Adding Renal Scan Data Improves the Accuracy of a Computational Model to Predict Vesicoureteral Reflux Resolution

Abstract

We previously developed a computational model to predict vesicoureteral reflux resolution 1 and 2 years after diagnosis. Previous studies suggest that an abnormal renal scan may be a predictor of the failure of vesicoureteral reflux to resolve. We investigated whether the addition of renal scan data would improve the accuracy of our computational model. Medical records and renal scans were reviewed on 161 children, including 127 girls and 34 boys, with primary reflux between 1988 and 2004. In addition to the 9 input variables from our prior model, we added renal scan data on decreased relative renal function (40% or less in the refluxing kidney) and renal scars. Resolution outcome was evaluated 1 and 2 years after diagnosis. Data sets were prepared for 1 and 2-year outcomes, and randomized into a modeling set of 111 and a cross-validation set of 50. The model was constructed using neUROn++. A logistic regression model had the best fit with an ROC area of 0.945 for predicting reflux resolution in the 2-year model. This was improved compared to our previous model without renal scan data. A prognostic calculator using this model can be deployed for availability on the Internet, allowing input variables to be entered and calculating the odds of resolution. This computational model uses multiple variables, including renal scan data, to improve individualized prediction of early reflux resolution with almost 95% accuracy. The prognostic calculator is a useful tool for predicting individualized vesicoureteral reflux resolution.