TORONTO, CANADA. Although the bleeding risk involved in warfarin therapy has been evaluated in numerous clinical trials, there is very little data on the actual incidence of hemorrhage amongst patients not participating in clinical trials. It is conceivable that there could be a significant difference between clinical trial incidence and the incidence found in the real world, since time spent in the usually recommended INR range of 2.0 to 3.0 is often as low as 50% in community surveys, but much higher in closely controlled clinical trials.

A group of Canadian researchers sponsored by the Ontario Drug Policy Research Network now reports the results of a study aimed at determining the actual incidence of hemorrhage in a population of 125,195 atrial fibrillation (AF) patients who were prescribed warfarin during the period April 1, 1997 to March 31, 2008. The average age of the study participants (evenly split between men and women) was 77 years with 57.5% being 76 years or older. The majority had one or more comorbid conditions such as hypertension (75%), congestive heart failure (35%), and diabetes (24%), which would increase their risk of ischemic stroke. Only 7% had a CHADS2 score of 0 and 69% had a score of 2 or higher. Patients were followed until one of the following events occurred:

• Visit to hospital for hemorrhage
• End of warfarin therapy
• Death
• Five years of follow-up
• End of study period (March 31, 2010)

If a patient had multiple hospital admissions during warfarin therapy, only the first admission was counted. The overall incidence of hemorrhage requiring a visit to an emergency department, or an admission to hospital was 3.8% over a 5-year follow-up period. The rate was significantly higher amongst patients aged 76 years or older (4.6%) than amongst those younger (2.9%). The rate of hemorrhage was particularly high (11.8%) during the first 30 days of warfarin administration.

The most common specified bleeding types were lower gastrointestinal at 36.5%, upper gastrointestinal at 26.1%, and intracranial (hemorrhagic stroke) at 5.1%. Eighteen percent of patients admitted to hospital with warfarin-related bleeding died in hospital or within 7 days of discharge. Mortality was particularly high (42%) amongst patients admitted for intracranial bleeding.

The researchers conclude that warfarin-related hospital admissions of AF patients is substantially higher than that reported in clinical trials designed to evaluate the safety and efficacy of warfarin therapy.
Gomes, T, et al. Rates of hemorrhage during warfarin therapy for atrial fibrillation. Canadian Medical Association Journal, November 26, 2012 [Epub ahead of print]

Editor's comment: It seems to me that this study is somewhat misleading in that patients were no longer followed-up once they had experienced a first bleeding event after having started on warfarin therapy. If all warfarin-related hemorrhages had been recorded, I would be very surprised if the annual incidence of bleeding would not have overshadowed any protective effect against ischemic stroke, except in cases where the patient had already suffered a previous stroke.