Background: The benefits of antithrombotic therapy in chronic atrial fibrillation (AF) have been established in clinical trials, but there is a paucity of data on outcomes in practice.
Aims: The objective was to establish a large ongoing database of patients with non-valvular AF, to enable the accurate determination of clinical outcomes.
Methods: A retrospective review of the medical records for consecutive patients who had AF documented on electrocardiogram at the major teaching hospital in Tasmania between 1 January 1997 and 30 June 1999 was performed. An extensive range of demographic and clinical variables was recorded for all patients with chronic or paroxysmal non-valvular AF.
Results: The 505 patients (60% males) included in the database had a median age of 76 years. According to risk stratification criteria, 79% of the patients with previously diagnosed chronic or paroxysmal AF had a high risk of developing stroke at the time of admission to hospital care. However, only one-third (34%) of these patients were receiving warfarin (or warfarin plus aspirin), with almost one-quarter (24%) receiving no antithrombotic agent. The annual incidence of ischaemic strokes was 3.4% (1.5-6.4%; 95% CI) when taking warfarin, compared to 7.0% (5.2-9.4%) for patients not taking warfarin and 7.8% (5.4-11.1%) for patients taking aspirin. The annual incidence of bleeding complications in patients taking warfarin was 14.2% (10.0-19.5%) overall and 3.4% (1.5-6.4%) for major bleeds. In patients not taking warfarin, the overall annual incidence of bleeds was 8.4% (6.3-10.9%) and 3.9% (2.5-5.7%) for major bleeds.
Conclusions: Warfarin is underused in patients with AF. In clinical practice, warfarin confers a similar stroke risk reduction to that observed in trials, with an increase in incidence of only minor bleeding complications. Aspirin did not appear to reduce the risk of stroke.