Noninvasive and Direct Measures of Kidney Size in Kidney Donors

Abstract

Kidney size is important for clinical assessment of kidney disease. This study was performed to determine the usefulness of radiologic methods for predicting direct measurements of kidney size. Diagnostic test study. 139 living kidney donors were enrolled. Body mass index, body surface area, and total-body water were estimated from body height and weight. Kidney lengths were estimated using ultrasound sonography (US) and computed tomography (CT), and kidney volumes were estimated from CT scans using the ellipsoid and voxel-count methods. Kidney length and weight were measured directly after donor nephrectomy. Mean measured kidney length and weight were 11.5 ± 0.9 cm and 188.5 ± 33.5 g, respectively. All body indexes correlated with measured size of the left kidney; the highest correlation was between body weight and measured kidney weight (γ = 0.54; P < 0.001). The difference between measured and estimated lengths was greater for US than for CT (-1.1 ± 0.9 cm for US [P < 0.001] vs -0.7 ± 0.7 cm for CT [P < 0.001]). Bland-Altman analysis showed that limits of agreement between CT estimates and measured kidney length (-0.62 to 2.75 cm) were narrower than those between US estimates and measured kidney length (-0.78 to 2.15 cm). Correlation coefficients between radiologic estimates and measured kidney weight were 0.41 and 0.49 for US and CT estimates of kidney length and 0.72 and 0.79 for volume estimates using the ellipsoid and voxel-count methods, respectively (all P < 0.01). No direct measurement of kidney volume. CT estimation of kidney length is more accurate than US estimation, and CT estimation of kidney volume using the voxel-count method is most useful to predict kidney weight.