patients with AFL may face a lower risk of ischemic stroke than patients with AF. However, AF often develops in those with AFL, and taking this into account, the difference in stroke risk between the two conditions grows smaller.