Analysis of the Predictive Capacity of the Clinical Variables That Indicate the Performance of Voiding Urosonography in the Study of Vesicoureteral Reflux in Children.

Abstract

Voiding urosonography (VUS) is a dynamic imaging technique which evaluates the lower urinary tract by introducing sonographic contrast into the bladder, preferably used in the diagnosis of vesicoureteral reflux (VUR). Our goals were to describe the clinical indications for performing a VUS for VUR's diagnosis and analyse its diagnostic reliability. Cross-sectional study carried on patients under 15 years old with a VUS taken between November 2013-2020. Sex, age, indications (lower urinary tract infection-LUTI; dilatated lower urinary tract-DLUT; duplex collecting system-DCS), results (presence/absence of VUR, score and side) and complications were analysed. U Mann-Whitney and Chi-squared tests were used and 2 predictive capacity models for the indications used to detect VUR were employed (logistic binomial regression and multilayer perceptron neural network). Statistical significance p < 0.05. 415 VUS were completed correctly (male 51.8%, median age 7.3 (3.1-15.3) months). Indications were: LUTI (67.5%), DLUT (33.5%) and DCS (10.2%); presenting 1, 2 and 3 indications in 86.5%, 12.8% and 0.7% respectively. VUR was diagnosed in 34.7% cases. A tendency towards statistically significance was showed related to male sex (p = 0.052) and the only significative clinical indication for VUR was DCS (p = 0.007). Patients with 2 simultaneous indications had higher probability of VUR (p = 0.012). DCS presence or male sex were the only predictive variables of VUR: DCS = OR 1.89 (1.54-6.52) (p = 0.006) and male sex = OR 1.56 (1.03-2.35) (p = 0.035). VUS is a thriving technique, radiation free and with a low complications rate. Presence of DCS, male sex or 2 simultaneous indications for VUS increase the probability of presenting VUR.