Treatment patterns and real-world effectiveness of warfarin in nonvalvular atrial fibrillation within a managed care system

Curr Med Res Opin. 2005 Oct;21(10):1583-94. doi: 10.1185/030079905X61956.


Objective: To examine warfarin utilization and clinical effectiveness among patients with nonvalvular atrial fibrillation within usual clinical care in a managed care system.

Research design and methods: A retrospective analysis of health care claims for an approximately four million member managed care organization was performed. Health plan members with a diagnosis of nonvalvular atrial fibrillation in calendar year 2000 were identified and stratified into two cohorts: Warfarin Therapy (newly initiating warfarin) or Warfarin Candidates (eligible for warfarin therapy according to the ACC/AHA/ESC Guidelines for the Management of Patients with Atrial Fibrillation, but did not receive warfarin).

Measurements: The occurrence of thromboembolism, ischemic stroke, and hemorrhage during a maximum 720-day follow-up were compared between cohorts, adjusting for age, gender, and other risk factors, using Cox regression.

Results: Among 12 539 subjects (mean age 78.0 +/- 8.8 years) with nonvalvular atrial fibrillation, 4895 (39.0%) initiated Warfarin Therapy and 7644 (61.0%) were Warfarin Candidates. Event occurrences among Warfarin Therapy vs. Warfarin Candidates were: ischemic stroke, 3.7% vs. 4.5%; any thromboembolism, 7.8% vs. 10.8%; and hemorrhage, 4.4% vs. 4.9%, respectively. Warfarin therapy was not associated with an increased risk for hemorrhage (hazard ratio [HR] = 0.97, 95% confidence interval [CI] = 0.82-1.15), while risks for ischemic stroke and any thromboembolism were significantly reduced, by 22% (HR = 0.78, 95% CI = 0.65-0.93) and 34% (HR = 0.66, 95% CI = 0.59-0.75), respectively.

Conclusions: Within usual clinical care for the managed care population examined, warfarin remains underused despite current guidelines recommending its use in nearly all patients with nonvalvular atrial fibrillation. Although utilization of anticoagulation clinics and INR values attained were unknown in this study, the observed risk reductions for ischemic stroke and thromboembolism were lower than those achieved in clinical trials, while no increased risk for hemorrhage was observed. These findings suggest that warfarin is used conservatively, and dosed cautiously, diminishing the full potential benefit of anticoagulant therapy in patients with nonvalvular atrial fibrillation.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Aged
  • Anticoagulants / administration & dosage
  • Anticoagulants / therapeutic use*
  • Atrial Fibrillation / complications
  • Atrial Fibrillation / drug therapy*
  • Female
  • Hemorrhage / etiology
  • Humans
  • Male
  • Managed Care Programs*
  • Retrospective Studies
  • Stroke / etiology
  • Thromboembolism / etiology
  • Warfarin / administration & dosage
  • Warfarin / therapeutic use*


  • Anticoagulants
  • Warfarin