A risk score for predicting stroke or death in individuals with new-onset atrial fibrillation in the community: the Framingham Heart Study

JAMA. 2003 Aug 27;290(8):1049-56. doi: 10.1001/jama.290.8.1049.


Context: Prior risk stratification schemes for atrial fibrillation (AF) have been based on randomized trial cohorts or Medicare administrative databases, have included patients with established AF, and have focused on stroke as the principal outcome.

Objective: To derive risk scores for stroke alone and stroke or death in community-based individuals with new-onset AF.

Design, setting, and participants: Prospective, community-based, observational cohort in Framingham, Mass. We identified 868 participants with new-onset AF, 705 of whom were not treated with warfarin at baseline. Risk scores for stroke (ischemic or hemorrhagic) and stroke or death were developed with censoring when warfarin initiation occurred during follow-up. Event rates were examined in low-risk individuals, as defined by the risk score and 4 previously published risk schemes.

Main outcome measures: Stroke and the combination of stroke or death.

Results: During a mean follow-up of 4.0 years free of warfarin use, stroke alone occurred in 83 participants and stroke or death occurred in 382 participants. A risk score for stroke was derived that included the following risk predictors: advancing age, female sex, increasing systolic blood pressure, prior stroke or transient ischemic attack, and diabetes. With the risk score, 14.3% of the cohort had a predicted 5-year stroke rate < or =7.5% (average annual rate < or =1.5%), and 30.6% of the cohort had a predicted 5-year stroke rate < or =10% (average annual rate < or =2%). Actual stroke rates in these low-risk groups were 1.1 and 1.5 per 100 person-years, respectively. Previous risk schemes classified 6.4% to 17.3% of subjects as low risk, with actual stroke rates of 0.9 to 2.3 per 100 person-years. A risk score for stroke or death is also presented.

Conclusion: These risk scores can be used to estimate the absolute risk of an adverse event in individuals with AF, which may be helpful in counseling patients and making treatment decisions.

Publication types

  • Research Support, Non-U.S. Gov't
  • Research Support, U.S. Gov't, P.H.S.

MeSH terms

  • Aged
  • Aged, 80 and over
  • Anticoagulants / therapeutic use
  • Atrial Fibrillation / complications*
  • Atrial Fibrillation / drug therapy
  • Female
  • Humans
  • Male
  • Middle Aged
  • Proportional Hazards Models
  • Risk Assessment
  • Stroke / epidemiology*
  • Stroke / etiology*
  • Stroke / prevention & control
  • Survival Analysis
  • Warfarin / therapeutic use


  • Anticoagulants
  • Warfarin