Remote Monitoring Alerts for Ventricular Arrhythmias in Implantable Cardioverter Defibrillator (ICD) Patients: A Retrospective Observational Cohort Study

Abstract

Background: Remote monitoring is the standard of care for implanted cardioverter defibrillator (ICD) patients. The early notification of actionable alert events via remote monitoring is associated with improved patient outcomes. It reduces in-clinic visits for patients and increases adherence to follow-up without compromising safety. Remote monitoring has resulted in large volumes of unscheduled alert transmissions that require daily triage, review, and documentation, with most transmissions occurring for non-critical arrhythmias. There are significant challenges for clinics in safely managing this unplanned workload. Questions remain on the clinical utility of the early notification of some ventricular arrhythmias, and how the differences between cardiac implantable device manufacturers influence the timing and frequency of alerts and impact workload.

Aim: The primary aim of this retrospective review is to investigate the extent of differences in the notification times, transmission type and frequency of ventricular arrhythmias between device manufacturers. The secondary aim is to document ventricular arrhythmia alert burden, its clinical relevance, and its impact on workload.

Method: The frequency of transmissions for shocks, anti-tachycardia pacing, and non-sustained VT were retrospectively reviewed in 557 remotely monitored ICD patients followed between January and December 2021. Manufacturer, ventricular arrhythmia alert type, notification time, and action taken by the clinic in response to each arrhythmia were analysed. Descriptive statistical analyses were performed using Pearson’s Chi-squared test for categorical variables and ANOVA or Kruskal Wallis for continuous variables.

Results: Of 2666 transmissions, 1263 (47.4%) were unplanned alert transmissions with 540 (42.6%) of incoming alerts for ventricular arrhythmias. Overall ventricular arrhythmia notification times varied by manufacturer; median 1 day (IQR 1 day) (Abbott and Biotronik), 7 days (IQR 46 days) (Boston Scientific), 28 days (IQR 67days) (Medtronic) p<.001. This difference was most significant for non-sustained VT episodes. Shock notification timing did not vary significantly between manufacturers. Ventricular arrhythmia transmissions occur in 24.2% of all remote monitoring transmissions and make up 42.8% of all alerts. Of ventricular arrhythmia alerts, 7% were for shocks, 12.6% for anti-tachycardia pacing, and 80.4% for non-sustained VT. Alert frequency for ventricular arrhythmias was significantly different between device manufacturers: Abbott 40.6%, Biotronik 47.0%, Boston 9.8%, Medtronic 2.6% (p <.001). No action was required for 78.3% of ventricular arrhythmia alerts. Clinician review was required for 76.9% of shock episodes compared to 5.7% of non-sustained VT alerts. Unscheduled transmissions were more likely to require clinician review than scheduled transmissions (p<0.01). The frequency of transmissions for ventricular arrhythmias was significantly increased in Māori patients (2.48 transmissions per patient per year) compared to European patients (0.85 transmissions per patient per year) (p=.005). Transmission frequency per patient per year varied significantly as a result of device type (single chamber ICD 3.71 +/- 6.99, dual chamber ICD 4.11 +/-5.38, cardiac resynchronisation ICD 7.11 +/-7.66 and subcutaneous ICD 8.30 +/-6.48) (p<0.001).

Conclusion: Unplanned remote monitoring alert transmissions comprise a significant proportion of device clinic workload. The most frequent alerts are for non-sustained VT with the majority of ventricular arrhythmia alerts requiring no action. There is an increased ventricular arrhythmia burden in Māori patients. There is marked variation in alert frequency and notification time based on manufacturer, highlighting challenges for clinics in managing this unplanned workload.