Accuracy of Screening Tests for the Diagnosis of Urinary Tract Infections in Young Children

BACKGROUND: The objective of this study was to compare the accuracy of available tests for pyuria, including newer automated tests, and to examine the implications of requiring them for the diagnosis of urinary tract infections (UTIs). METHODS: We included children between 1 and 36 months of age undergoing bladder catheterization for suspected UTIs who presented to 1 of 3 pediatric centers. Using a positive urine culture result as the reference standard, we compared the sensitivity of 5 modalities for assessing pyuria at the cutoffs most often used clinically for detecting children with a positive culture result: leukocyte esterase on a dipstick, white blood cell (WBC) count on manual microscopy with and without using a hemocytometer, automated WBC enumeration using flow cytometry, and automated WBC enumeration using digital imaging with particle recognition. RESULTS: A total of 4188 children were included. Among febrile children, the sensitivity of the 2 most widely available modalities, the leukocyte esterase test and WBC enumeration using digital imaging, had sensitivity values of 84% (95% confidence interval, 0.80–0.87) and 75% (95% confidence interval, 0.66–0.83), respectively. CONCLUSIONS: Our findings suggest that for febrile children <36 months of age undergoing bladder catheterization for suspected UTI, pyuria will be absent in ~20% of children who are eventually shown to have pure growth of a pathogen on a culture. This raises questions about the appropriateness of requiring pyuria for the diagnosis of UTIs.