To identify geographic variability in the imaging of infants with congenital hydronephrosis at initial pediatric urologic evaluation.

We performed a retrospective review of infants aged ≤12 months with congenital hydronephrosis seen as new patients from October 2010 to September 2011 at 3 regionally diverse pediatric urology practices: University of Virginia Hospital, Rady Children's Hospital, and Children's Hospital Colorado. Primary outcomes measured were the type and number of tests ordered at initial evaluation. Independent variables collected included the following: patient age, location, and initial ultrasound findings. Ultrasound findings were manually extracted from the attending pediatric urologist's clinic note. All other data were automatically extracted from the electronic medical record. Proportions were analyzed using Pearson's goodness of fit and Fisher exact tests. Medians were compared using the Kruskal-Wallis test.

Two hundred forty-one patients met the study criteria. Median patient age was 2 months and did not differ across sites. Most patients (64.7%) had Society for Fetal Urology grade 0-2 hydronephrosis; prevalence of high-grade hydronephrosis varied across sites (P = .002). Use of voiding cystourethrography also varied across sites (17.6%-88.9%); this difference persisted when controlling for age and hydronephrosis grade (P <.05). Use of other imaging studies did not significantly differ across sites.

Use of screening voiding cystourethrography for infants with congenital hydronephrosis varies across practices. This variation persists when controlling for differences in age and ultrasound findings, suggesting that regional differences in patient demographics, provider/parental preferences, or referral patterns might contribute to practice variations in the evaluation of these patients.