Clinical utility and prognostic implications of the novel 4S-AF scheme to characterize and evaluate patients with atrial fibrillation: a report from ESC-EHRA EORP-AF Long-Term General Registry

Aims The 4S-AF classification scheme comprises of four domains: stroke risk (St), symptoms (Sy), severity of atrial fibrillation (AF) burden (Sb), and substrate (Su). We sought to examine the implementation of the 4S-AF scheme in the EORP-AF General Long-Term Registry and compare outcomes in AF patients according to the 4S-AF-led decision-making process. Methods and results Atrial fibrillation patients from 250 centres across 27 European countries were included. A 4S-AF score was calculated as the sum of each domain with a maximum score of 9. Of 6321 patients, 8.4% had low (St), 47.5% EHRA I (Sy), 40.5% newly diagnosed or paroxysmal AF (Sb), and 5.1% no cardiovascular risk factors or left atrial enlargement (Su). Median follow-up was 24 months. Using multivariable Cox regression analysis, independent predictors of all-cause mortality were (St) [adjusted hazard ratio (aHR) 8.21, 95% confidence interval (CI): 2.60-25.9], (Sb) (aHR 1.21, 95% CI: 1.08-1.35), and (Su) (aHR 1.27, 95% CI: 1.14-1.41). For CV mortality and any thromboembolic event, only (Su) (aHR 1.73, 95% CI: 1.45-2.06) and (Sy) (aHR 1.29, 95% CI: 1.00-1.66) were statistically significant, respectively. None of the domains were independently linked to ischaemic stroke or major bleeding. Higher 4S-AF score was related to a significant increase in all-cause mortality, CV mortality, any thromboembolic event, and ischaemic stroke but not major bleeding. Treatment of all 4S-AF domains was associated with an independent decrease in all-cause mortality (aHR 0.71, 95% CI: 0.55-0.92). For each 4S-AF domain left untreated, the risk of all-cause mortality increased substantially (aHR 1.35, 95% CI: 1.16-1.56). Conclusion Implementation of the novel 4S-AF scheme is feasible, and treatment decisions based on this scheme improve mortality rates in AF.