New algorithm for the localization of accessory atrioventricular connections using a baseline electrocardiogram.

OBJECTIVES In this study, we propose a new algorithm for accessory atrioventricular pathway localization using a 12-lead electrocardiogram (ECG). BACKGROUND Radiofrequency catheter ablation produces a very discrete lesion, and ECG localization based on surgical dissection is obsolete. METHODS Stepwise discriminant analysis was used to assess the relation of 18 pre-excited ECG (QRS duration > 100 ms) variables to the site of successful ablation in 93 patients. The most discriminating variables were combined to form rules for each location. The ECGs were retested by these rules to determine predictive accuracy. RESULTS If the precordial QRS transition was at or before lead V1, the pathway had been ablated on the left side. If it was after lead V2, the pathway had been ablated on the right side. If the QRS transition was between leads V1 and V2 or at lead V2, then if the R wave amplitude in lead I was greater than the S wave by > or = 1.0 mV, it was right-sided; otherwise, it was left-sided (p < 0.0001, sensitivity 100%, specificity 97%). Right-side pathways. If the QRS transition was between leads V2 and V3, the pathway was right septal; if after lead V4, it was right lateral. If it was between leads V3 and V4, then if the delta wave amplitude in lead II was > or = 1.0 mV, it was right septal; otherwise, it was right lateral (p < 0.0001, sensitivity 97%, specificity 95%). In right lateral locations, if the delta wave frontal axis was > or = 0 degrees, or if it was < 0 degrees but the R wave amplitude in lead III was > or = 0 mV, it was anterolateral; otherwise, it was posterolateral (p < 0.0001, sensitivity 100%, specificity 87.3%). Anteroseptal pathways had a sum of delta wave polarities in leads II, III and aVF > or = +2(p < 0.0001, sensitivity 100%, specificity 100%). Posteroseptal pathways (inferior delta wave sum < or = -2) were less well discriminated from right midseptal pathways (inferior delta wave sum < or = 1 > or = -1) (p < 0.0001, sensitivity 76.5%, specificity 71%) [corrected]. Left-sided pathways. Two or more positive delta waves in the inferior leads or the presence of an S wave amplitude in lead aVL greater than the R wave, or both, discriminated left anterolateral pathways from posterior pathways (p < 0.001, sensitivity and specificity 100%). If the R wave in lead I was greater than the S wave by > or = 0.8 mV, and the sum of inferior delta wave polarities was negative, the location was posteroseptal; otherwise, it was posterolateral (p < 0.05, sensitivity 71.4%, specificity 100%). CONCLUSIONS Using the algorithm derived, a right-sided accessory pathway can be reliably distinguished from one that is left-sided, right free wall from right septal, right anterolateral from posterolateral and anteroseptal from other right septal pathways. Left anterolateral pathways can be distinguished from left posterior pathways and left posterolateral pathways from left posteroseptal pathways.