Novel Nonlinear and RR-based Methods for Inappropriate ICD Therapy Reduction

Abstract

Implantable cardioverter-defibrillators (ICD) are a commonly implanted device used to deliver arrythmia terminating therapy to the heart in the presence of life-threatening arrythmias such as ventricular fibrillation and ventricular tachycardia. However, not all therapies, shocks or anti-tachycardia pacing (ATP), are appropriate. In many instances, therapy is delivered when the heart is in an abnormal, but non-life-threatening arrythmia such as atrial fibrillation. While major medical device companies have devised numerous strategies to eliminate these cases of inappropriate therapy, there remains room for improvement. Thus, the goal of this thesis is to describe the development of novel strategies to discriminate between appropriate and inappropriate ICD therapy. The strategies use nonlinear measures and RR interval measures fed into principal component analysis or linear combination scores to discriminate. The final method using linear combination scores showed near 100% discrimination between appropriate and inappropriate therapy events retrospectively and 100% discrimination in a pseudo real time study. Thus, this strategy shows immense promise for use in a future large clinical study to eliminate inappropriate therapy.