Automatic remote monitoring of implantable cardioverter-defibrillator lead and generator performance: the Lumos-T Safely RedUceS RouTine Office Device Follow-Up (TRUST) trial.

BACKGROUND Monitoring performance of implantable cardioverter-defibrillator (ICD) generators and leads is important. Methods available are with in-person evaluations or by automatic remote home monitoring (HM). These were prospectively evaluated and compared in the TRUST trial. The HM technology tested performed daily self-checks and databasing with rapid event notifications for out-of-range (including asymptomatic) conditions. METHODS AND RESULTS Patients (n=1339) were randomly assigned after ICD implant 2:1 to HM or to conventional groups. Both groups underwent scheduled checks every 3 months and were followed for 15 months. In HM, in-person office visits were scheduled at 3 and 15 months. At 6, 9, and 12 months, HM only was used with subsequent office visits if necessary. Between these time points, ICDs triggered event notifications for system integrity problems. Patients randomly assigned to conventional follow-up were evaluated with office visits only. HM and conventional patients were similar (age, 63.3±12.8 versus 64.0±12.1 years; 72.0% versus 73.1% male; New York Heart Association II class, 55.9% versus 60.4%; left ventricular ejection fraction, 29.0±10.7% versus 28.5±9.8%; coronary artery disease, 64.8% versus 71.7%; primary prevention, 72.2% versus 73.8%; DDD devices, 57.8% versus 56.6%). Four patients crossed over from conventional to HM because of advisories. Scheduled checks were more successfully accomplished in HM (92.7% versus 89.2% in conventional, P<0.001). Sixty-two device-related events (53 in HM versus 9 in conventional) were observed in 46 patients (40 [4.4%] in HM versus 6 [1.39%] in conventional, P=0.004). Forty-seven percent were asymptomatic. HM detected generator and lead problems earlier (HM versus conventional: median, 1 versus 5 days; P=0.05). A total of 20 device problems (eg, lead fracture, elective replacement indicators) requiring surgical revision (0.012 per patient-year) were found, 15 in HM and 5 in the conventional groups. Other events were managed nonsurgically (eg, reprogramming, initiation of antiarrhythmics). CONCLUSIONS ICD lead and generator malfunction was infrequent and often asymptomatic. Only a minority of detected events required surgical intervention. Automatic HM enhanced discovery, permitted prompt detection, and facilitated management decisions. Longitudinal parameter trending, with component function evaluated daily by remote monitoring, may enable long-term performance assessment. CLINICAL TRIAL REGISTRATION URL: http://clinicaltrials.gov. Unique identifier: NCT00336284.