CHA2DS2-VASc Score Best for Stroke Risk Assessment in AF

September 19, 2013

Further data confirming that the CHA2DS2-VASc score is preferable to the CHADS2 score in identifying patients with atrial fibrillation (AF) at very low risk for stroke who do not need anticoagulation have come from a large German registry study.

In the AFNET (German Competence Network on Atrial Fibrillation) registry, which included 8847 patients with nonvalvular AF, more than one third of the strokes or other thromboembolic events that occurred in the mean follow-up period of 5 years were in patients assigned a CHADS2 score of 0 or 1, indicating that oral anticoagulation is not definitively recommended.

In contrast, using the CHA2DS2-VASc score, which adds age 65 to 74 years, vascular disease, and female sex as stroke risk factors to the CHADS2 score, led to the reclassification of more than half of the patients with CHADS2 scores 0 or 1 to a CHA2DS2-VASc score of 2 and higher, in whom oral anticoagulation is the recommended treatment.

Only 8 strokes and other thromboembolic events were observed in patients classified as CHA2DS2-VASc 0 at baseline (excluding strokes occurring in association with cardioversion or ablation).

"A CHADS2 score of 0 or 1 does not appear to be suitable to identify patients with AF at low risk for stroke while CHA2DS2-VASc picks up these patients," said Professor Michael Nabauer, MD, University Hospital Grosshadern, Munich, Germany, at a press conference held during the European Society of Cardiology (ESC) Congress last week.

"The risk of stroke in patients with a CHA2DS2-VASc score of 0 over a mean follow-up of 5.1 years was very low. Our data support the current recommendation that oral anticoagulation is not beneficial in patients with 'lone AF' or a CHA2DS2-VASc score of 0."

The analysis was based on the prospective AFNET registry, which included 9575 patients with AF, 8847 of whom had nonvalvular AF. Follow-up was for a mean of 5.1 years, and stroke events were adjudicated by a critical event committee.

On enrollment, almost half of patients with nonvalvular AF were assigned a CHADS2 score of 0 (16.2%) or 1 (31.5%), indicating that oral anticoagulation was not definitively recommended. But 145 of the 405 (38%) strokes and other thromboembolic events (transient ischemic attack, systemic embolism) during follow-up occurred in these patients classified as "low-risk."

Table 1. Events in Patients by CHADS2 Scores

Endpoint CHADS2 Score ≤ 2 CHADS2 Score = 1 CHADS2 Score = 0
Patients (n) 4631 2786 1430
Events (stroke/transient ischemic attack/systemic embolism) (n) 258 100 45


Application of the CHA2DS2-VASc score reclassified 54% of patients with CHADS2 scores 0 or 1 to a CHA2DS2-VASc score of 2 and higher, in whom oral anticoagulation is the recommended treatment. In addition, only about half the patients previously classified as having a CHADS2 score of 0 were assigned a CHA2DS2-VASc score of 0, indicating a very low stroke risk.

Table 2. Events in Patients With CHADS2 Score 0 and 1 Reclassified by CHA2DS2-VASc Score

Endpoint CHA2DS2-VASc ≤ 2 CHA2DS2-VASc Score = 1 CHA2DS2-VASc Score = 0
Patients (n) 2270 1133 795
Events (stroke/transient ischemic attack/systemic embolism) (n) 90 36 19


Of the 19 events, in the patients with CHA2DS2-VASc of 0, 4 were periprocedural events, and 7 occurred after progression of CHA2DS2-VASc score, leaving 8 residual events in this group.

Professor Guenter Breithardt, who was also involved in the study, explained to Medscape Medical News that the CHA2DS2-VASc score was already recommended in preference to the CHADS2 score in the guidelines. "But we need more confirmatory data and this is what this study provides."

… this is sound substantiation of what we are now coming to accept. Professor Guenter Breithardt

"This was a large registry with all levels of care included, not just one subgroup, so it is a realistic picture of the German population," he said. "Also there was a critical event committee to adjudicate all strokes, which is unique for a registry study. And we had over 95% follow-up which is again unusual for a registry study. So this is sound substantiation of what we are now coming to accept."

Professor Breithardt pointed out that a large number of strokes occur in low-risk patients. "Percentage-wise they are low risk but the numbers are substantial. This is similar to ejection fraction and MI [myocardial infarction]: while ejection fraction less than 35 is high risk, patients with ejection fractions greater than 35 account for the majority of sudden deaths. So looking at the low risk population is very valuable.

"Previously low-risk AF patients have not been treated properly. Many of them need anticoagulant therapy but have not been given it," he added. "The CHA2DS2-VASc score picks these patients up."

Also commenting for Medscape Medical News, Professor Robert Hatala, PhD, National Cardiovascular Institute, Bratislava, Slovakia, agreed that CHA2DS2-VASc score is less sensitive to misjudgment because it has more components. "It definitely gives a more precise prediction of risk."

More Education Needed

But he stressed that much more education of doctors is needed on these scores. He noted that at a recent symposium he attended, mostly with cardiologists and internists in the audience, attendees were asked if they knew what the CHADS2 score was and if they used it. "One third said they yes to both, one third said they knew what it was but didn't use it, and one third didn't even know what it was."

"Now that we have the new oral anticoagulant drugs that are easy to administer, we need to make sure their use is maximized to prevent stroke, and the most cost-effective way to do this is to make stroke risk scores such as CHA2DS2VASc become everyday routine by the primary care physician. We need to drum home this message."

Chair of the ESC press conference at which this study was discussed, Professor Harry Crijns, Maastricht University Medical Center, the Netherlands, added: "The most important point to take from this work is that during follow-up patients may develop new risk factors: they grow older or develop clinically significant hypertension or other vascular disease. Monitoring such clinical changes during follow-up and installing anticoagulation and other vascular prophylactic therapies as needed may save lives," he said.

"Another message here is that the very-low-risk CHA2DS2-VASc 0 AF patient may suffer cardiovascular events during subsequent interventional treatment indicating that low risk does not preclude adverse events of interventions such as ablation or cardioversion."

European Society of Cardiology (ESC) Congress 2013. Abstract 4381. Presented September 3, 2013.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.