961-81 Time Domain and Spectral Turbulence Analysis of Signal Averaged Electrocardiography as Prognostic Determinants in Idiopathic Dilated Cardiomyopathy

Abstract

Listen

Translate article

Identification of patients with idiopathic dilated cardiomyopathy (IDC) with poor prognosis [sudden death (SD) or progressive heart failure (PHF)] represents a major management problem. The signal averaged electrocardiogram (SAECG) is potentially a non-invasive marker of clinical outcome in IDC. This study compared the ability of time domain (Td) and spectral turbulence analysis (STA) of the SAECG to predict SD/PHF in IDC patients. SAECG were recorded in 58 consecutive patients with IDC (WHO criteria; age 41 ± 14 years) presenting to our hospital and followed over 26 ± 19 months. During the follow up 17 patients had SD/PHF and 41 remained symptomatically and echocardiographically stable. In this population, 27 patients with left bundle branch block or who took anti-arrhythmic drugs had been excluded. Conventiona Td (40 Hz, Butterworth filter) and STA analysis were performed using Del Mar 183 software. There were significant differences in all STA parameters (low slice correlation ratio, interslice correlation mean. interslice correlation standard deviation and spectral entropy) between SD/PHF and stable patients (p < 0.01), but in none of the Td parameters (tQRS, LAD40, RMS40). SD/PHF was more likely in patients with an abnormal ( ≥ 3 abnormal STA parameters) compared to a normal STA result (56% vs 19%; p = 0.005). There was a significant difference in event free survival at 1 year between patients with abnormal and normal STA (61% vs 88%; P = 0.03), but the presence of late potentials was not discriminatory (72% vs 82%; P = 0.4). The sensitivity, specificity and total predictive accuracy for predicting SD/PHF were 29%, 85% and 69% (p = 0.2) for Td and 53%, 83% and 74% (p = 0.005) for STA analysis. The positive predictive characteristics curves showed that higher positive predictive accuracies were achieved by STA compared to those of Td analysis. The differences were significant at different sensitivity levels (at 50%, 60%, 70%. p = 0.05; 80%. p = 0.01). The relative risk (95% Cl) of SD/PHF was 2.1 (0.7-6.0) for presence of late potentials and 3.5 (1.4-9.2) for abnormal STA result. STA analysis of SAECG is a more sensitive and specific predictor of SD/PHF than Td analysis for IDC patients and has an important role in clinical management.