Premium
Performance of Basic Ventricular Tachycardia Detection Algorithms in Implantable Cardioverter Defibrillators; Implications for Device Programming
Author(s) -
ANDERSON MARK H.,
MURGATROYD FRANCIS D.,
HNATKOVA KATERINA,
XIE BAIYAN,
JONES SUSAN,
ROWLAND EDWARD,
WARD DAVID E.,
CAMM A. JOHN,
MALIK MAREK
Publication year - 1997
Publication title -
pacing and clinical electrophysiology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.686
H-Index - 101
eISSN - 1540-8159
pISSN - 0147-8389
DOI - 10.1111/j.1540-8159.1997.tb05469.x
Subject(s) - medicine , implantable cardioverter defibrillator , ventricular tachycardia , cardiology , tachycardia , algorithm , computer science
Around 20% of patients with third generation implantable cardioverter defibrillators receive inappropriate therapy, usually triggered by atrial fibrillation. This is because the criteria used for ventricular tachycardia detection by current implantable cardioverter defibrillators are based on the analysis of a sequence of RR intervals and may be inappropriately satisfied by supraventricular tachyarrhythmias. Algorithms for ventricular tachycardia detection were challenged against the full RR interval sequences from 482 spontaneous episodes of atrial fibrillation and 260 spontaneous episodes of ventricular tachycardia to determine their ability to discriminate between the arrhythmias. The sensitivities and specificities of the algorithms were calculated over a wide range of programmable parameters. For a given window length and detection interval, the most stringent algorithms, that required all beats to be classified as “fast”, were more specific than those allowing a proportion of “normal” intervals, even after adjustment for differing sensitivity. These differences were less marked for faster tachycardias. Specificity increased with the detection window length to a limit of approximately 18 beats. We conclude that ventricular tachycardia is detected with the highest specificity if all beats in an analyzed sequence are required to be “fast” even after lengthening of the tachycardia detection interval to maintain sensitivity. Further improvement in algorithm performance may require the incorporation of criteria such as tachycardia onset and stability.