LEFT-VENTRICULAR EJECTION FRACTION IN PATIENTS WITH NORMAL AND DISTORTED LEFT-VENTRICULAR SHAPE BY 3-DIMENSIONAL ECHOCARDIOGRAPHIC METHODS - A COMPARISON WITH RADIONUCLIDE ANGIOGRAPHY

Background: Serial evaluation of left ventricular (LV) ejection fracti on (EF) is important for the management and follow-up of cardiac patie nts. Our aim was to compare LVEF calculated from two three-dimensional echocardiographic (3DE) methods with multigated radionuclide angiogra phy (RNA), in patients with normal and abnormally shaped ventricles. M ethods and Results: Forty-one consecutive patients referred for RNA un derwent precordial rotational 3DE acquisition of 90 cut-planes. From t he volumetric data set, LVEF was calculated by (a) Simpson's rule (3DS ) through manual endocardial tracing of LV short-axis series at 3 mm s lice distance and (b) apical biplane modified Simpson's method (BMS) i n 29 patients by manual endocardial tracing of the apical four-chamber view and its computer-derived orthogonal view. Patients included thre e groups: A, 17 patients with LV segmental wall motion abnormalities; B, 13 patients with LV global hypokinesis; and C, 11 patients with nor mal LV wall motion. For all the 41 patients, there was excellent corre lation, close limits of agreement, and nonsignificant difference betwe en 3DS and RNA for LVEF calculation (r = 0.99, [-6.7, +6.9] and p = 0. 9), respectively. For the 29 patients, excellent correlation and nonsi gnificant differences between LVEF calculated by both 3DS and EMS and values obtained by RNA were found (r = 0.99 and 0.97, p = 0.7 and P = 0.5, respectively). In addition, no significant difference existed bet ween values of LVEF obtained from RNA, 3DS, and EMS by the analysis of variance (p = 0.6). The limits of agreement tended to be closer betwe en 3DS and RNA (-6.8, +7.2) than between EMS and RNA (-8.3, +9.7). The intraobserver and interobserver variability of RNA, 3DS, and EMS for calculating LVEF(%) were (0.8, 1.5), (1.3, 1.8), and (1.6, 2.6), respe ctively. There were closer limits of agreement between 3DS and RNA for LVEF calculation in A, B, and C patient subgroups [(-3.5, +5), (-8.4, +5.6), and (-7.8, +8.6)] than that between EMS and RNA [(-8.1, +10.7) , (-11.9, +9.3), and (-9.1, +11.3)], respectively. Conclusions: No sig nificant difference existed between RNA, 3DS, and EMS for LVEF calcula tion. 3DS has better correlation and closer limits of agreement than E MS with RNA for LVEF calculation, particularly in patients with segmen tal wall motion abnormalities and global hypokinesis. 3DS has a compar able observer variability with RNA. Therefore the use of 3DS for seria l accurate LVEF calculation in cardiac patients is recommended.