Comparison of helical computerized tomography and plain radiography for estimating urinary stone size

We evaluated the accuracy of noncontrast spiral computerized tomography (CT) for determining urinary stone size compared with plain x-ray.

We retrospectively analyzed noncontrast helical CT and plain x-ray of the kidneys, ureters and bladder images of 39 patients with urolithiasis who underwent each study from July 1997 to February 1999. Stone size on x-ray was measured in the craniocaudal and transverse dimensions by a single radiologist (I.C.B.). The size of the same stone was then determined on blinded noncontrast spiral CT images. Stones that were less than 3 mm. were excluded from study. On CT estimated craniocaudal dimension was based on collimation thickness, the reconstruction interval and the number of images on which the stone was visualized.

We evaluated 58 stones 3 mm. or larger in the greatest dimension on plain x-ray of the kidneys, ureters and bladder, and noncontrast spiral CT, including 15 in the distal ureter, 7 in the mid or proximal ureter and 36 in the kidneys. Stone size was 3 to 18 mm. in the greatest dimension. Mean transverse dimension of the stone plus or minus standard deviation on noncontrast spiral CT was 5.1 +/- 1.08 versus 4.9 +/- 1.08 mm. on plain x-ray of the kidneys, ureters and bladder (paired t test; p = 0.335). The mean craniocaudal dimension on noncontrast spiral CT was 7.5 +/- 1.98 versus 6.7 +/- 1.98 mm. on plain x-ray paired t test (p = 0.005).

Noncontrast spiral CT enables a similar measurement of stone size along the transverse dimension as plain x-ray of the kidneys, ureters and bladder. In individuals noncontrast spiral CT does not accurately measure the craniocaudal dimension of stones compared with plain x-ray when standard acquisition and reconstruction techniques are used. In this regard it has a tendency to overestimate stone size by an average of 0.8 mm.