High-resolution and signal-averaged electrocardiography to separate post-myocardial infarction patients with and without ventricular tachycardia.

Abstract

High-resolution and signal-averaged ECG, 24 h Holter recording and ejection fraction were used to separate post-myocardial infarction patients with and without ventricular tachycardia (VT) among 150 individuals: 26 patients with an old myocardial infarction and documented sustained VT, 104 patients with an acute myocardial infarction without sustained VT, who were followed-up for 2 years, and 20 healthy volunteers. Bipolar orthogonal XYZ leads were recorded, high-pass filtered at cut-off frequencies of 25, 40, 60, 80 and 100 Hz, and combined to vector magnitude square root of X2 + Y2 + Z2. The filtered QRS duration, the root-mean-square voltages of different time intervals and the durations of low amplitude signals under different thresholds, both from the initial and terminal QRS, were calculated. The sensitivity and specificity of each parameter alone and in every combination of two, three and four parameters (17 million different combinations) were computed both from non-averaged and averaged data. The best separation was achieved by 12 combinations all including four signal-averaged ECG parameters, with a sensitivity of 81% and a specificity of 79%. The parameters represented most were: filtered QRS duration at 25 Hz, RMS voltage of the last 50 ms at 25 Hz, terminal LAS duration at 80 Hz, and RMS voltage of the last 20 ms at 80 Hz. Parameters of the initial QRS complex did not improve either the sensitivity or the specificity of the method. In logistic regression analysis, the best combinations of four signal-averaged ECG parameters separated VT patients better (P < 0.001) than non-sustained ventricular tachycardia at Holter (P = 0.001); left ventricular ejection fraction (P = 0.01); or age (P = 0.006). Parameters calculated from averaged data gave better results than parameters calculated from non-averaged data.