Comparison of the Accuracy of Three Algorithms in Predicting Accessory Pathways among Adult Wolff-Parkinson-White (WPW) Syndrome Patients in our Population

Abstract

Background: after the introduction of catheter ablation of atrioventricular accessory pathways (APs) in Wolff-Parkinson-White (WPW) syndrome, to facilitate planned ablation and minimize catheter-related injury, predicting the AP was required. Localization of APs can be anywhere around the atrioventricular (AV) annuli, left- or right-sided, or within the septum or a rare possibility of Mahaim fibers. Objective: to compare between three algorithms in their accuracy in predicting the exact site of accessory pathway. Patients and Methods: one hundred patients with manifest pre-excitation who underwent electrophysiological study (EPS) and successful ablation of accessory pathway who were subjected to history taking, complete physical examination stressing on 12-lead of pre-excited electrocardiogram (ECG). Results: the distributions of the accessory pathway with left lateral accessory pathway being the most prevalent (27%) while the postero-lateral being the least prevalent one (2%) and that Right-sided accessory pathway was most frequently predicted by Arruda algorithm in 31% of patients and least frequently predicted by D’Avila in 17% of patients while its actually present in 23% of patients according to the EP study, Left-sided accessory pathway was most frequently predicted by D’Avila (61%) and least frequently predicted by Arruda (49%) whereas it was truly present in 51% of patients according to EP study. Mid and antero-septal accessory pathways were most frequently predicted by Chiang and least frequently predicted by Arruda while it was present in 26% of cases proved by EP study. In all algorithms, 72 % of predictions were correct for Chiang, 81 % for D’Avila, and 71 % for Arruda and the percentage of predictive accuracy of all algorithms did not differ between the algorithms (p=1.000; p=0.076; p=0.064, respectively) The best algorithm for prediction of right-sided and left-sided accessory pathways was D’Avila (p<0.001). The best algorithm that is particularly useful in predicting antero-septal and mid-septal accessory pathways was Chiang (p<0.001). Conclusion: prior knowledge of the AP location allows better planning, faster and safer procedure, as well as decreased exposure to ionizing radiation and unnecessary punctures. In all algorithms, 72% of predictions were correct for Chiang, 81% for D'Avila, and 71% for Arruda and the percentage of predictive accuracy of all algorithms did not differ between the algorithms (p=1.000; p=0.076; p=0.064, respectively). The best algorithm for prediction of right-sided and left-sided accessory pathways was D'Avila (p<0.001). The best algorithm that is particularly useful in predicting anteroseptal and mid-septal accessory pathways was Chiang (p<0.001).