Abstract
INTRODUCTION: Implantable defibrillator devices (ICDs) can be used for remote monitoring of different variables, including some related to Heart Failure (HF). Two different algorithms (TriageHF and HeartLogic) arise by combining some of these variables to generate an estimation of HF decompensation risk in the following days. Until now, no other trial has evaluated both algorithms in a head-to-head comparison. The primary objective is to compare diagnostic accuracy of both algorithms in a similar cohort of patients. MATERIAL AND METHODS: Descriptive monocentric cohort study of a series of 64 patients who have been implanted with a Medtronic or Boston Scientific ICD with the TriageHF or Heart Logic algorithm available during the period between January 2020 and June 2022, with a total of 27 patients in the HeartLogic group and 37 patients in the TriageHF group. RESULTS: During the period of the study there were a total of 1142 alarms analyzed. There were no differences in the basal characteristics of both groups. We reported a risk alarm-patient ratio of 1.31 ± 1.89 in the HeartLogic group and of 3.32 ± 3.08 in the TriageHF group (p < 0.01). In the TriageHF group, we reported a lower specificity with (0.76), with higher sensitivity (0.97) and PPV (0.18), and similar NPV (1). Survival analysis shows no statistical differences between both algorithms in the 30 days following the alert. CONCLUSIONS: TriageHF algorithm had higher sensibility and PPV, leading to a higher number of alerts/patients, while HeartLogic algorithm had a better specificity. These differences should be considered to optimize patient follow-ups in home monitoring.