ABSTRACT
Objective:To compare the accuracy of different stone scoring systems for predicting the stone-free rate (SFR) after retrograde intrarenal surgery (RIRS).Methods:The clinical data of 227 patients with lithiasis undergoing RIRS from June 2017 to December 2020 in Affiliated Benq Hospital of Nanjing Medical University and Qingdao Fuwai Hospital were retrospectively analyzed. There were 152 males and 75 females. The average age was (53.0±10.4) years old. The average body mass index was (26.9±2.1)kg/m 2. The maximum diameter of the stone was (22.7±12.8)mm. The stone is located in left side in 133 cases and in right side in 94 cases. The stones of 44 cases were located in upper ureter, upper calyceal or renal pelvis, that of 23 cases were in medium calyceal, 157 cases in lower calyceal, and 3 cases in calyceal diverticulum.The average CT value of stone was (778.3±350.4)HU. American Society of Anesthesiology (ASA)scores: 86 cases of grade Ⅰ, 129 cases of grade Ⅱ, 12 cases of grade Ⅲ. Preoperative non-contrast CT was conducted and three-dimensional data were constructed. A single observer reviewed and entered the modified S.T.O.N.E., RUSS, modified S-ReSC, R. I.R.S., SHA.LIN, Ito nomogram, S. O.L.V.E., stone free index (SFI) scores. Logistic analysis were performed between every score and SFR. Receiver operating characteristic (ROC) curve was drawn to detect sensitivity and specificity of every score in predicting the SFR. The predictive accuracies of all scores were compared. Results:The SFR was 83.0%(189/227). There were statistically significant differences in modified S. T.O.N.E.(10.5±1.9 vs. 12.7±1.8), RUSS[1(0, 4) vs. 3(0, 6)], modified S-ReSC (8.2±5.6 vs. 11.8±6.0), R.I.R.S.(6.2±1.4 vs. 8.1±1.2), SHA.LIN (9.9±2.4 vs. 13.0±2.1), Ito nomogram (12.1±5.8 vs. 4.3±3.3), S. O.L.V.E. (6.8±1.6 vs. 8.7±1.2), SFI score (7.9±1.1 vs. 6.3±0.9) between the stone-free group and the stone remaining group ( P <0.05). Logistic regression revealed that modified S.T.O.N.E., RUSS, modified S-ReSC, R. I.R.S., SHA.LIN, Ito nomogram, S. O.L.V.E. and SFI score were significantly associated with SFR( P<0.05). There were no significant differences in the area under the curve (AUC) between the modified S. T.O.N.E., RUSS, R. I.R.S., SHA.LIN, Ito nomogram, S. O.L.V.E. and SFI score( P>0.05), but there were significant differences in the AUC between modified S-ReSC score and other score ( P<0.05). When the cutoff of SHA.LIN, SFI and R. I.R.S. score was determined as 10, 6 and 6 scores, the specificity of SHA.LIN, SFI and R. I.R.S. score was 94.7%, 92.6% and 89.5%, respectively. Conclusions:All score could predict the postoperative SFR of RIRS, while the SHA.LIN, SFI and R.I.R.S. score were more accurate than the other scores. The accuracy of the modified S-ReSC in predicting SFR after RIRS was slightly worse than other scores.