Skip to main content

29-06-2011 | Stroke | Article

Stroke risk scores underestimate true risk in elderly patients with AF


Free abstract

MedWire News: The seven most widely used stroke risk scores greatly underestimate the actual risk in older people with atrial fibrillation (AF) who are not taking warfarin, UK researchers have found.

Writing in the BMJ, they say there may be a "pragmatic rationale" for classifying all patients over 75 years of age as "high risk," at least until better tools are available.

The research, by Richard Hobbs (University of Oxford) and team, compared rates of stroke and thromboembolism among 665 patients aged 75 years and above who took part in the Birmingham Atrial Fibrillation in the Aged (BAFTA) trial.

BAFTA was a randomized trial comparing warfarin with aspirin for thromboprophylaxis in elderly patients with AF. However, the present study included only those patients who were not assigned to warfarin.

The researchers compared actual rates of thromboembolism with the rates predicted by retrospectively applied stroke risk scores. They tested seven scoring systems: CHADS2; Framingham; the UK National Institutes of Health and Clinical Excellence (NICE) guidelines; the American College of Cardiology/American Heart Association/European Society of Cardiology (ACC/AHA/ESC) guidelines; the American College of Chest Physicians guidelines, and the more recently developed CHA2DS2-VASc and Reitbrock-modified CHADS2 scheme.

Over a median follow-up period of 2.2 years, the actual rates of ischemic stroke, systemic embolism, and transient ischemic attack were 8.0%, 0.6%, and 2.0%, respectively.

Three risk scores - revised CHADS2, NICE, and ACC/AHA/ESC - performed similarly, each classifying around two-thirds of patients as "high risk" and the remainder as "moderate risk."

The original CHADS2 score classified the smallest proportion of patients as "high risk," at just 27%. Additionally, all three CHADS2-based scores failed to show an increase in risk at the upper range of scores, showing similar risk rates at scores of 2 and above.

Furthermore, the original CHADS2score failed to outperform simple chance, with a c-statistic of just 0.55. Predictive accuracy was comparable - and poor - across the other six scores, with c-statistics ranging from 0.59 (Framingham) to 0.62 (Rietbrock modified CHADS2).

Hobbs and co-authors conclude that the risk stratification schemes advocated in international guidelines appear to have only limited accuracy for predicting stroke risk in elderly AF patients.

"These data are important as high-risk AF patients should be offered vitamin K antagonists in most circumstances, yet vitamin K antagonists are underused for stroke prevention," they write. "Moreover, physicians' reluctance to use vitamin K antagonists has repeatedly been shown to be associated with a false understanding of the benefit-to-risk ratio.

"Until better scores are available, clinicians might consider all patients over 75 as at high risk and offer formal anticoagulation rather than aspirin."

MedWire ( is an independent clinical news service provided by Springer Healthcare Limited. © Springer Healthcare Ltd; 2011

By Joanna Lyford

Related topics