Purpose The diagnostic accuracy of contrast-enhanced voiding urosonography (ce-VUS) in diagnosing vesicoureteral reflux (VUR) is high compared with voiding cystourethrography. However, its reliability has not been yet adequately evaluated. The purpose of this study is to assess the reliability of ce-VUS in VUR detection and grading by estimating the inter- and intra-observer agreement of two radiologists.

Patients and methods Two hundred ten children (86 boys/124 girls, mean-age 2.7y) with 421 pelvi-ureteral-units underwent ce-VUS examination with a second-generation contrast-agent to detect possible (180) or follow-up known (30) VUR. The video-clips of all ce-VUS examinations were twice independently assessed by two paediatric radiologists 4–6 weeks apart. The inter- and intra-observer agreement was estimated by kappa statistic.

Results The inter- and intra-observer agreement of both radiologists regarding the presence or grading of VUR was perfect (k=0.90–0.94). There were only two disagreements regarding the presence of VUR (grade I and II false-negative and false-positive respectively). There were 5 cases of disagreement in VUR grading: three cases of VUR grade II-III and two cases grade III-IV. VUR was detected in 123(29%) pelvi-ureteral-units of 87(41.4%) children and it was more common in completely duplicated ureters (6/7) than in single ones (p = 0.03). The rate of VUR was independent of sex/age/presence of hydronephrosis (p > 0.05).

Conclusions The reliability of ce-VUS in VUR detection and grading is high. VUS with a second generation ultrasound contrast-agent could be used as a radiation-free alternative.