Major Hemorrhage on Warfarin in Elderly Patients

Abstract & Commentary

By John P. DiMarco, MD, PhD, Professor of Medicine, Division of Cardiology, University of Virginia, Charlottesville. Dr. DiMarco is a consultant for Novartis, and does research for Medtronic and Guidant.
This article originally appeared in the July 2007 issue of Clinical Cardiology Alert. It was edited by Michael Crawford, MD, and peer reviewed by Rakesh Mishra, MD.

Synopsis: Stroke prevention among elderly patients with atrial fibrillation remains a challenging and pressing health concern.

Source: Hylek EM, et al. Major hemorrhage and tolerability of warfarin in the first year of therapy among elderly patients with atrial fibrillation. Circulation. 2007;115:2689-2696.

Recently, the importance of anticoagulation with warfarin in patients with atrial fibrillation and risk factors for stroke has been established in clinical trials and emphasized in recent guidelines. The major complication of anticoagulant therapy is bleeding, and there is always a risk/benefit ratio which must be considered whenever warfarin is prescribed. In this paper, Hylek and colleagues report an inception cohort study designed to define the rate of major hemorrhage in atrial fibrillation patients starting warfarin therapy. Patients were eligible for inclusion if they were over 65 years of age, had atrial fibrillation verified by an ECG, and were starting warfarin therapy, which would be managed by a single, onsite anticoagulation clinic. Patients were enrolled on the first day of warfarin and followed throughout the first year of therapy. The end points analyzed were: major hemorrhage, time to termination of warfarin, and physician reasons for discontinuation. For this study, a hemorrhage was considered major if it was either fatal, required greater than or equal to 2 units of packed red blood cells, or involved a critical site. Each patient was classified according to the CHADS2 scheme for known risk factors of stroke. Concurrent medications were obtained from electronic medical records. Use of over-the-counter agents, such as aspirin or nonsteroidal anti-inflammatory drugs was also recorded.

The study cohort eventually included 472 patients. Of these, 47% were female and 54% were over 75 years of age, with 153 patients (32%) over 80 years of age. With the use of the Outpatient Bleeding Risk Index, 95.3% were classified as intermediate risk and 4.7% as high risk for major hemorrhage with warfarin therapy. During the study, the following proportion of patient times were spent within the following INR ranges: 2.0 to 3.0 (58%), below 2.0 (29%), between 3.1 and 4.0 (11%), and 2% greater than or equal to 4.0.

During the first year after starting warfarin, major hemorrhages were noted in 26 patients (5.5%). These included: 9 intracranial hemorrhages, 11 gastrointestinal bleeds, one retroperitoneal bleed, one hemothorax after a fall, one ocular bleed, one hemarthrosis, and 2 nose bleeds severe enough to require transfusions. The major hemorrhage rate was 7.2 per 100 person-years, with a rate of intracranial hemorrhage of 2.5%. Increased age was associated with increased risk of bleeding. Patients 80 years of age or older had a bleeding rate of 13.1 per 100 person-years vs 4.75 for those under age 80. Several risk factors for hemorrhage were identified: INR range greater than or equal to 4, older age, and first 90 days after initiation of therapy. During the first year of therapy, warfarin was discontinued in 134 patients. Although perceived maintenance of sinus rhythm was the most common cause (65 patients), safety concerns led to discontinuation in many others, particularly in those 80 years of age or older. There was a strong correlation between bleeding risk and an elevated CHADS2 score, demonstrating an important overlap between risks for bleeding and stroke.

Hylek et al conclude that in general clinical practice, the risk of bleeding may be higher than has been estimated from prior randomized trials. These observations indicate that stroke prevention among the highest risk patients remains a challenge.


Stroke is the most serious complication of atrial fibrillation. Randomized clinical trials have clearly shown that warfarin decreases this risk by about 67% in an intention-to-treat analysis, and has an even greater effect in an on-therapy analysis. However, many of the patients studied in these trials had already tolerated warfarin for some period of time before enrollment, and patients thought to be at particularly high risk of bleeding were excluded. For some studies, this was more than 35% of the patients screened for participation. As a result, the annual rates of major bleeding were relatively low (1.4% to 4.2%) in these trials and in some other large scale trials in atrial fibrillation. In this paper, Hylek et al show that "real world" anticoagulation of elderly patients with atrial fibrillation is still a major clinical problem. By studying an inception cohort, they mimicked the situation when a clinician makes a decision to start warfarin therapy. Even though all patients were followed systematically by an experienced anticoagulation clinic, the major hemorrhage rate in the first year of therapy was 7.2%, and much higher in those patients 80 years of age or older. An important observation was the risk factors for stroke and for bleeding overlap, making clinical decisions even more difficult.

How should clinicians react to these data? Clearly both stroke risk and bleeding risk must be considered before starting anticoagulation. The need for concomitant drugs that might increase bleeding risk (eg, aspirin, anti-inflammatory drugs, other platelet inhibitors) must be evaluated at each visit and, in many cases, these drugs should be discontinued. Very careful monitoring to prevent excessive anticoagulation is needed, especially in patients over 80.