Comparison of Electrocardiogram Characteristics of Two Commercially Available Implantable Loop Monitors: Impact of These Characteristics in the Correct Adjudication of Recorded Events and Minimized Undersensing and Oversensing of Events.

Abstract

Implantable cardiac monitors (ICMs) are useful in the detection of tachycardias, bradycardias, and atrial fibrillation. Undersensing and oversensing of events occur despite complex algorithms. The devices available have subtle differences, which may account for a difference in recorded characteristics. The electrocardiogram (ECG) characteristics of different monitors may influence their ability to correctly identify the events recorded. The objective is to compare the ECG characteristics of two commercially available implantable loop monitors and the ability to improve diagnostic accuracy. Two cohorts of patients were examined. Cohort 1 consisted of 30 patients with a Reveal LINQ I (Medtronic, Minneapolis, MN, USA) implanted, which was replaced with a BIOMONITOR III (Biotronik, Berlin, Germany) when the former had reached end of life. The new monitor was implanted at the same site in all patients. This provided a unique opportunity to compare ECGs obtained by both devices with no other confounding variable. Cohort 2 consisted of patients who had undergone implantation of either device at the discretion of the physician. This was therefore a retrospective analysis of 106 patients who had been implanted with one of the devices within a 2-year period. In both cohorts, we compared R-wave amplitude, the ability to accurately detect P-waves, and the frequency of undersensing and oversensing of events. In cohort 1, the mean R-wave was 0.35 ± 0.2 mV with the LINQ I as compared to 0.98 ± 0.4 with the BIOMONITOR III (P = .001). A P-wave in sinus rhythm was present in 19 (63%) subjects with the LINQ I implants versus 28 (93%) with the BIOMONITOR III implants (P = .004). Undersensing of events was noted in five (16%) patients with the LINQ I versus five (16%) with the BIOMONITOR III (P > .5). Oversensing was seen in 4 patients (13.33%) with the LINQ I versus 0 with the BIOMONITOR III (P = .012). In cohort 2, the mean R-wave with the BIOMONITOR III was significantly greater than with the LINQ I (0.65 ± 0.37 vs. 0.48 ± 0.38; P = .02). A visible P-wave was seen in 33 of 53 patients with the BIOMONITOR III as compared to 16 of 536 patients with the LINQ I monitor (P = .01). The number of patients identified as having under- or oversensing was, however, not significantly different (P = .08) in this cohort. In both patient cohorts, the BIOMONITOR III was noted to have significantly greater R-wave amplitude as compared with the LINQ I. A visible P-wave was also more commonly seen in patients with a BIOMONITOR III. In the paired cohort, but not in the unpaired cohort, the BIOMONITOR III was less likely to oversense R-waves. There was no significant difference in undersensing between the devices.