my personal edition > stroke > news


  E-Mail this DGReview to a colleague

DGReview

Risk Scores Predict Stroke and Mortality Among Patients with New Onset Atrial Fibrillation

The risk of stroke alone or stroke and death over a 5-year period among individuals with new onset atrial fibrillation can be estimated with a set of risk scores established using findings from the Framingham Heart Study.

"An understanding of absolute risk is fundamental to many clinical decisions involving patients with atrial fibrillation," writes Emelia J. Benjamin, MD, ScM, and colleagues with the Framingham Heart Study, Massachusetts, United States. Among a prospective, community-based, observational cohort living in Framingham, the researchers used Cox proportional hazards models to assess predictors for developing stroke (either ischaemic or hemorrhagic) and stroke or death by censoring the initiation of warfarin treatment during follow up. Of the overall 868 patients with new onset atrial fibrillation, 705 did not receive warfarin at baseline.

After censoring for warfarin use during a mean follow up of 4.0 years the researchers found that 83 stroke and 382 stroke or death events occurred. Risk predictors for stroke, including advancing age, female sex, increasing systolic blood pressure, diabetes, and prior transient ischaemic attack or stroke, were used to derive a risk score for stroke. These scores, when added together, can estimate the 5-year risk of stroke based on the individualised data of each patient with new onset atrial fibrillation.

"With the risk score," Dr. Benjamin and colleagues write, "14.3% of the cohort had a predicted 5-year stroke rate less than or equal to 7.5% and 30.6% of the cohort had a predicted 5-year stroke rate of less than or equal 10%."

Similar predictors, with the addition of prior congestive heart failure or myocardial infarction, smoking, significant heart murmur, and electrocardiographic left ventricular hypertrophy, were used to derive a risk score for stroke or death.

The investigators note that the risk associated with atrial fibrillation is highly variable. "These risk scores can be used to estimate the absolute risk of an adverse event in individuals diagnosed with atrial fibrillation," the authors write, "which may be helpful in counselling patients and in making treatment decisions."

Dr. Benjamin and colleagues conclude however, "because several risk schemes for atrial fibrillation now exist, it will be critical to validate this risk prediction instrument in an independent cohort and to assess its performance relative to other risk schemes."
JAMA 2003;290:1049-1056.

E-Mail this DGReview to a colleague

To print, use this version