Construction and validation of a nomogram-based predictive model for uric acid stones with area under the curve evaluation

Xinfeng Chen^#, Xiaoyi Li^#, Xin Xu, Suning Kang, Jie Jiang, Zhenmin Liu, Yong Zhang, Zhan Chen, Cheng Shen, Bing Zheng, Hua Zhu^*
Department of Urology, Affiliated Nantong Clinical College of Nantong University, Nantong First People's Hospital, Nantong, China
* Corresponding Author: Hua Zhu. Email: email
# These authors contributed equally to this work

Canadian Journal of Urology https://doi.org/10.32604/cju.2026.080517

Received 10 March 2026; Accepted 09 June 2026; Published online 17 August 2026

Download PDF

Abstract

Objective: Uric acid stones and non-uric acid stones differ significantly in treatment strategies, yet accurately identifying stone composition preoperatively remains challenging, which directly affects individualized treatment planning. Therefore, this study investigates the clinical determinants of uric acid stone formation to construct a predictive model for clinical application. Methods: The demographic and clinical data of 229 individuals with urinary calculi were retrospectively analyzed. Participants were classified into two groups according to stone composition: a non-uric acid stone group (n = 135) and a uric acid stone group (n = 94). The general information, blood and urine biochemical indices, and imaging findings of the two groups were compared. Binary logistic regression was used to identify independent predictors of uric acid stone formation, leading to the development of a nomogram using R software and evaluation of its performance. Results: Independent predictors identified through logistic regression included age, BMI, urinary pH, CT attenuation values, and diabetes history. A nomogram incorporating these variables identified an optimal cutoff of 0.49. Discriminative performance of the model was high, yielding an area under the curve (AUC) of 0.92 (95% CI: 0.88–0.97) in the training cohort and 0.89 (95% CI: 0.80–0.98) in the internal validation cohort. Calibration analysis demonstrated good agreement between predicted probabilities and observed outcomes. Clinical utility was further supported by decision curve analysis (DCA). In temporal validation, the model achieved an AUC of 0.87 (95% CI: 0.79–0.96). Conclusions: Age, BMI, stone CT attenuation value, urinary pH, and diabetes history emerged as independent predictors of uric acid stone development. For surgically treated patients with urinary stones ≥5 mm, the nomogram model developed on the basis of these factors exhibited good predictive performance.