Regional left ventricular ejection fraction with single crystal gammacamera and gated equilibrium technique: repeatability and inter-observer variability.

Regional left ventricular ejection fraction calculated by means of gated blood pool scintigraphy was compared with that obtained from contrast ventriculography in a group of 20 patients, submitted to both procedures within a month. To evaluate repeatability both at rest and during stress, a second scintigraphic study was carried out after three weeks, with no change in drug regimen. Inter-observer variability was assessed in a randomly selected subgroup of 16 patients, in whom a computation was repeated by a second operator. Comparison of scintigraphic results showed a satisfactory agreement, and measurement agreement for global left ventricular ejection fraction was very good. As regards interobserver variability, each of the three myocardial segments examined (septal, lateral and infero-apical) showed good agreement. As regards repeatability, global left ventricular ejection fraction, proved readily reproducible both for rest and stress studies, with relatively small differences in observed values. Regional ejection fraction, on the other hand, had a high variability, especially during stress. It is suggested that for regional ejection fraction evaluation, reproducibility studies should be performed in order to better correlate result changes to clinical outcome after medical or surgical therapy.