Comparison Of Guy’s Stone Score, Stone Score, And Seoul National University Renal Stone Complexity Score On CT Scan As Predictors Of Stone-Free Rate After Percutaneous Nephrolithotomy: A Retrospective Study

Abstract

Background: Percutaneous nephrolithotomy (PCNL) is the primary treatment of choice for large (>2 cm) or complex renal calculi. Procedural success is highly dependent on preoperative factors, which can be evaluated via non-contrast computed tomography (NCCT). Several CT-based predictive scoring systems have been developed, including the Guy’s Stone Score (GSS), STONE Score, and Seoul National University Renal Stone Complexity Score (S-ReSC). Direct comparative studies evaluating the performance of all three scoring systems simultaneously in the same population remain limited, particularly in Indonesia. Objective: To compare the predictive ability of GSS, STONE Score, and S-ReSC in predicting the stone-free rate (SFR) following PCNL at Dr. Soetomo General Hospital. Methods: This was a retrospective observational analytic study involving 44 patients who underwent unilateral single-tract PCNL at Dr. Soetomo General Hospital between January and December 2025, who met the inclusion and exclusion criteria. The independent variables were the GSS, STONE Score, and S-ReSC assessed on preoperative NCCT. The dependent variable was SFR (defined as residual fragments <4 mm) evaluated via ultrasonography and/or plain kidney-ureter-bladder (KUB) radiography postoperatively. Data were analyzed using the Mann-Whitney test, bivariate and multivariate logistic regression, and receiver operating characteristic (ROC) curve analysis to determine the area under the curve (AUC). Results: The overall SFR was 50.0% (22/44 patients). Bivariate analysis demonstrated significant associations with SFR for all three scores (GSS p=0.013; STONE p=0.004; S-ReSC p<0.001). In the multivariate model, only S-ReSC remained a significant independent predictor (OR=2.32; p=0.025; 95% CI: 1.11–4.85). The S-ReSC yielded the highest AUC (0.840; 95% CI: 0.709–0.947), compared to the STONE Score (0.777; 95% CI: 0.625–0.902) and GSS (0.748; 95% CI: 0.602–0.874). The optimal cut-off value for S-ReSC was ≥5, providing a sensitivity of 81.8% and specificity of 81.8%. Conclusion: The S-ReSC demonstrated superior predictive performance compared to the GSS and STONE Score in predicting SFR after PCNL and was the only independent predictor in the multivariate model. An S-ReSC cut-off of ≥5 can serve as a valuable clinical reference for preoperative risk stratification and procedural planning.