Clinical prediction modeling of baseline differential renal function based on ultrasound in congenital unilateral ureteropelvic junction obstruction.

OBJECTIVE

To develop a clinical prediction model using general clinical and renal ultrasound parameters to noninvasively monitor baseline renal function in children with unilateral ureteropelvic junction obstruction (UPJO), thereby reducing the need for invasive procedures.

MATERIALS AND

METHODS

This prospective study included 379 children with unilateral UPJO treated between March 2019 and December 2022. Patients were randomly divided into training and validation sets (7:3 ratio). Binary logistic regression identified independent predictors of impaired differential renal function (DRF), which were incorporated into a predictive nomogram. Model performance was evaluated using ROC analysis, calibration curves, and decision curve analysis (DCA).

RESULTS

A total of 379 children with UPJO were included in this study. Renal parenchymal thickness (PT) min, Hydronephrosis volume/Renal volume (HV/RV), and corrected contralateral renal volume (RV) were independent predictors of impaired baseline DRF. A nomogram based on these predictors was developed to estimate the risk of baseline DRF impairment in children with UPJO. The area under the receiver operating characteristics curve (ROC) was 0.778 (95% CI: 0.715-0.842) and 0.925 (95% CI: 0.866-0.985) in the training and validation sets, respectively, showing good differentiation ability, and the calibration curve and clinical decision analysis curve validated the better calibration ability and clinical net benefit of the model.

CONCLUSIONS

A nomogram incorporating PTmin, HV/RV, and corrected contralateral renal volume effectively predicts baseline DRF impairment in children with UPJO, showing robust performance in internal validation. External validation is warranted to confirm its clinical applicability.