Article Text


Prediction of stroke in patients without atrial fibrillation using the CHADS2 and CHA2DS2-VASc scores: a justification for more widespread thromboprophylaxis?
  1. Keitaro Senoo,
  2. Gregory Y H Lip
  1. University of Birmingham, Centre for Cardiovascular Sciences, City Hospital, Birmingham, UK
  1. Correspondence to Professor Gregory Y H Lip, University of Birmingham, Centre for Cardiovascular Sciences, City Hospital, Birmingham B18 7QH, UK; g.y.h.lip{at}

Statistics from

The CHADS2 (congestive heart failure, hypertension, age ≥75 years, diabetes mellitus, previous stroke/transient ischaemic attack (TIA) (double)) and more recently, the CHA2DS2-VASc (congestive heart failure, hypertension, age ≥75 years (double score), diabetes mellitus, previous stroke/TIA (double score), vascular disease, age 65–74 years, sex category (female)) scores are validated clinical prediction tools commonly used to estimate the stroke risk in atrial fibrillation (AF) patients. More recently, the focus has been on initially identifying those patients at ‘low risk’ (ie, CHA2DS2-VASc score=0 for men, 1 for women) who do not need any antithrombotic therapy. Subsequent to this step, effective stroke prevention can be offered to those patients with AF with ≥1 stroke risk factors.

The CHADS2 and CHA2DS2-VASc scores represent clusters of common cardiovascular risk factors related to thromboembolism. Perhaps unsurprisingly, both scores have been reported to be associated with the incidence of stroke, thromboemblism and mortality, not only in patients with AF but also in general population or non-AF patient populations.

Mitchell et al1 show that the CHADS2 and CHA2DS2-VASc scores could predict the incidence of stroke/TIA in a population of over 20 000 patients with acute coronary syndrome (ACS) without AF. Both scores had acceptable discrimination performance (CHADS2 and CHA2DS2-VASc scores; C statistic=0.68 and 0.71, respectively). The present report added to this literature by confirming previous findings and by refining their estimations in a large cohort of ACS patients.

Given that the CHADS2 and CHA2DS2-VASc scores were introduced as tools to predict stroke risk in patients with AF, were the results of the present study surprising?

First, components of these two scores are recognised to be associated with mortality and ischaemic strokes in non-AF patients. Recently, some studies have shown that mortality was higher as the CHADS2 score increased in patients with ACS without AF.2 ,3 Furthermore, Welles et al4 found that in patients with coronary heart disease without AF, the CHADS2 score was strongly predictive of ischaemic stroke/TIA. The present study validates these previous findings and refines their estimates from a Canadian large cohort.

Second, the CHADS2 and CHA2DS2-VASc scores include the main risk factors for developing incident AF. In the study by Brent et al the CHADS2 and CHA2DS2-VASc scores also predicted subsequent hospital discharge with a diagnosis of AF in those without pre-existing AF. The annual incidence of AF increased in a stepwise fashion and reached 3.85% in patients with CHADS2 ≥3 and 2.52% in those with CHA2DS2-VASc ≥4. Each score had reasonable discrimination performance for prediction of incident AF (c-statistic=0.71 for each score). There was a graded increase in adverse event rates with increasing CHADS2 score, which was seen even after adjusting for incident AF. Therefore, higher scores identify patients more likely to develop AF, which may be asymptomatic. Indeed, such asymptomatic AF may only be first diagnosed when the patient presents with a serious complication, such as stroke or heart failure.

This is in agreement with the finding of Welles et al4 where patients with hypertension admitted for stroke demonstrated a higher prevalence of heart failure, diabetes, coronary heart disease and older age in patients found to have silent AF than those in whom monitoring did not reveal AF. Indeed, Chao et al5 also showed that the incidence of AF progressively increased when patients had more complicated systemic diseases represented by the high CHADS2 score. In the ASSERT trial, 10% of device-implanted patients without known AF had asymptomatic atrial arrhythmias lasting at least 6 min in the first 3 months of monitoring, which associated with subsequent ischaemic stroke.6

Third, cerebral atherosclerosis is more common in patients with higher CHADS2 scores. Kim et al7 indicated that patients at high risk by CHADS2 score had more extracranial and intracranial stenosis as well as proximal stenotic lesions of the symptomatic artery, leading to an increased risk of atherothrombotic stroke. Overall, patients with higher CHADS2 scores have higher risks of atherothrombotic stroke even if patients have no AF.

Finally, each of the component comorbidities of the CHADS2 and CHA2DS2-VASc scores have all been associated with remodelling of the left atrium (LA). This remodelling may result in blood stasis, a prothrombotic state, and give an increased risk of thromboembolism independent of rhythm. These risk scores may, therefore, contribute to stroke risk via LA remodelling.

Given the many reasons above, it is perhaps unsurprising that the CHADS2 and CHA2DS2-VASc scores used to assess the risk of thromboembolic events in patients with non-valvular AF have broadly similar utility for predicting thromboembolism in patients without known AF.

More controversially, should we routinely prescribe oral anticoagulants (OACs, eg, warfarin) for those patients with ACS and without AF in high CHADS2 and CHA2DS2-VASc scores? As these patients are taking antiplatelet therapy for ACS, the risk of major bleeding and intracranial bleeding in combination with OAC for stroke prevention must be well considered, as part of a ‘triple therapy’ regime.

What about non-AF patients with cardiovascular risk factors, that is, primary prevention? Current clinical practice is to prescribe aspirin for patients at high cardiovascular risk, but without AF—but this practice of ‘routine’ aspirin for primary prevention has been questioned, given the fine balance between benefits and risks from regular aspirin use in the primary prevention of cardiovascular disease.8 One may even argue that the incremental benefit of aspirin over contemporary cardiovascular prevention strategies, such as ACE inhibitors and statins, might be very small. In AF patients unsuited for warfarin, or where warfarin was refused, the Non-Vitamin K OACs (NOACs) are even superior to aspirin for preventing thromboembolism, with no significant difference in major bleeding or intracranial bleeding, and with better tolerability.

Thus, if approximately one in three of patients with risk factors develop new onset AF over a year, further clinical trials with (N)OACs may need to address this important question, especially since the harder one looks in ‘high risk’ populations (eg, by improved screening or by using implantable devices), more AF is more likely to be detected in an easy and cost-effective manner.9 What we can do now is that we follow-up such patients regularly and screen carefully for AF. Time will tell, but randomised trials may need to address the issue of whether OACs would be needed.


View Abstract


  • Contributors Both authors wrote the editorial.

  • Competing interests None.

  • Provenance and peer review Commissioned; internally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles