Residual stones after percutaneous nephrolithotomy: comparison of intraoperative assessment and postoperative non-contrast computerized tomography

Objectives To compare the intraoperative surgeon perspective for detection of residual fragments (RFs) after percutaneous nephrolithotomy (PNL) with postoperative NCCT. Methods A prospective study of adult patients who underwent PNL between March and September 2014 was conducted. Stone complexity was evaluated using the Guy's stone score (GSS). All patients were evaluated by pre- and postoperative NCCT. After the procedure, the surgeon had been asked whether there were residual stones or not. The sensitivity, specificity and predictive values were tested against postoperative NCCT. Predictors of accurate intraoperative assessment were determined using univariate and multivariate statistical analyses. Results The study included 306 consecutive patients. The surgeons reported no residual stones in 236 procedures; of whom 170 (72%) were reported stone-free by NCCT. On the other hand, 65 out of 70 procedures (93%) reported with residual stones by the surgeons were true by NCCT. The sensitivity was 50% and the NPV was 72%, while the specificity was 97% and the PPV was 93%. On multivariate analysis, only lower GSS (p < 0.001) was independently associated with true negative surgeon opinion. Conclusions Although there was a high surgeon ability to detect post-PNL residual stones, postoperative imaging is mandatory because of the high false negative rates and low NPV. The surgeon opinions can be judged only in stones with lower GSS. The NPV could be enhanced if a consistent definition of clinically significant RFs is introduced.