Does success in percutaneous nephrolithotomy depend only on stone size? Analysis of the predictive capacity for success and complications of the current nephrolithometry scoring systems and their relationship with the stone surface

Abstract

ObjectiveTo analyze the predictive capacity of the nephrolithometry scoring systems (GSS, STONE, CROES and S-ReSC) and stone surface regarding success and complications following percutaneous nephrolithotomy (PCNL). MethodsWe studied 392 patients who had undergone PCNL in our center. Only patients with a non-contrast CT (n = 240) were finally included for analysis. The predictive capacities for success and complications of the different scoring systems were evaluated using ROC curves and their area under the curve (AUC). ResultsRegarding success, the S-ReSC system had the highest predictive capacity with an AUC of 0.681 (95% CI 0.610−0.751), followed by the CROES with 0.667 (95% CI 0.595−0.738), the STONE with 0.654 (95% CI 0.579−0.728) and finally the GSS with 0.626 (95% CI 0.555−0.698). The stone surface as a single variable had an AUC of 0.641 (95% CI 0.565−0.718). As for complications, the S-ReSC had the highest AUC with 0.664 (95% CI 0.57−0.758), followed by STONE with 0.663 (95% CI 0.572−0.755), GSS with 0.626 (95% CI 0.555).−0.698) and CROES with 0.614 (95% CI 0.518−0.7). The stone surface alone had an AUC of 0.616 (95% CI 0.522−0.715). ConclusionThe nephrolithometry scales analyzed show a moderate predictive capacity for success and complications in patients undergoing PCNL in our center. Moreover, stone surface as an independent variable demonstrates moderate predictive capacity for both outcomes.