To determine whether the success and complication rates of oral anticoagulation obtained in large, well controlled trials, upon which recommendations are based, are comparable with routine cardiology practice.
An observational, prospective cohort study collected data on all patients followed at an anticoagulant clinic over one calendar year.
One thousand and seventy-eight patients anticoagulated for cardiovascular indications, mainly atrial fibrillation, prosthetic valves and ventricular dysfunction, were followed for 804 patient-years of treatment. No patient was lost to follow-up.
Telephone conversations and regular verification of medical files were used to record and classify all bleeding and thromboembolic events according to severity. International normalized ratios (INR) were compared with target ranges.
One hundred and twelve bleeding events, ie, 13.9/100 patient-years (% p-y), were recorded, of which 61 required medical attention. Major hemorrhages, defined as those requiring treatment or hospital observation for more than 24 h, occurred in 15 instances (1.9% p-y). Among these, 9 (1.1% p-y) were classified as life threatening, with four being fatal (0,5% p-y). Twenty-two thromboembolic events (2.7% p-y) occurred, of which 10 were major (1.2% p-y), leaving three patients (0.4% p-y) with long term sequelae and causing two deaths (0.25% p-y). INRs were within target range 62.3% of the time, with 2.2% of values recorded above 5 and 0.3% above 10.
The low failure and complication rates obtained in large, controlled trials are similar to those observed in actual cardiology practice.