Heterogeneity in clinical judgment of septal lead position and capture type in left bundle branch area pacing

BACKGROUND: Determining capture type and septal lead location during left bundle branch area pacing (LBBAP) relies on criteria obtained during implantation. However, during follow-up, the interpretation of left bundle branch (LBB) capture largely depends on QRS morphology, which is not so straightforward in LBBAP. OBJECTIVE: To investigate the inter- and intra-observer agreement, as well as the accuracy of clinical judgment of the ECG in determining LBB-capture and septal lead position in patients undergoing LBBAP implantation. In addition, the role of vectorcardiographic QRS-area in determining LBB-capture was evaluated. METHODS: Unipolar paced ECGs during LBBAP implantation from 50 patients with baseline narrow QRS were collected. LBB-capture was attempted in all patients and assessed using MELOS-criteria and the EHRA consensus-statement. Eight blinded cardiologists classified 100 ECGs for capture type and septal location. RESULTS: The inter-observer and intra-observer agreement for capture type had a Light's kappa of 0.43 and 0.62 respectively. Concordance between clinical judgment and intra-procedural confirmation averaged 72%. Inter-observer and intra-observer agreement for septal lead position had a Light's kappa of 0.43 and 0.77 respectively. QRS-area was significantly higher for LVSP than nsLBBP, while QRS duration was not. A QRS-area cutoff of 26 mV.ms had 77% accuracy in distinguishing LVSP from nsLBBP. Clinical judgment accuracy averaged 72%. CONCLUSION: Inter-observer agreement and correlation with intraprocedural confirmation (gold standard) are only moderate, while intra-observer agreement on ECG-based differentiation of capture type and septal lead location is substantial. Vectorcardiographic QRS-area slightly outperforms clinical judgment in distinguishing capture types and may be a useful objective alternative.