Variability of digitized left ventricular M-mode echocardiography: a study in healthy subjects and patients with repaired tetralogy of Fallot.

To assess possible factors affecting the variability of digitized left ventricular M-mode echocardiograms, the influence of respiration and the variability due to different beats and observers were analysed in 11 healthy subjects and 11 patients with repaired tetralogy of Fallot. Left ventricular end-diastolic dimension (LVEDD) decreased from end-expiration to end-inspiration in the healthy subjects, but not in the patients. The maximal rate of dimension change decreased in both healthy subjects and patients from end-expiration to end-inspiration. The beat-to-beat variability assessed by the coefficient of variation (CV,%) between measurements of one cardiac cycle was twice the CV for three cycles, whereas the CV for three and five cardiac cycles was not different. The CV for intraobserver variability was less than 5.0% for dimensions and less than 13.0% for the rates of dimension change, whereas the interobserver variability had CV of 17.1% for rates of dimension changes. The influence of respiration and different observers on the variability of LV end-systolic dimension and shortening fraction was larger in the patients than in the healthy subjects. Thus, to obtain optimal technique for analysis of digitized LV M-mode echocardiograms in serial patient studies, the number of observers should be kept at a minimum and at least 3 beats at end-expiration should be used.