Repeatability and agreement of real time three-dimensional echocardiography measurements of left ventricular mass and synchrony in young patients

Background Left ventricular mass (LVM) and synchrony have prognostic value for many cardiovascular disease states. We report the agreement and repeatability of LVM estimation by three-dimensional (3D) versus M-mode and repeatability of 3D estimation of systolic dyssynchrony. Methods 3DLVM was computed by subtraction of endocardial from epicardial volume X1.05 both at end-diastole and end-systole. M-mode measurements were made at end-diastole. This prospective study comprised 40 subjects, 20 patients with chronic kidney disease or treated neuroblastoma and 20 healthy individuals. The median age was 17 (range 6-29 years). Results Intra- and inter-observer intraclass correlation was excellent for 3D systolic LVM (0.99, 0.87), 3D diastolic LVM (0.99, 0.93), M-mode LVM (0.88, 0.93), moderate for 16-segment SDI (0.77, 062), moderate to low for 12-segment SDI (0.48, 0.73), and 6-segment SDI (0.37, 0.69). The median (range) LVM measurement for 3D diastolic LVM was 125 g (50-253), 3D systolic LVM 109 g (40-195), and M-mode LVM 115 g (range 40-207). There was a significant bias for diastolic 3DLVM to be higher than systolic 3D or M-mode. Limits of agreement between methods were wide. The median (range) systolic dyssynchrony measurements were 2.0 (0.4-7.0), 1.5 (0.3-4.3), and 1.4 (0.3-4.5) for 16-segment, 12-segment, and 6-segment models, respectively. Conclusion 3D and M-mode measurement of LVM are highly repeatable. Measurement bias and wide limits of agreement mean that the same echocardiographic technique should be used during follow-up. Measurement of 3D systolic dyssynchrony is most repeatable using a 16-segment model.