P2872Remote management of heart failure patients with the multisensor ICD alert: preliminary results from the Italian pilot experience

Background: The HeartLogic (Boston Scientific) index combines data from multiple implantable cardioverter-defibrillator (ICD)-based sensors and has proved to be a sensitive and timely predictor of impending heart failure (HF) decompensation. Objective: To describe a preliminary experience of remote HF manage- ment of patients who received a HeartLogic-enabled ICD or cardiac resyn- chronization therapy ICD (CRT-D) in clinical practice. Methods: The HeartLogic feature was activated in 101 patients (74 male, 71±10 years, ejection fraction 30±7%). From implantation to activation (blinded phase), the HeartLogic index trend was not available, thus no clin- ical actions were taken in response to it. After activation (active phase), remote data reviews and patient phone contacts were performed monthly and at the time of HeartLogic alerts (when the index crossed the nominal alert threshold value of 16), to assess the patient decompensation status. In-office visits were performed when deemed necessary. Results: During the blinded phase, the HeartLogic index crossed the threshold value 24 times (over 24 person-years, 0.99 alerts/pt-year) in 16 patients. HeartLogic alerts preceded all hospitalizations and unplanned in- office visits for HF (sensitivity: 100%, median early warning: 38 days for hospitalizations, 12 days for HF visits). No clinical events were detected during or within 30 days of recovery of 10 HeartLogic alerts (unexplained alert rate: 0.41 per patient-year). Thus, the positive predictive value was 58% (14/24). During the active phase, 44 HeartLogic alerts were reported (over 46 person-years, 0.95 alerts/pt-year) in 30 patients. 26 (59%) Heart- Logic alerts were judged clinically meaningful (i.e. associated with worsen- ing of HF and/or influenced the clinician’s decision to make changes to the subject’s management). Conclusions: In this first description of the use of HeartLogic in clinical practice, the algorithm demonstrated its ability to detect gradual worsening of HF. The results of the blinded phase of our experience favorably com- pare with those reported in the validation study. In the active phase, the HeartLogic index provided clinically meaningful information for the remote management of HF patients.