The agonizing pain experienced by patients during the development of kidney stones greatly impacts their physical and mental well-being. Moreover, kidney stone patients incur substantial direct and indirect economic costs every year [12]. Consequently, there is a pressing need for timely and effective treatment of kidney stones, particularly in complex cases involving renal stone clearance. Endoscopic Combined Intra Renal Surgery (ECIRS) has emerged as an effective means of mitigating this issue, as studies have shown improved stone-free rate and a reduced incidence of complications after ECIRS [8, 13]. However, due to the complexity of the pelvic structure, and the size and the hardness of the stones, surgeons often struggle to predict the probability of successful stone removal before surgery. Therefore, the development of a computationally tractable model could greatly improve physicians' understanding of the difficulty of removing stones before surgery.

In this study, we used logistic regression and lasso regression to establish the key variables influencing the stone-free rate after ECIRS. The results showed hydronephrosis, RIL, RIW, stone burden, and number of calyces involved as the main factors affecting the stone-free rate. Traditional factors such as average Hounsfield unit and skin-to-stone distance, considered to influence stone clearance, were excluded from the study. Using the five identified parameters, we developed a nomogram to predict stone clearance rates. We then validated the effectiveness of the model using the calibration, DCA, and ROC curves. Furthermore, the DCA curves were plotted, showing a significant probability of range threshold, demonstrating that the nomogram had excellent predictive accuracy and that patients would benefit significantly from its use.

The stone burden is historically believed to be an independent factor in the kidney stones' free rate [14]. However, there is no universal standard for stone burden. It is not even explicitly addressed in parts of the literature. Therefore, the first question concerning the definition of stone burden is how to measure it. Clinics usually conduct Kidney-ureter-bladder radiography (KUB) or renal ultrasound (KUS). However, when a patient has symptoms or plans to undergo surgery, then Non-contrast-enhanced CT (NCCT) screening may be considered. The KUS technique is commonly used in outpatient settings and has the advantage of being radiation-free. Moreover, it can be checked multiple times. Studies have revealed that ultrasonography (US) is 45% sensitive and 88% specific to kidney stones [15]. However, KUS findings are largely operator-influenced, and results can vary widely among experienced sonographers. Consequently, KUS is only used typically as a preliminary exam. Although the NCCT is considered the most accurate diagnostic test, it does carry some radiation exposure risks. This risk can, however, be mitigated by low-dose CT, which has demonstrated a combined sensitivity of 93.1% and specificity of 96.6% in diagnosing urolithiasis [16]. As a preoperative test, we used low-dose CT to confirm stone size, location, and other associated details of renal stones. The sensitivity and specificity of KUB x-rays were estimated at 57% and 76%, respectively [17]. In addition, we used CT images as a reference and defined stone burden as Σ (0.785*length (max)* width (max)). In the process, we dropped the three-dimensional calculus size and adopted the planar calculation approach, mainly designed for surgeons to estimate quickly. We also opted not to use the longest path of stones, commonly used in clinics, because this calculation method is inaccurate for assessing alien stones. In this study, the stone burden was found to be a significant prognostic factor, showing a strong association with residual stone rates after ECIRS (OR = 0.995, p < 0.001). Moreover, the predictive ability of stone burden as a variable exhibited a high level of accuracy (AUC = 0.727, CI = 0.668–0.785). Based on the experience of the surgeons in this study, patients with larger stone burden are more likely to produce larger stone fragments and disperse them into the surrounding calyces intraoperatively. If the surgeon can confirm the presence of stones by ultrasound before perfecting the operation, the stone clearance rate will be greatly improved and the pain of patients with recurrent kidney stones will be better alleviated. This is an area where we believe further improvements can be made. Recent relevant studies have also mentioned that the Urologist should carefully evaluate patients with multiple calyceal stones and consider routine use of flexible nephoscopy to retrieve migrated fragments in order to improve their prediction of SFS.

In previous studies examining stone removal rates, hydrocephalus was classified into only two categories: with or without hydronephrosis. Consistent with other studies [18], our results suggested that hydronephrosis is a favorable factor for stone removal. We also discovered that different grades of hydronephrosis had varying effects on stone clearance. In our study, hydrocephalus as a prognostic factor had a strong positive association with the stone-free rate after ECIRS. However, the predictive ability of hydronephrosis as a variable showed a low level of accuracy (AUC = 0.590, CI = 0.531–0.650).

As commonly observed in various studies, the anatomy of the inferior pole of the kidney is an important factor affecting the stone removal rate. The characteristics of IPA, RIL and RIW are often analyzed to understand their effect on the stone clearance rate following retrograde intrarenal surgery (RIRS) [1, 19]. Studies have shown that IPA has a significant effect on the stone clearance rate of lower pole stones after RIRS. When reviewing the literature, we also found that IPA is measured differently. In this study, we used the inner angle formed at the intersection of ureteropelvic axis and the central axis of the lower pole infundibulum as IPA measurement[20]. An IPA < 45° is generally considered to significantly impact renal stone clearance after RIRS [1]. When doing an RIRS surgery on lower Renal Pelvic Stones, surgeons often encounter the anomaly of being able to see stones but not the impact of the holmium laser. However, in our study, IPA did not differ significantly between the SF and SR groups. Because of the ECIRS approach, we suggest that smaller IPA stones can be removed more effectively using PCNL. For RIL and RIW several standards exist [1, 11, 21]. We discovered that stone residue is more likely to occur when RIL ≥ 30 mm and RIW < 5 mm., which can be linked to the difficulty in passing PCNL endoscopes and RIRS ureteroscopes through the neck of calyces. These findings are consistent with other studies as RIL ≥ 30 mm significantly correlated with stone residue rate after ECIRS (OR = 0.205, p = 0.000). However, the predictive accuracy of the RIL variable was low (AUC = 0.601, CI = 0.548–0.655). On the other hand, RIW < 5 mm showed a negative and significant correlation with stone residue rate following ECIRS (OR = 0.519, p = 0.012). However, this scoring scheme may not be appropriate in all situations, such as when IPA is too small, RIL is too long, and RIW is too narrow for ECIRS. Furthermore, the predictive ability of the variable RIW demonstrated poor accuracy (AUC = 0.558, CI = 0.504–0.611).

In our clinical studies, we discovered that as the number of stones increased, the likelihood of being stone-free reduced. In the score of stone removal following PCNL, the S.T.O.N.E. score also highlighted the number of calyces involved as a factor in the surgical removal of stones [22, 23]. Although not explicitly stated, Guy's stone score grades also suggested that stone clearance rates are associated with the number of calyces occupied by stones [24, 25]. The number of calyces involved as a prognostic factor was significantly associated with the stone-free rate after ECIRS in our study. Additionally, the predictive ability of the variable number of calyces involved showed high accuracy (AUC = 0.751, CI = 0.703–0.798).

This study has several limitations. Firstly, it is retrospective, which introduces gaps in patient data collection and clinical variables such as the patient's history of urolithiasis surgery. Secondly, the study population was limited to patients from only two hospitals; hence the sample size was relatively small. As a result, the outcomes of this study have only been internally validated, with a significant number of procedures requiring external validation in subsequent studies. Lastly, this study's nomogram may not apply to patients with anatomic renal anomalies.