A new and reliable method of individual ventricular capture identification during biventricular pacing threshold testing.

Biventricular (BV) pacing for the treatment of heart failure is in clinical investigation. In the absence of independent outputs for separate pacing of each ventricle, a method is needed to determine the respective LV versus RV thresholds. A technique was developed and validated to distinguish BV capture from LV or RV capture from a multilead surface ECG. The QRS axes were determined at the time of implant by comparing multilead surface ECGs during BV, RV, and LV pacing in 63 patients (42 men, age 63 +/- 12 years) who received pacemakers or ICDs capable of BV pacing. Differences between BV and LV, and between BV and RV axes were examined to determine which ECG leads best indicate a change from BV to univentricular capture. The axis shift from BV to RV pacing was positive while the axis shift from BV to LV pacing was negative. The morphology change associated with LV versus RV capture is best examined in the ECG lead that is perpendicular to the axis shift. A change from BV to LV capture was best identified as increasing positivity of the QRS in lead III, while a change from BV to RV capture was best recognized as increasing positivity of the QRS in lead I. When performing a BV pacing threshold test, mean QRS vector changes derived from standard ECG can be used to distinguish LV or RV capture from BV capture.