Work

Optimized Strategy for Improving the Accuracy of Determining Left Bundle Branch Capture

Year: 2026
Type: article
Study design: Observational Study
Abstract: Background The accuracy of current ECG criteria for left bundle branch (LBB) pacing (LBBP) remains inadequate. Objectives To establish an optimized strategy to distinguish LBBP from left ventricular septal pacing (LVSP) in patients with a narrow QRS complex. Methods LBB po ‐V 6 RWPT, Native‐V 6 RWPT, and Paced‐V 6 RWPT were measured from LBB potential (LBB po ) onset, intrinsic QRS onset, and stimulus end to the R‐wave peak in V 6 , respectively. ΔV 6 RWPT and ΔLBB po ‐V 6 RWPT were defined as Paced‐V 6 RWPT minus Native‐V 6 RWPT and Paced‐V 6 RWPT minus LBB po ‐V 6 RWPT. Receiver operating characteristic analysis identified optimal cut‐offs for these two differences, Paced‐V 6 RWPT, and the V 6 ‐V 1 interpeak interval in patients with transition‐confirmed capture. Areas under the curves (AUCs) were compared by LBB po status. Results LBBP was confirmed in 292 (61.6%) of 474 patients with LBBAP. The optimal cut‐off values of ΔLBB po ‐V 6 RWPT, ΔV 6 RWPT, Paced‐V 6 RWPT, and the V 6 ‐V 1 interpeak interval for determining LBBP were 6 ms, 30 ms, 67 ms, and 31 ms, respectively. In 211 of 292 patients, LBB po was recorded and the ΔLBB po ‐V 6 RWPT criterion achieved the largest AUC (1.000), with 98.5% sensitivity and 99.5% specificity. LBBP was misclassified as LVSP by the ΔV 6 RWPT criterion in 3 patients but was correctly identified by the ΔLBB po ‐V 6 RWPT criterion. In 81 of 292 patients, LBB po was not recorded and the ΔV 6 RWPT criterion showed the largest AUC (0.990), with 98.8% sensitivity and 97.3% specificity. Conclusions Selecting ΔLBB po ‐V 6 RWPT for patients with LBB po and ΔV 6 RWPT for those without LBB po may help classify LBB capture in unclassified LBBAP with narrow QRS complexes. Diagnostic accuracy in this population requires prospective validation.
Language: English
Field: Medicine
Open Access status: bronze
OpenAlex ID: W7237194635
FWCI:
Cites: 7
Cited by:
Related to: 10