Development and validation of a predictive model for symptomatic urinary stones in pregnant women

Background This study aims to construct a predictive model by analyzing clinical data of patients, providing a reference for selecting treatment strategies for symptomatic urinary tract stones in pregnant women. Methods We performed a retrospective review of 252 pregnant women with symptomatic urinary tract stones from January 2018 to January 2025. Multivariate logistic regression analysis was used to identify factors influencing the efficacy of conservative treatment, and a nomogram was constructed. The model was evaluated using ROC curves, the area under the curve (AUC), and the Hosmer-Lemeshow test. Results A total of 252 patients were included in the study, of whom 198 showed no response to conservative treatment. Identified independent risk factors included white blood cell count, number of stones, stone size, and degree of hydronephrosis. The AUC for the training set was 0.84 (95% CI: 0.77–0.91), and for the validation set, it was 0.81 (95% CI: 0.70–0.93). The optimal cutoff value for the ROC curve in the training set was 0.738, corresponding to a sensitivity of 0.78 and a specificity of 0.76. Using the same cutoff value in the validation set, the sensitivity was 0.78 and the specificity was 0.83. The Hosmer-Lemeshow test indicated good model fit. Conclusions This study identified white blood cell count, number of stones, stone size, and degree of hydronephrosis as factors influencing the efficacy of conservative treatment. The predictive model constructed based on these factors provides guidance for treatment decision-making in pregnant patients with symptomatic renal stones.