Development and validation of a simple stone score to estimate the probability of residual stones prior to percutaneous nephrolithotomy.

ABSTRACT

BACKGROUND: The aim of the present study was to develop and internally validate a simple stone score (SSS) to estimate the probability of clinically significant residual fragments (CSRF) prior to percutaneous nephrolithotomy (PNL). METHODS: The files of 1170 PNL procedures between January and December 2015 were evaluated. CT-derived stone characteristics were examined. Caliceal stone distribution (CSD) was assigned three grades based on the number of calices involved regardless of the renal pelvis (I = no or single calix; II = more than one calix; and III = more than 2 calices or complete staghorn stones). CSRF was defined as any residuals >4 mm in postoperative CT. A logistic regression model to predict the CSRF was fitted, and coefficients were used to develop the SSS. The SSS was validated by discrimination, calibration, and decision curve analysis (DCA). RESULTS: Patients' data were split into training (936, 80%) and validating (234, 20%) datasets. In the training partition, independent predictors of CSRF were CSD-grade II (OR 4.2; 95%CI 2.5-7; P<0.001), grade III (OR 7.8; 95%CI 4.2-14.4; P<0.001) and largest stone diameter (LSD) (OR1.3; 95%CI 1.1-1.6; P<0.001). Score points 0, 1, 2, and 0, 3, 9 were given to LSD<30, 30-40, >40 mm, and CSD grades I, II, III, respectively. Discrimination of the SSS was 0.79 and after 10-fold cross-validation and internal validation was 0.86. The calibration plot and DCA highlighted the validity and clinical significance of the SSS. CONCLUSIONS: The novel SSS could be used to describe the risk of CSRF prior to PNL. Further studies are invited for external validation.