Embolic stroke of undetermined source (ESUS) accounts for up to one quarter of ischemic strokes, and occult atrial fibrillation (AF) remains a key underlying cause. This study published in the International Journal of Stroke analyzed 543 consecutive ESUS patients systematically monitored with implantable cardiac monitors (ICMs). The objective was to develop and validate a clinically applicable tool for predicting early and long-term atrial fibrillation (AF) detection. Variable selection was performed using LASSO-penalized Cox regression. Model performance was assessed through time-dependent receiver operating characteristic curves, restricted mean survival time (RMST) analysis, 10-fold cross-validation, and internal-external cross-validation across seven participating centers to evaluate geographic generalizability.

Over 1558.5 patient-years of follow-up, 118 patients (22%) developed new AF. The final point-based CATCH-AF score included age, coronary artery disease, heart failure, and prior transient ischemic attack or ischemic stroke. The model demonstrated strong discrimination with an area under the curve (AUC) of 0.85 (95% CI 0.82–0.89), remaining stable over 4.5 years (AUC range 0.84–0.87). Compared with low-risk patients (0–2 points), high-risk individuals (≥5 points) had a 19-fold higher hazard of AF detection (HR 19.2; 95% CI 9.4–39.4; p<0.001) and 918 fewer AF-free days (95% CI −1080 to −757).

The CATCH-AF score offers a clinically interpretable and practical approach to estimating AF risk after ESUS and may support more targeted and cost-conscious rhythm monitoring strategies. Further validation in broader populations may help clarify its role in guiding rhythm monitoring strategies.