P5648Efficiency and accuracy of arrhythmia detection using implantable cardiac monitor: A prospective multicenter randomized clinical trial comparing Reveal LINQ and Confirm Rx

Abstract

Abstract Background Implantable cardiac monitor (ICM) has been used to detect occult cardiac arrhythmias in a variety of clinical situations. The reliability and accuracy of diagnosing cardiac arrhythmia could impact patient care. However, it is not clear the efficiency and reliability of detecting cardiac arrhythmias between currently available two ICMs. Purpose To compare the efficiency and accuracy of diagnosing cardiac arrhythmias between Reveal LINQ™ and Confirm Rx™. Methods In the prospective multicenter randomized study, a total of 80 patients (age: 61±17 years, men: 41 [51%]) with cryptogenic stroke (n=52) or recurrent unexplained episodes of palpitations (n=3) or syncope (n=25) received an ICM for detection of arrhythmias. Patients were randomized one to one into receiving either Reveal LINQ™ or Confirm Rx™. Arrhythmic events are defined as pauses of ≥3.0 seconds, bradycardia with heart rate of ≤40 bpm, tachycardia with heart rate of ≥150 bpm, and atrial fibrillation (AF). The time of detected arrhythmic events and patient activated events to the time of data availability on respective website site for analysis were compared between the two ICM. All of the arrhythmia events were adjudicated by two cardiologists to assess the accuracy. Results A total of 680 arrhythmic events including 352 AFs, 109 bradycardias, 167 tachycardias and 52 pauses were transmitted over a mean follow-up of 4.2±2.5 months. The mean time to data transmission was significantly faster in Confirm Rx™ than in Reveal LINQ™ (24±103 vs 475±426 min, P<0.0001). One-hundred fifty-nine of 352 AF events (45%) and 237 of 328 non-AF events (72%) were accurately detected by ICMs (P<0.0001). Twenty-eight of 51 AF events (55%) were detected accurately by Reveal LINQ™ and 131 of 301 AF events (44%) were accurately detected by Confirm Rx™ (P=0.13). The reason for inaccurate AF detection was frequent ectopic beats (n=134), P-wave oversensing (n=15), T-wave oversensing (n=8), both P- and T-waves oversensing (n=18) and sinus tachycardia with R-wave undersensing (n=18). Conclusions Data transmission of arrhythmic events using Confirm Rx™ is significantly faster than Reveal LINQ™. The accuracy of AF detection with current algorithms in both ICM remains suboptimal. More patients and longer follow-up is required to confirm these findings.