Noncontrast computed tomography in obstructive anuria: a prospective study.

OBJECTIVES: To evaluate the role of noncontrast computed tomography (NCCT) in the determination of the cause of obstructive anuria and to compare its accuracy with that of the traditional methods of combined plain abdominal x-ray (KUB) and gray-scale abdominal ultrasonography (US).

METHODS: The study included 40 consecutive patients with obstructive anuria. In addition to the routine evaluation, which included history, clinical examination, biochemical profile, KUB, and US, all patients underwent NCCT. The study patients were tested against an age and sex-matched control group that included the normal contralateral kidneys of 57 consecutive patients who underwent KUB, US, and NCCT for acute flank pain during the same study period. The reference standard for the determination of the cause of obstruction was retrograde or antegrade ureterography with or without ureteroscopy or open surgery. The absence of obstruction in the control group was confirmed by nonequivocal normal intravenous urography of the side free of flank pain. Both NCCT and combined KUB and US were compared regarding the sensitivity, specificity, and overall accuracy.

RESULTS: The study group had 48 renal units, because obstruction was bilateral in 8 patients and of a solitary kidney in 32. Of the 42 renal units with calculus obstruction, the site of stone impaction was identified in all renal units by NCCT (sensitivity 100%) and in only 25 by combined KUB and US (sensitivity 59.5%)-a significant difference (P = 0.0001). Of the 6 renal units with noncalcular obstruction, both NCCT and US diagnosed the cause of obstruction in 3. The overall sensitivity of NCCT in the determination of the cause of obstructive anuria was 94% and that of combined KUB and US was 58%-a significant difference (P = 0.0001). The specificity of NCCT was not significantly different from that of combined KUB and US (96.5% versus 93%, respectively). The overall accuracy of NCCT was 95% and that of combined KUB and US was 77%-a significant difference (P = 0.0003).

CONCLUSIONS: In patients with obstructive anuria, conventional KUB and US could not identify the cause of ureteral obstruction in about 40% of the patients. Under such conditions, NCCT can accurately provide the diagnosis, obviating the need of invasive and expensive diagnostic procedures.