Resumen:
Introduction and objectives: HeartLogic is a multiparametric algorithm incorporated into implantable cardioverter-defibrillators (LCD). The associated alerts predict impending heart failure (HF) decompensations. Our objective was to analyze the association between alerts and clinical events and to describe the implementation of a protocol for remote management in a multicenter registry. Methods: We evaluated study phase 1 (the investigators were blinded to the alert state) and phases 2 and 3 (after HeartLogic activation, managed as per local practice and with a standardized protocol, respectively). Results: We included 288 patients from 15 centers. In phase 1, the median observation period was 10 months and there were 73 alerts (0.72 alerts/patient-y), with 8 hospitalizations and 2 emergency room admissions for HF (0.10 events/patient-y). There were no HF hospitalizations outside the alert period. In the active phases, the median follow-up was 16 (95%CI, 15-22) months and there were 277 alerts (0.89 alerts/patient-y); 33 were associated with HF hospitalizations or HF death (n = 6), 46 with minor decompensations, and 78 with other events. The unexplained alert rate was 0.39 alerts/patient-y. Outside the alert state, there was only 1 HF hospitalization and 1 minor HF decompensation. Most alerts (82% in phase 2 and 81% in phase 3; P = .861) were remotely managed. The median NT-proBNP value was higher within than outside the alert state (7378 vs 1210 pg/mL; P < .001). Conclusions: The HeartLogic index was frequently associated with HF-related events and other clinically relevant situations, with a low rate of unexplained events. A standardized protocol allowed alerts to be safely and remotely detected and appropriate action to be taken on them.