External Validation of Croes Stone Score System to Predict Treatment Success and Complications of Percutaneous Nephrolithotomy

Introduction: To validate CROES scoring system for predict treatment success and complication rates of patients who underwent percutaneous nephrolithotomy (PNL). Material and Methods: We retrospectively analysed 568 patients who underwent PNL from January 2012 to August 2015 at Izmir Bozyaka Research and Training Hospital Urology Department. Sixty-one patients who had more than one access were excluded from the study. All patients evaluated with non-contrast spiral computed tomography (CT) scan before operation. Stone-free status was assessed on 1-month postoperative visit using kidney-ureter-bladder (KUB) radiograph. CT is reserved for only symptomatic patients, radiolucent stones, and/or suspicion of residual fragments seen on KUB radiography. Residual stone size of <4 mm on KUB or CT was described as clinically insignificant residual fragments, and these patients were considered as stone free. All patients were evaluated one by one with CROES scoring system. Stone-free status, operation time, fluoroscopy time, length of hospital stay and estimated blood loss were correlated with CROES score. Postoperative complications were scored with Modified Clavien scoring systems and correlated with CROES score. Results: The mean CROES score was 203,70. Overall stone free rates was %77,9. The mean stone size was 502,01 ± 517,5 (55-2869) mm.. 112 (%22,1) of the patients had residue stone. 159(32,5%) patients experienced postoperative complications. There was positive correlation between CROES stone scoring system with stone-free rates (p<0,001) and negative correlation with complication rates (p<0,001). There was negative correlation with statistically significant between CROES scoring system with operative time (p<0,001), stone burden (p<0,001), estimated blood loss (>250 mL) (p=0,007) and length of hospital stay (p=0,026). CROES score was found to be independent predictor of SFR in a multivariate logistic regression analysis. (p<0,001). CROES score was not an independent factor for predicting complication. Conclusion: In the present study, we demonstrated that CROES scoring system can be effectively and reliably used to predict SFR after PCNL. Also, CROES was effective to predict postoperative complications. This scoring system can guide the clinician to predict perioperative and postoperative PCNL outcomes