Comparison of CHA2DS2-VA and CHA2DS2-VASc scores for predicting residual risk of stroke and adverse outcomes in anticoagulated patients with atrial fibrillation: a report from a European registry

Background The 2024 European Society of Cardiology (ESC) guidelines on atrial fibrillation (AF) recommend the use of the non-sex CHA2DS2-VASc (ie. CHA2DS2-VA) score to stratify thromboembolic risk in AF patients. Comparative data on the performance of the CHA2DS2-VA vs. CHA2DS2-VASc scores in predicting residual thromboembolic risk in anticoagulated AF patients are limited. Purpose To compare the performance of the CHA2DS2-VA and CHA2DS2-VASc scores in predicting residual risk of adverse events in an oral anticoagulant (OAC)-treated AF population. Methods From a prospective observational registry of European AF patients, we included those treated with OAC. The CHA2DS2-VA and CHA2DS2-VASc scores were calculated according to their original schemes. The primary endpoint was the occurrence of any stroke or thromboembolic events. The secondary endpoint was a composite of all-cause death and major adverse cardiovascular events (MACEs). Receiver operating characteristic (ROC) curves and reclassification analysis were used to compare the predictive ability of the scores. We evaluated the interaction between sex and the CHA2DS2-VA score by plotting an adjusted Cox regression analysis with a restricted spline curve for both primary and secondary outcomes. Results A total of 9,437 patients (mean age 69.7 ± 3.5 years, 40.6% female) treated with OAC were included. After a median follow-up of 729 days (IQR 365-735), 182 (2.1%) events of the primary endpoint (stroke/thromboembolism) and 1,213 (14.3%) events of the secondary composite outcome were observed. Both scores demonstrated modest predictive ability for the primary endpoint (CHA2DS2-VA AUC, 0.60; 95% confidence interval [CI]: 0.56-0.64 vs. CHA2DS2-VASc AUC, 0.61; 95% CI: 0.56-0.65, p = 0.665, Figure 1). The CHA2DS2-VA score showed marginal but statistically better performance compared to CHA2DS2-VASc for the secondary composite outcome (AUC, 0.66; 95% CI: 0.65-0.68 vs. 0.65; 95% CI: 0.64-0.67, p < 0.001, Figure 1). Reclassification analysis revealed no significant difference for stroke/thromboembolism, but a marginal, statistically significant improvement in IDI for the CHA2DS2-VA score vs. CHA2DS2-VASc (IDI = 0.004; 95% CI: 0.000-0.006, p < 0.001) for the composite outcome. No significant interaction effects were found between CHA2DS2-VA score and sex for the risk of both the primary and secondary outcomes (p for interaction = 0.322 and 0.777, respectively, Figure 2). Conclusion In a contemporary cohort of OAC-treated AF patients, the CHA2DS2-VA and CHA2DS2-VASc scores showed broadly similar performance for the prediction of adverse events. Excluding sex from current risk stratification models does not result in clinically meaningful improvements of risk stratification in OAC-treated AF patients.Figure 1ROC curves. Figure 2.Spline curves