QRS complex shortening by cardiac resynchronization therapy (CRT) has been associated with improved outcomes.

We hypothesized that the absence of QRS duration (QRSd) prolongation by right ventricular mid-septal pacing (RVP) may indicate a complete left bundle branch block (cLBBB).

We prospectively collected 12-lead surface ECG and intracardiac electrograms during CRT implant procedures. Digital recordings were edited and manually measured. The outcome measure was a change in QRSd induced by CRT (delta CRT). Several outcome predictors were investigated: native QRSd, cLBBB (by Strauss), Q-LV interval, and a newly proposed index defined by the difference between RVP and native QRSd (delta RVP).

133 consecutive patients were included in the study. Delta RVP was 27 ± 25 ms and delta CRT was -14 ± 28 ms. Delta CRT correlated with native QRSd (r = -0.65), with the presence of ECG-based cLBBB (-0.40), with Q-LV (r = -0.68), and with delta RVP (r = 0.72), P <0.00001 for all correlations. On multivariable analysis, delta CRT was most strongly associated with delta RVP (P <0.00001), followed by native QRSd and Q-LV, while the ECG-based cLBBB became a non-significant factor.

Baseline QRSd, delta RVP, and LV electrical lead position (Q-LV) represent strong independent predictors of electrocardiographic response to CRT. The absence of QRSd prolongation by RVP may serve as an alternative and more specific marker of the cLBBB. Delta RVP correlates tightly with the CRT effect on QRSd and outperforms the predictive value of ECG-based cLBBB.