Current issues of ACP Journal Club are published in Annals of Internal Medicine


Long-term warfarin did not affect quality of life in patients with nonvalvular nonrheumatic atrial fibrillation unless they bled

ACP J Club. 1992 Jan-Feb;116:4. doi:10.7326/ACPJC-1992-116-1-004

Source Citation

Lancaster TR, Singer DE, Sheehan MA, et al. The impact of long-term warfarin therapy on quality of life. Evidence from a randomized trial. Arch Intern Med. 1991 Oct;151:1944-9.



To evaluate the effects on patients' perceived health status of long-term warfarin therapy when given to prevent stroke in patients with nonrheumatic atrial fibrillation.


Survey of participants in a randomized, unblinded, controlled trial (the Boston Area Anticoagulation Trial for Atrial Fibrillation—BAATAF).


Community, university, and Veterans Affairs hospitals and clinics in several U.S. cities.


416 patients randomized into BAATAF, excluding 51 patients who had died or had had a stroke and 8 patients who could not speak English or who had developed cognitive or sensory deficits, were asked to return a questionnaire. 93% of the remaining 357 patients responded. Warfarin patients had been in the trial an average of 875 days; comparison patients, an average of 856 days. The mean age for all patients was 68 years.


The intervention group (n = 177) received warfarin at a rate adjusted to produce a prothrombin time ratio of 1.2 to 1.5 and was not allowed aspirin. The control group (n = 156) received "normal medical care" and no warfarin and was permitted aspirin or other antiplatelet medications.

Main outcome measures

These included patients' responses to questions about their well-being, functional status, role functioning, and health perceptions (current health; health distress, outlook, and concern). Patients assigned to warfarin were asked specific questions about lifestyle restrictions and side effects associated with warfarin use.

Main results

Warfarin and control groups did not differ on any of the outcome measures nor did they differ when the analysis was adjusted for potential confounders. Only 7% of patients treated with warfarin felt warfarin restricted their lifestyle and 13% worried about side effects. However, the 23 patients who experienced a "mild" bleeding episode while taking warfarin (3 of these episodes required hospitalization and 1, a transfusion) had greater concern and distress over their health than the remainder of the warfarin-treated patients (P < 0.05).


Self-reported perceptions of health status were not damaged by long-term warfarin therapy in patients with nonrheumatic atrial fibrillation unless they bled.

Sources of funding: National Heart, Lung, and Blood Institute; Eliot B. Shoolman Fund; Louise U. Snell; Vera and J.W. Gilliland Fund.

Address for article reprint: Dr. D.E. Singer, General Internal Medicine Unit, Bulfinch 1, Massachusetts General Hospital, Boston, MA 02114, USA.


4 randomized trials have shown that warfarin decreases the rate of stroke and non-central nervous system emboli in patients with nonrheumatic atrial fibrillation. The annual absolute risk reduction for stroke varied from 1.5 to 4.7 events per 100 person-years of follow-up (1). This small but clinically important benefit must be weighed against the risk for major and minor bleeding caused by warfarin, as well as against the inconvenience and anxiety associated with its use.

The risk for major bleeding was low in these studies (an absolute increase of less than 1% per year, compared with control). It is reassuring that Lancaster and colleagues found no change in quality of life in warfarin-treated patients and that only a few patients worried about side effects or had to restrict their lifestyle. On the one hand, this study may overestimate patients' concern about warfarin use because at the time it was conducted there was no evidence that warfarin decreased the risk for stroke. Now that there is, patients may perceive more of an improvement in their health while on the drug. On the other hand, because patients volunteered for the drug trial, it is likely that their reactions to chronic medication use were more favorable than most patients with nonrheumatic atrial fibrillation. Overall, however, the results of this study should be reassuring to patients and clinicians when they embark on long-term anticoagulation for nonrheumatic atrial fibrillation.

This study also highlights the negative effect of medically "minor" bleeding on patients' concern about their health. Health care workers might attempt to decrease this concern by reassuring patients that minor bleeding, such as epistaxis, does not mean that they are at higher risk for a major bleeding event.

Andreas Laupacis, MD
Ottawa Civic HospitalOttawa, Ontario, Canada

Andreas Laupacis, MD
Ottawa Civic Hospital
Ottawa, Ontario, Canada


1. Albers GW, Atwood JE, Hirsh J, et al. Stroke prevention in nonvalvular atrial fibrillation. Ann Intern Med. 1991;115:727-36.