Implantable cardioverter-defibrillators (ICD) are a standard therapy to prevent sudden cardiac death (SCD). Remote monitoring (RM) of ICD patients provides healthcare quality improvement and resource savings compared with standard in-hospital visits. Only limited data exist about RM organizations.

We aimed to evaluate and validate our institutional optimized RM organization model for ICD patients.

This observational study compared two RM models with an iterative qualitative and quantitative approach in 562 ICD patients: RM1 with device diagnostics evaluation by nurses and cardiologists, and RM2 with a selected approach with decision trees for actions. The main endpoints were in-hospital professional actions and times related to RM alerts.

During RM1, 1134 alerts occurred in 427 patients (286 patient-years) of which 376 (33%) were submitted to cardiologist review whereas during RM2, 1522 alerts occurred in 562 patients (458 patient-years) of which 273 (18%) were submitted to cardiologist review (P<0.001). An active decision was triggered by 73/376 (19.4%) alerts in RM1 versus 77/273 (28.2%) in RM2 (P=0.009). Mean time to manage a RM alert was 4minutes 31seconds for RM1 versus 2minutes 10seconds for RM2 (P<0.001). Comparing RM1 vs. RM2, annual rates of alert-related hospitalizations were 9.4 and 7.9 per 100 patient-years (P=0.50), while annual rates of alert-related in-hospital visits were 9.4 and 6.1 per 100 patient-years (P=0.11).

In ICD patients, optimized RM strategy based on automatic alerts and decision trees allows to focus on clinically relevant events and to reduce healthcare resources without compromising quality.