PO-04-155 LONG TERM POST-ABLATION ATRIAL FIBRILLATION MONITORING USING AUTOMATED SINGLE-LEAD ELECTROCARDIOGRAM ANALYSIS

Abstract

Wearable and handheld consumer electrocardiogram devices may allow for noninvasive, frequent, and symptom driven monitoring after atrial fibrillation (AF) ablation. We sought to evaluate the performance of the KardiaMobile device and AliveCor algorithm in monitoring for arrythmia recurrence after AF ablation. Patients in a prospective randomized controlled trial of adjunctive therapy in AF ablation (NCT04625946) were provided with a KardiaMobile and asked to obtain symptomatic or at least weekly recordings before ablation and during 1 year of follow up. A total of 3174 recordings from 50 patients were obtained. The performance of the automated diagnosis was compared to the human interpretation which was considered the gold standard. A total of 772 tracings (24%) demonstrated AF occurring in 38/50 patients. The automated analysis had an overall sensitivity of 90.2% and specificity of 94.3% for the diagnosis of AF. The precision of the automated analysis had significant inter-patient variation (interquartile range 66.6-100%). In 3/12 patients without tracings of AF the automated analysis assigned a diagnosis of AF to one or more tracings. Inaccurate identifications of AF were most frequently adjudicated as sinus rhythm (47.7%), sinus rhythm with premature atrial contractions (23.5%), atrial flutter (8.8%), or uninterpretable (6.0%). A total of 58 tracings demonstrated atrial tachycardia or atrial flutter and were classified as ‘tachycardia’ (26%), AF (21%), or ‘unclassified’ (53%) by the automated analysis. The use of intermittent home electrocardiogram monitoring may facilitate the diagnosis of recurrent atrial arrythmias after AF ablation. Manual adjudication of the automated diagnosis or additional clinical event monitoring is likely required especially in patients with ectopy or postoperative atrial flutter.