9. Pacemakers: Clinical Results, Practical Aspects and Late-Break Clinical Trials

CRT requires persistent myocardial capture in LV to assure hemodynamic benefit. The feasibility of an evoked response (VER) morphology based algorithm for capture verification in both ventricles was investigated. Right and left VERs were recorded from 30 pts (24 m, 72±7 y) during implant of a biventricular (BiV) device. Leads of different companies were tested. ECG and VER were recorded for 10 s in each pacing conditions: RV, LV, and BiV with different inter-ventricular delays. VERs were classified offline for capture and non-capture by the morphology based algorithm and classification was confirmed or rejected manually, based on ECGs. A total of 4548 LV and 4450 RV paced events were collected. In RV and LV, the sensitivity of the algorithm was 93.0% and 96.8% and the specificity was 88.1% and 95.0%, respectively. The lower sensitivity in RV resulted from signal blanking in both channels during BiV pacing with inter-ventricular delay ≠ 0 ms. Conclusions The signal morphology based algorithm for RV and LV capture classification showed to be safe and effective with all leads tested in the study.