Indian Pacing and Electrophysiology Journal | 2019
Reducing inappropriate therapy in defibrillators-can we count on mathematical models?
Abstract
Inappropriate arrhythmia detection is a vexing clinical issue which has detrimental clinical effects. Tackling this is an important aspect in clinical electrophysiology, and innovations in device programming have been proved useful. A risk predictive model to identify those who are more likely to have inappropriate arrhythmia detection by the device would be a useful innovation that will permit tailored programming of the device and protect from the inherent drawbacks of inappropriate arrhythmia therapy. In the study published in this issue of the journal, Lebedeva et al. have designed a predictive logistic model for detecting episodes of true ventricular tachycardia (VT) in patients who had an implantable cardioverter-defibrillator (ICD). Of the 83 patients who underwent an ICD implantation for primary prevention indication, the majority of the devices (65) had an atrial lead (45 CRT-D devices, and 20 dual chamber ICDs), and all the patients had a uniform 2zone programming for VT detection. A change from this standard programming was reserved for those with an inappropriate VT detection. Apart from the standard discrimination algorithm used by the device, the electrograms were also reviewed by two electrophysiologists and a device support specialist. Arguably, this methodology has imparted a uniformity of device programming and also improved the diagnostic accuracy of the arrhythmia episode. The use of a decision tree model using the relevant parameters derived by logistic regression analysis predictive of inappropriate ICD therapy makes the study interesting. During a mean 10.75 (2.3e24.5) month follow up period, there were 256 episodes of non-sustained ventricular tachycardia (NSVT), and 114 episodes of VT/Ventricular fibrillation (VF) detected by the ICDs. Of more concern, there were 181 episodes of inappropriate shocks and 22 (out of 108) episodes of inappropriate antitachycardia pacing (ATP). These figures are comparable to other studies with similar follow up, like theMADIT II trial which reported an incidence of 31.2% [1]. Further, the authors observed