Beyond Statistical Accuracy: Decision Curve Analysis and Single-Center Independent Validation of the SHA.LIN Score for PCNL Outcomes

Abstract

Objective: To perform a single-center independent validation of the SHA.LIN score and compare its predictive accuracy and clinical net benefit with the Guy’s Stone Score (GSS), S.T.O.N.E. score, and CROES nomogram in a large cohort of patients undergoing percutaneous nephrolithotomy (PCNL).Materials and Methods: We retrospectively analyzed 271 adult patients undergoing PCNL between 2020 and 2025. Stone-free status (SFS) was defined as the absence of stones or asymptomatic fragments <4 mm on non-contrast computed tomography at 1 month. Predictive performance was assessed using area under the curve (AUC), calibration plots, and Decision Curve Analysis (DCA).Results: The SFS rate was 73.1%. The CROES nomogram demonstrated the highest predictive accuracy (AUC: 0.802, 95% CI: 0.765–0.839), followed by the SHA.LIN score (AUC: 0.756, 95% CI: 0.715–0.797). The SHA.LIN score significantly outperformed both GSS (AUC: 0.719, p=0.023) and S.T.O.N.E. (AUC: 0.690, p<0.0001). In both univariate and multivariable analyses, the CROES nomogram and SHA.LIN score emerged as significant independent predictors of stone-free status, whereas Guy and STONE scores failed to retain significance after multivariable adjustment. In DCA, the SHA.LIN score showed superior net clinical benefit compared with GSS and S.T.O.N.E. across threshold probabilities between 40% and 80%.Conclusion: The SHA.LIN score is a reliable, independently validated single-center tool for predicting PCNL outcomes. Although the CROES nomogram demonstrates high statistical accuracy, the SHA.LIN score provides superior clinical utility and net benefit in decision curve analysis. Furthermore, CROES and SHA.LIN scores remained independent predictors of stone-free status, whereas Guy and STONE scores failed to demonstrate independent predictive value in multivariable analyses. Further prospective studies with larger cohorts are warranted to validate these findings.