A-V=1 1 CARDIAC-ARRHYTHMIA DETECTION BY VA INTERVAL-ANALYSIS/

Dual-chamber-sensing implantable-cardioverter defibrillators are soon expected to replace ventricular sensing devices. The addition of an at rial sensing lead will dramatically improve the specificity of arrhyth mia detection. Even when using combined ventricular and atrial rate cr iteria, ambiguity in the case of atrial tachycardia with: anterograde conduction versus ventricular tachycardia with: retrograde conduction still remains. The introduction of dual-chamber sensing in antitachyca rdia devices allows for additional features, such as the measurement o f atrioventricular (AV) and ventriculoatrial (VA) intervals. This stud y investigated relationships between AV and VA intervals to address pr oblems arising in tachycardias with confounding 1:1 relationships. Thi rty-one passages of 1:1 anterograde conduction from nine patients duri ng atrial pacing at cycle lengths of 600-300 ms and 24 passages of 1:1 retrograde conduction from eight patients during ventricular pacing a t cycle lengths of 600-300 ms were analyzed. Moving averages of three successive VA interval measurements were used to develop a criterion t o be implemented into an algorithm to reduce ambiguity. Five randomly selected ventricular pacing passages were used as a training set. Uppe r and lower VA interval boundaries (234 ms and 132 ms) determined from the training set were used to classify 1:1 retrograde activation. To account for premature beats and outliers, the boundary criterion requi red 9 of 12 of the most recent moving averages to fall within the uppe r and lower limits. Of the 19 analyzed passages of ventricular pacing, 18 (95%) were correctly classified using the VA interval as an added feature. Of the 31 atrial pacing passages, 24 (77%) were correctly cla ssified. Using only atrial or ventricular rates, all 1:1 tachycardias in this patient sample would be classified as ventricular tachycardia, resulting in false shocks. Specificity of diagnosis in ambiguous 1:1 tachycardias can be increased using VA interval measurements at the co st of minimum loss in sensitivity for ventricular tachycardia detectio n. This algorithm imposes little in additional computation for dual-ch amber-sensing implantable-cardioverter defibrillators and greatly redu ces the possibility of false shocks in 1:1 supraventricular tachycardi as.