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


Warfarin was cost-effective for stroke prophylaxis in nonvalvular atrial fibrillation with additional risk factors

ACP J Club. 1996 May-June;124:81. doi:10.7326/ACPJC-1996-124-3-081

Source Citation

Gage BF, Cardinalli AB, Albers GW, Owens DK. Cost-effectiveness of warfarin and aspirin for prophylaxis of stroke in patients with nonvalvular atrial fibrillation. JAMA. 1995 Dec 20;274:1839-45. [PubMed ID: 7500532]



To examine the cost-effectiveness of warfarin sodium and aspirin for prophylaxis of stroke in patients with nonvalvular atrial fibrillation (NVAF), with or without additional risk factors for stroke (previous stroke or transient ischemic attack, diabetes, hypertension, or heart disease).


Cost-effectiveness analysis using a decision-analytic model. The probabilities of stroke, hemorrhage, and death were obtained from randomized controlled trials. Estimates of quality of life were derived from interviews with 74 patients who had atrial fibrillation. Costs were estimated from a literature review, a telephone survey, and Medicare reimbursement.


United States.


In the base case, patients who were 65 years old with chronic NVAF and who were good candidates for warfarin and aspirin therapy.


Treatment with warfarin, aspirin, or no therapy.

Main Cost and Outcome Measures

Quality-adjusted life expectancy and marginal cost-effectiveness of warfarin compared with aspirin or no therapy. Costs were estimated from a societal perspective in 1994 U.S. dollars. Net costs for 10 years of treatment were estimated for each therapy.

Main Results

The rate of stroke was the most important variable affecting cost-effectiveness. The quality-adjusted life expectancy in 65-year-old patients at high risk for stroke (i.e., patients with NVAF and ≥ 2 additional risk factors for stroke) was 6.51 years with warfarin, 6.27 years with aspirin, and 6.01 years with no therapy. Because of their high annual rate of stroke (4.9% to 17.6%), warfarin improved quality-adjusted life expectancy and reduced costs in high-risk patients. In patients with NVAF and 1 additional risk factor for stroke (annual stroke rate 2.6% to 4.6%), the cost-effectiveness of warfarin compared with aspirin was $8000 per quality-adjusted life-year (QALY) gained. In patients aged 65 years with NVAF alone (annual stroke rate 1.1% to 2.1%), the cost-effectiveness of warfarin compared with aspirin was $370 000 per QALY saved. In patients aged 75 years with NVAF alone, the cost-effectiveness of warfarin was $110 000 per QALY saved. Compared with no therapy, aspirin prolonged quality-adjusted life expectancy and saved costs in all patients who were not candidates for warfarin.


Warfarin therapy was cost-effective for prophylaxis of stroke in patients aged 65 years who had nonvalvular atrial fibrillation and at least 1 additional risk factor for stroke.

Sources of funding: Palo Alto Institute for Research and Education and the Veterans Affairs Health Services Research and Development Field Program.

For article reprint: Dr. B.F. Gage, Washington University School of Medicine, Campus Box 8005, 660 South Euclid Avenue, St. Louis, MO 63110-1093, USA.


In this analysis, Gage and colleagues made the decision to offer warfarin, aspirin, or no treatment to prevent stroke in patients with NVAF. The authors used a combination of newly collected data on cost and quality of life together with published data on stroke and bleeding. A strength of their model is that the probabilities of stroke and hemorrhage are based on meta-analytic estimates from well-done, randomized trials.

This study lends formal support to the clinical strategy of anticoagulating patients who are aged 65 to 74 years and are at medium to high risk for stroke. This strategy is clinically effective and actually saves health care costs among the highest risk patients (≥ 2 risk factors for stroke), which is unusual for a preventive intervention. In low-risk patients (65-year-old persons with NVAF only), the authors suggest that the small increment in effectiveness of warfarin compared with aspirin may not justify its additional costs. This conclusion should be interpreted cautiously because evidence for the effectiveness of aspirin is limited. In any case, the societal cost of warfarin compared with aspirin in low-risk patients should not be overly emphasized because these patients are a small proportion of the population with NVAF. Not surprisingly, cost-effectiveness is sensitive to both stroke and bleeding rates. Because concern exists that the rates found in closely monitored clinical trials may not be replicable in actual practice, more data on these rates for different groups of patients are needed.

As a public health problem, NVAF is an issue mainly of elderly persons (> half of patients are aged > 75 y [1]). This study focused on patients aged 65 to 74 years, where the data from clinical trials are most applicable. In sensitivity analysis, however, the model suggests that the cost-effectiveness of warfarin might improve in patients who are aged 75 years. Empirical data to support this prediction are urgently needed.

Tim Lancaster, MD
Radcliffe InfirmaryOxford, England, UK


1. Feinberg WM, Blackshear JL, Laupacis A, Kronmal R, Hart RG. Prevalence, age distribution, and gender of patients with atrial fibrillation. Arch Intern Med. 1995;155:469-73.