The Detection of Atrial Fibrillation by Implantable Event Recorders: A Study of Patients Who Have Had a Cryptogenic Stroke.

BACKGROUND: Arial fibrillation (AF) plays a role in approximately one-third of ischemic strokes. In embolic strokes of undetermined source (ESUS), paroxysmal atrial fibrillation often escapes detection; if detected, it could be an indication for oral anticoagulation. In the ARTESiA trial, the absolute benefit of anticoagulation was higher in patients who had already had a stroke or transient ischemic attack (TIA) than in those who had not. METHODS: We retrospectively analyzed data from 233 consecutive ESUS patients at the Jena University Hospital who had received an implanted loop recorder (ILR) over the period 2021-2025. We additionally provide a narrative review on the detection rates associated with various monitoring strategies based on pertinent publications that were retrieved by a literature search. RESULTS: The 12-month detection rate of paroxysmal AF was 24.0% (56/233, 95% confidence interval [19.0; 29.9]). 73.2% of the initial detections were within the first 90 days. All patients in whom AF was detected were orally anticoagulated. The 12-month detection rate was higher than in the CRYSTAL-AF trial (12.4%) and the PER-DIEM trial (15.3%), but lower than was reported by Pikija et al. (31.3%). The cutoff values with the highest positive predictive value were a stroke risk analysis (SRA) index of 2 or above (38.4%) and age 75 or above (37.9%). CONCLUSION: Early, risk-adapted diagnostic testing with ILR in selected ESUS patients can detect clinically relevant AF. Advanced age and an elevated SRA index may be useful patient selection criteria. A robust number needed to treat (NNT) for the ILR strategy cannot be derived from the available one-armed registry data. The indication for ILR should be linked to its therapeutic relevance, the probability of detecting AF, and the individual hemorrhagic risk.