Get access

Acute Clinical Evaluation of a Left Ventricular Automatic Threshold Determination Algorithm Based on Evoked Response Sensing


  • Disclosures: All authors have disclosed relationships with device and/or pharaceutical companies, and these are on file and available.

  • J. H. Lobban: None.

  • S. Shome and C. Delaney: Boston Scientific Corp. Employees.

Kenneth A. Ellenbogen, M.D., Virginia Commonwealth University, Division of Cardiology, PO Box 980053, Richmond, VA 23298. Fax: 651-582-2872; e-mail:


Introduction:Automatic pacing threshold (AT) testing may simplify device follow-up and improve device longevity. This study's objective was to evaluate the performance of a left ventricular (LV) evoked response sensing-based AT algorithm, for cardiac resynchronization therapy (CRT) devices.

Methods:Patients scheduled for CRT-D/P implant were enrolled. A manual step-down threshold test and a Left Ventricular Automatic Threshold (LVAT) test in each of four pacing vectors—LVTip→Can, LVTip→right ventricle (RV),= LVRing→Can, and LVRing→RV—were conducted. Patients were randomized to either 0.4-ms or 1.0-ms pacing pulse width and in the manual and LVAT test order. A blinded core lab electrophysiologist (EP) determined the threshold using the surface electrocardiogram (gold standard).

Results:Data from 70 patients were analyzed. Bipolar LV leads from three major manufacturers were used. A total of 273 AT tests were performed; 12 AT tests did not result in a threshold due to improper testing conditions. Of 261 eligible tests, 234 AT tests (89.6%) returned a threshold measurement. Of the 234 tests, in 233 tests (99.5%) the algorithm-determined threshold matched the EP-determined threshold for that test. A total of 16,689 capture and 526 noncapture beats were collected and the accuracy for detecting capture and noncapture were 98.5% and 99.7% with a two-sided 95% confidence level of (98.4%, 98.7%) and (99.4%, 100%), respectively. No AT threshold measurement was lower than the EP-determined threshold.

Conclusion:In this study, the results suggest that the LVAT algorithm is accurate at determining pacing thresholds in multiple pacing configurations and a wide range of LV leads in CRT-D/P patients. (PACE 2011;1–5)