Objective: To determine whether trial efficacy of prophylaxis with warfarin for patients with atrial fibrillation at high risk of stroke translates into effectiveness in clinical practice.
Design: Two year prospective cohort study.
Setting: District general hospital.
Participants: 167 patients with atrial fibrillation and at high stroke risk who were eligible for anticoagulation.
Interventions: Long term anticoagulation with warfarin at adjusted doses to maintain an international normalised ratio of 2.0-3.0.
Main outcome measures: Comparison of patient characteristics, comorbidity, anticoagulation control, stroke rate, and haemorrhagic complications with pooled data from five randomised controlled trials.
Results: Patients in the study group were seven years older (95% confidence interval 4 to 10) and comprised 33% more women than patients in the pooled trials. The international normalised ratio was in the target range for 61% of the time (range 37%-85%), below for 26% of the time (range 8%-32%), and above for 13% of the time (range 6%-26%). The time that patients in the study group spent in the target range was significantly less than in the pooled analysis. The incidence of stroke in the study group (2.0% per year, 0.7% to 4. 4%) was comparable to that of patients receiving warfarin in pooled studies (1.4%, 0.8% to 2.3%). Per year the incidence of major (1.7% v 1.6%) and minor (5.4% v 9.2%) bleeding complications was also similar.
Conclusion: Rates of stroke and major haemorrhage after anticoagulation in clinical practice were comparable to those obtained from pooled data from randomised controlled studies for patients with atrial fibrillation at high risk of stroke.