Model for predicting high-grade vesicoureteral reflux in young children presenting with febrile urinary tract infection

• Main Authors: Anuvat Klubdaeng, Thanaporn Chaiyapak, Achra Sumboonnanonda, Nuntawan Piyaphanee
• Other Authors: Siriraj Hospital
• Format: Article
• Published: 2022
• Subjects: Medicine
• Online Access: https://repository.li.mahidol.ac.th/handle/123456789/74954
• Institution: Mahidol University

Introduction: High-grade vesicoureteral reflux (VUR) is associated with long-term renal outcome, and can be missed if voiding cystourethrogram (VCUG) is not performed. The current febrile urinary tract infection (UTI) guidelines have narrowed down the indications for VCUG. This study aimed to determine the predictors and develop a model to predict high-grade VUR (grade III-V) in young children presenting with febrile UTI. Methods: We retrospectively reviewed 324 children aged 2–60 months who presented with first or recurrent febrile UTI and underwent both renal and bladder ultrasonography (RBUS) and VCUG during 2004–2013. Multivariate logistic regression was used to identify independent predictors of high-grade VUR, and a predictive model and scoring system were developed. Sensitivity and specificity of the predictors were evaluated. Results: Sixty-two (67.4%) of 92 children with VUR were high-grade VUR. The independent risk factors for high-grade VUR were recurrent UTI (p < 0.001), non-Escherichia coli (non-E. coli) pathogen (p = 0.011), and abnormal RBUS (p < 0.001) including pelvicalyceal dilation only (p = 0.003), and with ureteric dilation (p < 0.001). Abnormal RBUS showed highest sensitivity (58.1%), while ureteric dilation had highest specificity (96.6%) to predict high-grade VUR. Combination of these predictors improved the specificity but decreased the sensitivity. The prediction score was calculated, as follows: recurrent UTI = 3 points + non-E. coli = 1 point + pelvicalyceal dilation only = 1 point + pelvicalyceal and ureteric dilation = 3 point. A cutoff score of 1 showed sensitivity of 82.3% and specificity of 63.7%. The specificity would increase (87.8, 92.7, 97.7, 100, 100, and 100%) for higher scores of 2, 3, 4, 5, 6 and 7, respectively. Discussion: High-grade VUR could be predicted by simple model using single or combined predictors. Sensitivity and specificity varied according to each predictor. The specificity of ureteral dilation was as high as recurrent UTI. The weighing score for each presenting predictor and the summation of the scores provided more detailed in sensitivity and specificity. Conclusions: Children with febrile UTI who have the predictor(s) of high-grade VUR are recommended to undergo VCUG. The high-grade VUR prediction score may help for decision-making regarding whether to perform VCUG in a particular child.