A novel morphology based algorithm for prediction of antitachycardia pacing success

Purpose: Currently commercialized implantable cardioverter defibrillators (ICDs) can deliver high energy shocks or antitachycardia pacing (ATP) to treat an arrhythmia, based only on the cycle length of the tachycardia. We developed an algorithm to predict the success of an ATP therapy based on the morphology affinity of the QRS complexes of the arrhythmia. The purpose of this study is to test the performance of the algorithm retrospectively. Methods: The algorithm was implemented using signal analysis software. The episodes treated with ATP that were stored in our ICD remote monitoring system were analyzed by the algorithm. Results: Of 84 patients included in the system, 9 had 169 episodes treated with ATP. Episodes with a high algorithm score (i.e. high intra-episode QRSs affinity) were more likely to be terminated with the first sequence of ATP (p=0.002, 44% of episodes if score was 80%). The percentage of shock-terminated episodes was lower for those episodes with higher algorithm scores (p<0.001, 38% if score was <60%, 11% if score was 60%-80%, and 0% if score was 80%. No significance differences were seen in number of ATP sequences delivered between groups of different scores. Conclusion: In a retrospective simulation of the algorithm described, it was able to identify those episodes that are more likely to be terminated with ATPs. This algorithm could be implemented in future devices, allowing the physician to program different therapy strategies based not only on rate, but also on QRSs affinity.