Reproducibility of signal-averaged electrocardiography

Abstract

To evaluate the reproducibility of results with signal-averaged ECG (SAECG), three recordings (initial, 10 minutes later and 1 week later) were compared in 20 healthy male subjects. Subsequently, two recordings (initial and 10 minutes later) in 50 healthy male subjects (Group N) and 42 patients with myocardial infarction (Group M 1) were also compared. All healthy male subjects showed normal ECG. In Group M 1 which consisted of 37 males and 5 females, SAECG was recorded at least 4 weeks after the first attack. The recording system was ART1200EPX (Arrhythmia Research Technology : USA) which showed superior reproducibility to VCM-3000 (Fukuda Denshi : Tokyo) on preliminary studies. SAECG was obtained from orthogonal XYZ leads and analyzed by the time domain method (TDM) and frequency domain with fast Fourier transformation (FFT) technique. In TDM, diagnostic criteria were f QRS > 114msec, LAS > 38msec and RMS40 < 20μV, and the definition of late potential (LP) was the presence of more than 2 abnormal indices. In FFT, when the area ratio (AR : 20-50Hz/ 10-50Hz) was greater than 107, LP was considered positive. In 20 healthy male subjects, the diagnostic concordance ratio (positive and positive, or negative and negative) was 90% between the two initial and the second recording or the initial and third recording by TDM. It was 100% between the two initial and the second recording or the initial and third recording by FFT. In Group N, the concordance ratio was 92% by TDM and 100% by FFT. In Group M 1, the concordance ratio was 93% by TDM and 90% by FFT. In conclusion, the reproducibility of the diagnostic results on SAECG by TDM and FFT were excellent despite variation in individual measurements in both healthy subjects and patients with myocardial infarction.