Abstract
BACKGROUND: Insertable cardiac monitors (ICMs) provide long-term continuous monitoring for arrhythmia diagnosis and management for various clinical indications. However, little data exists on comprehensive real-world arrhythmia diagnostic yield and therapy rates in patients indicated for ICMs with validated artificial intelligence (AI) algorithms enabling large-scale, automated adjudication of ICM-detected episodes. We report the largest real-world analysis of arrhythmia detection as well as medical and procedural therapies in patients with ICMs implanted for guideline-approved indications with long-term monitoring. METHODS: Patients who received a Reveal LINQ ICM between October 1, 2016, and June 30, 2020, with ≥ 1 year of follow-up were identified in two databases (Medtronic CareLink data warehouse, N = 12 020, and Optum Clinformatics Data Mart claims database, N = 17 037) to analyze arrhythmia detections and therapeutic interventions, respectively. Patients were categorized by clinical indication for ICM placement. All device-detected ECGs were identified and processed through arrhythmia-specific AI algorithms. Therapeutic interventions included procedural interventions (cardiovascular implantable electronic device implantation, cardioversions, and ablations) and medication initiation or titration (antiarrhythmics, rate-control medications, and oral anticoagulants) after ICM implant. RESULTS: Mean (SD) follow-up in the CareLink and Clinformatics claims databases was 24.6 (12.7) and 40.8 (15.6) months, respectively. Of the 12 020 patients in the arrhythmia detection analysis, 7284 (60.6%) had ≥ 1 arrhythmia detected (56.3% in the suspected AF population; 80.1% in the AF management population), and 376 (28.9%) had ≥ 2 arrhythmias detected during long-term follow-up. Among syncope patients with arrhythmia(s) detected, 71.2% had a finding other than pause/bradycardia; 50.4% of cryptogenic stroke patients and 62.6% of AF management patients with arrhythmias had ≥ 1 finding other than AF. Of the 17 037 patients in the therapeutic interventions analysis, 9820 (57.6%) had a therapeutic action post-ICM insertion, with 25% of all patients receiving a procedural intervention, and > 50% undergoing a medication adjustment. Mean (SD) follow-up to first arrhythmia detection was 7 (9) months. Mean (SD) duration from ICM insertion to therapeutic action was 13 (13) months for procedures and 7 (11) months for medication initiation. CONCLUSIONS: Long-term continuous monitoring with ICMs enables identification of multiple arrhythmias that may have otherwise remained undetected and rules out arrhythmias in ~40% regardless of indication. Medication adjustments and/or procedural interventions related to the management of arrhythmias were observed in over half of ICM recipients during long-term follow-up.