Comparison of Seven Algorithms to Localize the Accessory Pathway in Patients with Wolff-Parkinson-White Syndrome

The purpose of thi s study was to compare the sensitivity, (+) predictive value (PPY), inter and intraobserver variability of published 7 a lgori thıns (Skeberis et al, Xie et al., Chiang et al. , Fitzpatrick et al. , Davilla et al., Adalet et al., and Arruda et al.) which were developed to localize atrioventricular accessoı·y pathways (AP's) in patients with the Wolff- Parkinson- White (WPW) syndrome. One h undred consecutive patients, diag nosed to have WPW syndrome from 1 2-lead electrocardiogram (ECG), who underwent successful radiofrequency catheter ablation of a single overt AP, were included in the study. The 12-lead ECGs were evaluated for two of each in diffe rent times by three independent and experienced electrophysiologist. The sensitivity and PPY was found to vary between 29-51% and 3 1-60%, respecti vely. Chiang's algorithm was shown to have the highest sensitivity and PPV amongst the 7 a l gorith ıns (5 ı % and 60%, respectively). In general, for localizing the APs, best results were obtained in left lateral APs (mean 70%). Adalet's algorithm had the highest sensit ivity to differentiate the right and left APs (75% and 77%, respectively). The PPY tended to be lower when delta-wave polarity was not included in the algorithm's architecture (34% and %43.3, respectively, p=0.045). The strongest correlation ininter and intraobserver variation was found in Chiang's algorithm (kappa=0.78-0.82 and kappa= 0.76, [%95 CI: 0.73-0.79], respectively), nevertheless Fitzpatrics's algorithm exhibited the weakest intra and interobserver agreement (kappa=0.56-0.72 and kappa= 0.40, [%95 CI: 0.36-0.44], respectively) . In conclusion, the sensitivity and PPY of the included a l g orithıns were clearly lower than those repor ted by the corresponding authors. Left lateral APs were the only pathways that could easily be recognized by all algorithms. The algorithms that included the delta-wave polarity was shown to increase PPY s. In o ur series, consisting of 100 patients with the WPW syndrome, Chiang's algorithın had the highest PPV and inter and intraobserver agreeınent aınong st all a lgorithıns. These findings should be considered when using these a lgorithıns in elinical settings or when building new ones.