BackgroundIn the 1990s, epidemiological studies estimated the prevalence of stroke caused by atrial fibrillation (AF) at about 15 %. Given the aging population, there is a rise in the number of AF patients. AF prevention guidelines based on clinical practice and the literature have been published and updated since 2001. Implementation seems to have an impact on the prescription of vitamin K antagonist (VKA). During the last 20 years, few population-based studies have focused on the prevalence of atrial arrhythmia (AA) in patients with stroke. The objective of the present prospective study, using data from 2008, was to evaluate the prevalence of AA (atrial fibrillation/flutter) in patients with stroke and the impact of implementing AF guidelines.MethodsThe prevalence of AA was studied in patients diagnosed with stroke from January 1 to December 31, 2008 in the population-based Stroke Registry of Brest, France (total population, 363,760 according to the 2008 census, with 295,553 aged 15 years or older). Guidelines implementation was assessed in terms of antithrombotic therapy (VKA, antiplatelet agent, none), and the CHADS2 (Congestive heart failure, Hypertension, Age > 75 years, Diabetes mellitus, and prior Stroke or transient ischemic attack).Results851 cases of stroke were identified. The prevalence of AA was 31.7 % (n = 264), and increased with age from < 20 % in patients aged 45 to 54 years to nearly 50 % in patients ≥ 85 years. In patients with AA, 231 strokes were ischemic, 28 hemorrhagic and 5 undetermined. At time of stroke, AA was known in 207 patients (78.4 %). 54 of the 152 patients with CHADS2 score ≥ 2 (35.5 %) were treated with VKA; this proportion decreased with age: 50 % between 50 and 74 years, 43.8 % between 75 and 84 years, and 25 % at 85 years and older.ConclusionThe prevalence of AA in the population-based Brest Stroke Registry in 2008 was higher than that reported by studies conducted 20 years ago. Despite publication of AF prevention guidelines, VKA prescription and use in elderly patients were significantly low.