A novel model using computed tomography parameters to predict shock wave lithotripsy success in ureteral stones at different locations.

Objective: To combine non-contrast computerized tomography (NCCT)-based parameters with stone and patient characteristics that are already known to affect shock wave lithotripsy (SWL) success and assess this novel model's effectiveness in predicting SWL success for single ureteral stones in different locations.

Materials And Methods: Data of patients treated by SWL for a single ureteral stone between January 2017 and January 2019 were retrospectively reviewed. Demographic parameters of patients and stone characteristics were combined with NCCT-based parameters. NCCT-based parameters included the presence or absence of hydronephrosis, perinephric stranding, periureteral edema, diameter of the proximal ureter, ureteral wall thickness (UWT) at ureteral stone site. The logistic regression method was used for the development of a useful predictive model. Subsequently, the receiver operating curve was used to determine cut-off levels, and a scoring system was developed for prediction of SWL success.

Results: Stone-free rate was 77,1% (267/346) in the entire cohort. Univariate analysis revealed that age, stone volume, density, perinephric stranding, diameter of proximal ureter, and UWT, were associated with SWL success. In multivariate analysis, proximal ureteral stone location, stone volume, density, and UWT were independent predictors of SWL success. The formula used during logistic regression analysis was: 1/[1 + exp {-8.856 + 0.008 (stone volume) + 0.002 (stone density) + 0.673 (UWT) + 1026 (proximal ureteral stone)}]. The scores of 0, 1, 2, 3 and 4 were associated with 97,8%, 83,4%, 60,8%, 33,2% and 11,1% success rates, respectively, in the prediction model based on these parameters.

Conclusion: We conclude that our model can facilitate decision-making for SWL treatment of ureteral stones in different locations.