ABSTRACT
OBJECTIVE: To evaluate the role of CHADS2 score on predicting ischaemic stroke or transient ischaemic attack (TIA) and death in patients with coronary artery disease (CAD), irrespective of the presence or absence of atrial fibrillation (AF). METHODS: We searched for cohort studies that reported risk estimates for incidence of ischaemic stroke/TIA or mortality by levels of CHADS2 score in Medline/PubMed and Embase. Random effects models were used to calculate pooled risk ratios (RRs) and 95% confidence intervals (CIs). RESULTS: Eight cohort studies (7 prospective and 1 retrospective) enrolling 31,509 patients with CAD were included. The pooled RR of mortality was 2.38 (95% CI 1.63-3.47) for CHADS2 score≥2, and of stroke/TIA incidence was 2.19 (1.55-3.08). In patients without AF, CHADS2 score≥2 was associated with increased mortality (pooled RR 3.14 95% CI 2.14-4.61) and stroke/TIA incidence (pooled 2.81, 2.08-3.78) In patients with AF, the pooled RR of mortality for CHADS2 score≥2 was 1.57 (1.07-2.28), but no significant association was found between CHADS2 score and stroke/TIA incidence (pooled RR 1.21, 95% CI 0.84-1.73). CONCLUSIONS: CHADS2 score can predict mortality in patients with CAD. However, higher CHADS2 score is associated with increased incidence of stroke/TIA only in patients without AF.