Morphology Discrimination: A Beat-to-Beat Algorithm for the Discrimination of Ventricular from Supraventricular Tachycardia by Implantable Cardioverter Defibrillators

2001 
GRONEFELD, G.C., et al.: Morphology Discrimination: A Beat-to-Beat Algorithm for the Discrimination of Ventricular From Supraventricular Tachycardia by Implantable Cardioverter Defibrillators. Inappropriate therapy of SVTs by ICDs remains a major clinical problem despite enhanced detection criteria like “sudden onset” and “rate stability” in third-generation devices. Electrogram morphology discrimination offers an additional approach to improve discrimination of supraventricular tachycardia (SVT) from ventricular tachycardia (VT). In a prospective, multicenter study, patients received an ICD with a beat-to-beat algorithm for morphological analysis of the intracardiac electrogram (Morphology Discrimination, MD). A nominal programming for standard enhancement criteria and morphology discrimination was required at implant. Electrogram storage of tachycardia episodes irrespective of delivery of therapy was used to assess sensitivity and specificity of the morphology algorithm alone and in combination with established detection criteria. During a 12 6 6-month follow-up, 886 episodes of device stored electrograms from 82 of 256 patients were evaluated. At nominal settings, the MD algorithm correctly identified 423 of 551 episodes as VT resulting in sensitivity of 77%. The classification of SVT was met in 239 of 335 episodes resulting in specificity of 71%. In combination with sudden onset, sensitivity increased to 99.5% at the expense of specificity (48%). In conclusion, SVT-VT discrimination based on morphological analysis alone results in limited sensitivity and specificity. Programming the monitor mode allows individual assessment of the performance of this detection enhancement feature during clinical followup without compromising device safety. Only in patients with documented efficacy of morphology discrimination should this feature be subsequently activated.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    17
    References
    28
    Citations
    NaN
    KQI
    []