Support for the Use of a New Cutoff to Define a Positive Urine Culture in Young Children.

Abstract

Conventional urine culture selects for a narrow range of organisms that grow well in aerobic conditions. In contrast, examination of bacterial gene sequences in the urine provides a relatively unbiased evaluation of the organisms present. Thus, by using 16S ribosomal ribonucleic acid (rRNA) gene amplicon sequencing as the reference standard, we now have the ability to assess the accuracy of urine culture in diagnosing urinary tract infection (UTI). We enrolled febrile children 1 month to 3 years of age that underwent bladder catheterization for suspected UTI. Using 16S rRNA gene amplicon sequencing as the reference standard, we calculated the accuracy of urine culture at various cutoffs (10 000, 50 000, and 100 000 colony forming units per milliliter). Children with ≥80% relative abundance of any organism on 16S rRNA gene amplicon sequencing with elevated urinary markers of inflammation were defined as having a UTI. When using a cutoff of 10 000 CFU/mL, the sensitivity and specificity of urine culture were 98% (95% confidence interval [CI]: 93%-100%) and 99% (95% CI: 97%-100%), respectively. Using a cutoff of 50 000 colony forming units per mL decreased sensitivity to 80% (95% CI: 68%-93%) without changing the specificity. Using a cutoff of 100 000 further decreased sensitivity to 70% (95% CI: 55%-84%). Conventional culture remains an accurate method of diagnosing UTIs in young children; however, these data suggest that a cutoff of 10 000 colony forming units per mL provides the optimal balance between sensitivity and specificity for children undergoing bladder catheterization.