Background: Inappropriate ICD therapy for supraventricular tachycardia (SVT) remains a significant problem. A morphology-based algorithm (Wavelet) compares baseline and tachycardia electrograms (EGM). For this analysis different EGM sources can be programmed. This study evaluates the performance of Wavelet using two different EGM configurations (SVC-Can and RV-Can) for the detection of exercise-induced SVT.
Methods: Patients with a Medtronic model 7230 single chamber ICD and a dual coil lead were included. For each EGM source (SVC-Can or RV-Can), a baseline EGM template was acquired and the morphology similarity to this template (match percentage) was evaluated for 10-15 beats at different heart rates during exercise testing. The lower VT detection limit was programmed to 600 ms (therapies off).
Results: A total of 28 patients (66.9 +/- 4.7 years, 93% men) and 5,824 intracardiac QRS complexes were analyzed. With the RV-Can source, a consistently high similarity to the baseline EGM template was observed (< or =100 bpm: 90.90 +/- 0.56%; >100 bpm: 90.24 +/- 0.55%, P > 0.05). In contrast, SVC-Can was associated with a lower match percentage at baseline and a significant decrease at higher heart rates (< or =100 bpm: 77.91 +/- 2.65%; >100 bpm: 59.05 +/- 5.65%, P < 0.005). Accordingly, the specificity for appropriate detection of exercise-induced SVT was higher with RV-Can (21/21 episodes) than with SVC-Can (8/18 episodes, specificity 100% vs 44%; P < 0.0001).
Conclusion: The RV-Can configuration appears to be superior to SVC-Can as EGM source for appropriate SVT detection with the Wavelet algorithm.