Programming optimal atrioventricular delay in dual chamber pacing using peak endocardial acceleration: comparison with a standard echocardiographic procedure.

Optimization of programmed atrioventricular delay in dual chamber pacing is essential to the hemodynamic efficiency of the heart. Automatic AV delay optimization in an implanted pacemaker is highly desirable. Variations of peak endocardial acceleration (PEA) with AV delay at rest correlate well with echocardiography derived observations, particularly with end-diastolic filling and mitral valve closure timings. This suggests the possibility of devicing a procedure for the automatic determination of the optimal AV delay. The aim of this study was to compare a proposed algorithm for optimal AV delay determination with an accepted echocardiographic method. Fifteen patients with high degree AV block received BEST-Living pacing systems. Automatic AV delay scans were performed at rest (60-300 ms in 20-ms steps with 60 beats per step) in DDD at 90 ppm, while simultaneously recording cycle-by-cycle PEA values, which were averaged for each AV delay to obtain a PEA versus AV delay curve. Nonlinear regression analysis based on a Boltzmann sigmoid curve was performed, and the optimal AV delay (OAVD) was chosen as the sigmoid inflection point of the regression curve. The OAVD was also evaluated for each patient using the Ritter echocardiographic method. Good sigmoid fit was obtained in 13 of 15 patients. The mean OAVD obtained by the PEA sigmoid algorithm was 146.9 +/- 32.1 ms, and the corresponding result obtained by echocardiography was 156.4 +/- 34.3 ms (range 31.8-39.7 ms). Correlation analysis yielded r = 0.79, P = 0.0012. In conclusion, OAVD estimates obtained by PEA analysis during automatic AV delay scanning are consistent with those obtained by echocardiography. The proposed algorithm can be used for automatic OAVD determination in an implanted pacemaker pulse generator.