Electrocardiographic Predictors of Concealed Left-Sided Accessory Pathways in Patients with Narrow-Complex Tachycardia: Development and Internal Validation of the CLAP Score

Objectives: Aim of this study was to identify electrocardiographic predictors of electrophysiological study (EPS)-confirmed concealed left-sided accessory pathways in patients with narrow-complex tachycardia and to develop and internally validate an ECG-based prediction score. Methodology: This cross-sectional diagnostic accuracy and prediction-model development study included 290 consecutive adults undergoing EPS for narrow-complex tachycardia. Twelve-lead ECGs recorded during tachycardia were independently evaluated by two electrophysiologists blinded to EPS findings. Candidate ECG predictors included RP interval, pseudo-R wave in V1, pseudo-S waves in the inferior leads, ST-segment depression, P-wave morphology, and heart rate. The primary outcome was an EPS-confirmed concealed left-sided accessory pathway. Results: Among 290 patients, 238 (82.1%) had atrioventricular nodal reentrant tachycardia (AVNRT), 29 (10.0%) had left-sided atrioventricular reentrant tachycardia (AVRT), 12 (4.1%) had right-sided AVRT, and 11 (3.8%) had non-inducible tachycardia. Absence of a pseudo-R wave in V1 was independently associated with a concealed left-sided accessory pathway (adjusted OR 16.14, 95% CI 1.95–133.27; p=0.010), as was an intermediate RP interval (>90 ms but RP < PR; adjusted OR 13.72, 95% CI 4.56–41.28; p<0.001). ST-segment depression and heart rate >200 beats/minute were retained in the exploratory CLAP score but did not reach statistical significance in the multivariable model. The resulting 0–6-point CLAP score assigned 2 points each for intermediate RP interval and absence of pseudo-R wave in V1 and 1 point each for ST-segment depression and heart rate >200 beats/minute. At a cutoff of ≥4, the score demonstrated 79.31% sensitivity, 88.51% specificity, 43.40% positive predictive value, and 97.47% negative predictive value. Discrimination across the complete score range was high (AUC 0.938, 95% CI 0.902–0.974), with an optimism-corrected AUC of 0.938, calibration intercept of 0.012, and calibration slope of 0.983 after bootstrap internal validation. Conclusion: Selected ECG characteristics recorded during narrow-complex tachycardia were associated with EPS-confirmed concealed left-sided accessory pathways. The preliminary CLAP score demonstrated promising discrimination and internal calibration in this single-center cohort.