Radionuclide determination of absolute LV volumes: interstudy, interobserver and intraobserver variances.

The results of 22 absolute left ventricular volume (LVV) determinations by a radionuclide (RN) method are compared to the results obtained by contrast ventriculography (CV). Another 10 patients were analysed in order to evaluate the interstudy, interobserver and intraobserver variances. Good correlation was shown between the RN and CV measurements of the end diastolic volume (EDV), end systolic volume (ESV), stroke volume (SV) and ejection fraction (EF), but the RN method overestimates the EDV and ESV. The EF was underestimated, but no difference could be shown for the SV. On the inter- and intraobserver levels, regression analysis yielded excellent correlation (r greater than 0.99 in all cases) with no statistically significant difference (P less than 0.05). The interstudy variance was minimal as indicated by regression analysis (r greater than 0.87) and no statistically significant difference (P less than 0.05) could be shown between studies. The results indicate that the RN method of LVV determination can be used in intervention studies over a limited period.