Does vesicoureteral reflux severity impact breakthrough UTI risk: a multi-institutional analysis and risk...

Vesicoureteral reflux (VUR) is a common urological disorder in children. To date, VUR severity is assessed by grading, and contemporary models have expanded on developing predictors for breakthrough UTI (bUTI) in VUR. Here, we developed models to predict the risk of bUTI in children with VUR using clinical features, combined with traditional VUR grading or novel quantitative VUR (qVUR) measures.

An international multi-institutional retrospective cohort database between 2015 and 2022, including four pediatric centers, was queried for VUR cases with available voiding cystourethrograms (VCUG). Each VCUG was reviewed, and VUR grade and qVUR features (including ureteral tortuosity and dilatation) were assigned by multiple raters. The outcome of interest, development of bUTI, was analyzed using multivariable Cox proportional hazards and boosted Cox regression. The models were trained and externally validated to provide individualized survival curves. Model performance, calibration, and net benefit were assessed.

Six hundred eighty-four children were included in this study (308 training, 376 validation), of which 109 (16%) experienced a bUTI within the study period. Age, sex, VUR grade, and ureteral dilation were associated with bUTI on survival analysis. The qVUR-based Cox regression and boosted Cox models performed with c-indices of 0.64 (95%CI 0.60, 0.68) and 0.75 (95%CI 0.67, 0.80), both superior to VUR-based models. qVUR-based Cox models provided improved calibration, and net benefit. The bUTI risk algorithm, based on the Cox model, is available in a free-to-use web application ( https://sickkidsurology.shinyapps.io/qVUR-bUTI/ ).

Our findings suggest that individualized bUTI risk assessment incorporating an objective measure of VUR severity can be modelled with modest performance. With further works, qVUR holds promise in refining predictive models in VUR.