From Washington University School of Medicine
New method of predicting stroke in heart patients
St. Louis, June 13, 2001 — Researchers at Washington University School of Medicine in St. Louis have developed a formula to predict the risk of stroke in patients with an irregular heart rhythm called atrial fibrillation.
“Our hope is that this new classification scheme will help physicians select the appropriate course of treatment for patients with atrial fibrillation,” says Brian F. Gage, M.D., who led the study. Gage is assistant professor of medicine at the School of Medicine and medical director of Barnes-Jewish Hospital’s blood thinner clinic. The results are published in the June 13 issue of the Journal of the American Medical Association.
Patients with atrial fibrillation, an irregular, uncoordinated contraction of heart muscles, are estimated to have a fivefold increased risk of stroke. A blood thinner called warfarin sodium (sold as Coumadin® and others) often is used to reduce this risk, but the drug itself can cause hemorrhage and other side effects. It also is more expensive and more difficult to administer and monitor than the alternative treatment, aspirin.
To help predict when the benefits of warfarin outweigh the risks, two earlier studies completed by two other research groups determined independent factors that significantly increase the risk of stroke. However, the studies reached somewhat different conclusions: The Atrial Fibrillation Investigators (AFI) found that stroke risk correlated with prior stroke, advanced age, hypertension and diabetes; the Stroke Prevention and Atrial Fibrillation (SPAF) team found that prior stroke, blood pressure, recent heart failure and the combination of being over 75 years old and female increased the risk of stroke.
“The two predictor models were helpful, but discrepancies between them sometimes led to confusion,” says Gage. “We needed a simple, uniform system to help select warfarin for patients at moderate or high risk of stroke, while avoiding this potentially dangerous blood thinner in low-risk patients.”
So Gage and colleagues combined the factors from both models and developed a points system called CHADS2, an acronym for the five factors: Congestive heart failure, Hypertension, Age, Diabetes and Stroke. Since both the AFI and SPAF found that a history of stroke is the best predictive factor, it was given a value of two points, delineated by the “2” at the end of the mnemonic. The other factors each are allocated one point. Patients therefore are assigned a score ranging from 0 to 6. In general, the researchers suggest prescribing warfarin to patients with a CHADS2 rating of one or greater, depending on the patient’s preferences and risk of hemorrhage.
In collaboration with Peer Review Organizations representing seven states, the team obtained data from 1,733 Medicare beneficiaries aged 65 to 95 years. They followed each patient for an average of 1.2 years and assembled a National Registry of Atrial Fibrillation (NRAF). They then compared the predictive value of each of the three models — CHADS2, AFI and SPAF.
The AFI and SPAF schemes both predicted stroke better than chance, but CHADS2 yielded significantly more accurate results than either of these models. In addition, the risk of stroke as estimated using CHADS2 ranges from less than two percent to roughly 18 percent. Both AFI and SPAF include only three categories — low, moderate and high risk — with stroke risk ranging from roughly one percent to ten percent.
“Having a wider range of scores provides a more quantitative approach to predicting stroke, which is very helpful,” explains Gage. “For example, even for high-risk patients, it’s important to know how high their score is so that you can take extra precautions if necessary during future surgeries and other medical treatments.”
Gage BF, Waterman AD, Shannon W, Boechler M, Rich MW, Radford MJ. Validation of clinical classification schemes for predicting stroke. Journal of the American Medical Association, 285(22), 2864-2870, June 13, 2001.
Funding from the Agency for Healthcare Research and Quality supported this research.
The full-time and volunteer faculty of Washington University School of Medicine are the physicians and surgeons of Barnes-Jewish and St. Louis Children's hospitals. The School of Medicine is one of the leading medical research, teaching and patient-care institutions in the nation. Through its affiliations with Barnes-Jewish and St. Louis Children's hospitals, the School of Medicine is linked to BJC Healthcare.