Correlation of the endocardial fragmented electrogram with body surface signal-averaged electrocardiographic mapping.

We compared signal-averaged electrocardiography (SAE), SAE mapping, and left ventricular catheter mapping in 60 patients with ischemic heart disease. Using the data obtained in patients with no fragmented electrograms (FE) in the left ventricle, the late potential was defined by SAE as a filtered QRS duration > 131 msec or a root mean square voltage < 16 microV for the last 40 msec of the QRS complex. SAE mapping was performed by recording the signal-averaged electrocardiogram at 48 sites on the body surface. With SAE mapping, the filtered QRS duration and the area in the last 20 msec of the QRS complex were significantly different between the patients with and without FEs. The late potential was defined by SAE mapping as a filtered QRS duration > 136 msec or an area < 28 microV.msec for the last 20 msec of the QRS complex. The sensitivity and specificity of detecting FEs were 46% and 88%, respectively, by the SAE filtered QRS criterion, while they were 66% and 88% by the root mean square criterion. In contrast, SAE mapping gave values of 66% and 92% by the filtered QRS criterion, as well as values of 100% and 92% by the area criterion. Thus, SAE mapping provided better detection of the FE and was more closely correlated with the results of catheter mapping, suggesting its potential for clinical application.