Variation in the documentation of findings in pediatric voiding cystourethrogram

Background Few standards exist for reporting results of voiding cystourethrogram (VCUG). Objective To assess the variation in reporting of VCUG findings from different facilities using a standardized assessment tool. Materials and methods VCUG reports were evaluated for demographic, technical, anatomical and functional information. Reports were categorized by age, gender, indication and vesicouretal reflux (VUR) status. Institutions were classified as a free-standing pediatric hospital (n=3), pediatric hospital within a hospital (n=11), or non-pediatric facility (n=24) and reports were classified as having been read by a pediatric radiologist or not. Each category of outside reports (n=152) was randomly matched with a twice-larger group of Hospital A reports from the same category (n=304). Multivariate linear regression was used to analyze the association between the primary outcome (percentage of items described in dictated VCUG report) and the type of radiologist and institution. Results Of the 456 studies, 66% were in girls, 56% were in those <12 months old, and the indication was urinary tract infection (UTI) in 81%. The mean percentage of items reported was 67 +/- 14% (74 +/- 7% at free-standing pediatric hospitals, 61 +/- 10% at pediatric hospitals within a hospital, and 48 +/- 11% at non-pediatric facilities). In multivariate analysis, VCUG reports generated at non-pediatric facilities had 17% fewer items included (95% CI: 14.5-19.7%, P<0.0001), and pediatric hospitals within a hospital had 9% fewer items included (5.9-12.5%, P<0.0001) when compared to free-standing pediatric hospitals. Reports read by a pediatric radiologist had 12% more items included (9.1-15.3%, P<0.0001) compared to those read by a non-pediatric radiologist. Conclusion More complete VCUG reports were observed when generated at free-standing pediatric hospitals and when interpreted by a pediatric radiologist.