Atrial fibrillation: warfarin was more cost-effective than aspirin at preventing stroke in patients with one or more other risk factors

Clinical bottom line (level 1b)

  1. Warfarin was more cost-effective than aspirin or no treatment in US elderly patients with chronic atrial fibrillation at moderate or high-risk for stroke.
  2. Aspirin was more cost-effective than warfarin in elderly patients at high-risk of major haemorrhage or at low-risk for stroke (< 1% per year).
Gage et al: Journal of the American Medical Association 1995; 274 (23): 1839-1845
Expires November 2003

The study

cost-effectiveness analysis based on a Markov decision model comparing warfarin, aspirin and no treatment over 10 years for 65 year old patients with chronic atrial fibrillation
Setting: NHS provider or HMO



  • Viewpoint: society
  • Benefit assessment: stroke, death, major haemorrhage, intracranial haemorrhage
  • Resources and costs: Annual costs of stroke care taken from US national figures, and cost of prophylaxis from 8 pharmacies and 8 laboratories throughout the USA. Costs were discounted at 5% per year, and were in 1994 US dollars. Costs for routine care of patients with AF were excluded as were indirect costs.
  • Sensitivity analysis: probability of stroke, major haemorrhage, patient demographics

    • The evidence

    intervention cost
    warfarin v. aspirin: low-risk for stroke (AF alone) $370000
    ( marginal cost per QALY )
    warfarin v. no treatment: low-risk for stroke $14 000
    ( marginal cost per QALY )
    warfarin v. aspirin: medium-risk for stroke (AF plus one other risk factor) $8000
    ( marginal cost per QALY )

    Effect of sensitivity analysis: The analysis was affected by the risk of stroke (<1% per year favours aspirin), major haemorrhage, and the disutility of long-term warfarin therapy.
    • Warfarin was always preferred to no treatment for medium-risk patients.
    • Warfarin was always preferred to aspirin or no treatment in high-risk patients (AF plus 2 or more other risk factors).

    Comments

    1. Cost-effectiveness of aspirin in NVAF was very sensitive to the efficacy estimate, suggesting caution in interpretation.
    2. The costs were calculated from the Medicare system, and may not immediately translate into different health care systems.

    Citation

    1. Gage BF, Cardinalli AB, Albers GW, et al: cost-effectiveness of warfarin and aspirin for prophylaxis of stroke in patients with nonvalvular atrial fibrillation. Journal of the American Medical Association 1995; 274 (23): 1839-1845
    Search Terms:
    Contributor: Chris Ball and Clare Wotton, November 1999
    Reviewer: William Rhoton

    Clinical Question.
    Patient elderly with chronic atrial fibrillation
    Intervention or Exposure warfarin
    Comparison aspirin
    Outcome cost-effectiveness