Residual stones after percutaneous nephrolithotomy; sensitivities of different imaging methods in renal stone detection.

Kidneys of 35 renal stone patients treated with percutaneous nephrolithotomy (PNL) were examined at 12-36 months (mean 23 months) after the procedure by plain radiography (PR), conventional linear tomography (LT), computed tomography (CT), and ultrasonography (US) to find out the long-term stone-free rate as well as the accuracy of these imaging methods to detect residual or recurrent renal stones. The present series included 36 kidneys with a 56% postoperative stone-free status examined by plain radiography and a 58% one examined by linear tomography at 12-36 months. CT showed 47% and US 72% of the kidneys to be stone-free, respectively. The sensitivity of CT was superior to plain radiography (P < 0.05), linear tomography (P < 0.05) or US (P < 0.001) in the detection of residual or recurrent stones after percutaneous nephrolithotomy analyzed statistically by McNemar's test. In the same group of 35 patients, which consisted of the above 36 postoperative kidneys together with 19 contralateral kidneys, a total of 55 kidneys were evaluated by the above methods, revealing 56 stones in 31 kidneys. Of these stones 95% were detected by CT, 70% by plain radiography, 68% by linear tomography, and 32% by renal US. Of the 55 kidneys, 14 contained 20 stone fragments smaller than 3 mm, and 26 kidneys carried 43 concrements smaller than 6 mm. The sensitivities of the four imaging methods in the detection of these small fragments were as follows: plain radiography 60% and 60%, linear tomography 55% and 58%, CT 95% and 93%, and US 15% and 26%.(ABSTRACT TRUNCATED AT 250 WORDS)