Incidence and Cost of Stroke and Hemorrhage in Patients Diagnosed with Atrial Fibrillation in France

Background: Stroke represents a major complication of atrial fibrillation (AF). The current anticoagulation options for stroke prevention increase hemorrhage risk. The objective of the study was to estimate the incidence and costs of hospitalization for stroke and hemorrhage in patients with AF who are eligible for stroke prevention. Methods: Patients hospitalized for AF were identified from the French National hospital database (Programme Medicalise des Systemes d'Information) and a calculated stroke risk score (congestive heart failure, hypertension [blood pressure consistently >140/90 mm Hg], age >= 75 years, diabetes mellitus, and previous stroke, transient ischemic attack [CHADS2]). Adult patients eligible for stroke prevention (CHADS>2.0) were enrolled. The incidence of hospitalization for stroke and hemorrhage was calculated over a 2-year period. Costs of acute care were determined using diagnosis related groups (DRGs) and corresponding National Hospital Tariffs. Rehabilitation costs were analyzed for patients with strokes and classified by stroke severity. Results: Sixty-one thousand five hundred eighty-two patients were identified with a mean age of 75.0 +/- 11.0 years and a mean CHADS2 score of 1.90 +/- 0.99. The 24-month cumulative incidence of any stroke was 32.1 cases/1,000 patients with AF (ischemic, 60%; hemorrhagic, 24%; unspecified, 16%), and that of hemorrhage was 53.1 cases/1,000 patients with AF (gastrointestinal, 26%; intracranial, 5%; other, 69%). The mean costs of ischemic and hemorrhagic strokes were (sic)4,848 and (sic)7,183 (mild), (sic)10,909 and (sic)14,298 (moderate), (sic)29,065 and (sic)29,701 (severe), and (sic)6,035 and (sic)4,590 (fatal), respectively. The mean costs of hemorrhage by type were (sic)3,601 (gastrointestinal), (sic)7,331 (intracranial), (sic)3,941 (other major), and (sic)2,552 (nonmajor). Conclusions: The incidence and cost of hospitalization for hemorrhage should be considered in the global burden of AF. These data should be useful for pharmacoeconomic evaluation of new oral anticoagulant medications. Such real-world studies may be relevant for monitoring mid-to long-term morbidity and mortality in the AF population.