Study: Automated urinalysis still misses up to 1 in 5 children with UTIs

Key takeaways:

• Automation has not improved the accuracy of UTI screening tests, researchers found.
• Choosing not to order urine cultures due to absence of pyuria could leave children with untreated UTIs.

One in five febrile infants and toddlers with UTIs were missed by automated screening methods because they did not have pyuria, according to a study published in Pediatrics.

Now that many dipstick and urinalysis tests have been automated, Nader Shaikh, MD, MPH, professor of pediatrics and clinical and translational science at University of Pittsburgh School of Medicine and Children’s Hospital of Pittsburgh, and colleagues wanted to see if automation made them more accurate.

“We focus on tests assessing pyuria because the last guideline from the AAP (now retired) suggested that a positive test result for pyuria should be required in all cases in which UTI is diagnosed,” Shaikh and colleagues wrote. “This requirement has led many centers to send urine cultures only for children with positive test results for pyuria.”

In 2011, the AAP published guidelines requiring the presence of pyuria and at least 50,000 colony-forming units of a single pathogen to diagnose a UTI in children aged 2 to 24 months.

Subsequent studies reported that many children may be left with untreated UTIs due to this requirement, and the organization retired these guidelines in 2021. WikiGuidelines released updated recommendations for UTI diagnosis and treatment in November, which cautioned providers not to use urinalysis alone to determine whether to order a urine culture.

Shaikh and colleagues analyzed data on 4,188 children aged 1 to 35 months (72% girls; mean age, 12.2 months; standard deviation, 9 months), who underwent catheterization for a suspected UTI at Children’s Hospital of Pittsburgh, Children’s National Medical Center and Cincinnati Children’s Hospital EDs between June 2019 and April 2023.

The researchers compared the sensitivity and specificity of five types of tests using previously recommended cutoffs, as well as a lower cutoff of 10,000 or more CFU/mL and the allowance of a second pathogen.

Out of 4,188 children, 3,377 (81%) had a fever, and 407 (9.7%) had positive urine cultures, Shaikh and colleagues found. They calculated the sensitivity for five modalities among all children, including leukocyte esterase on a dipstick (0.81; 95% CI, 0.78-0.85); white blood cell count on a microscope with (0.82; 95% CI, 0.75-0.9) and without a hemocytometer (0.78; 95% CI, 0.66-0.89); and automated WBC enumeration with flow cytometry (0.88; 95% CI, 0.82-0.93) and digital imaging with particle recognition (0.76; 95% CI, 0.69-0.84). They found similar results with different cutoffs and pathogens, as well.

“Although pyuria combined with bacteriuria (ie, pyuria or bacteriuria) had a slightly higher sensitivity (ie, > 90% for 4 of the 5 modalities examined), the specificity of these combinations was unacceptably low in many instances,” the authors wrote.

The researchers noted that pyuria was less common with pathogens other than Escherichia coli.

They reported that 1 in 5 children with fever did not have pyuria, but did have a positive urine culture. Additionally, Shaikh and colleagues found that digital imaging with particle recognition, which is the most widely available modality, missed 35% of febrile children with positive urine cultures.

“It appears that automation has not improved the accuracy of the available screening tests for UTI,” the researchers wrote. “The increasingly commonplace practice of ‘reflexing’ cultures based on the absence of pyuria alone using this modality is particularly prone to missing children with positive culture results.”

References:

• Nelson Z, et al. JAMA Netw Op. 2024;doi:10.1001/jamanetworkopen.2024.44495.
• Roberts KB. Pediatrics. 2011;doi:10.1542/peds.2011-1330.
• Shaikh N, et al. Pediatrics. 2024;doi:10.1542/peds.2024-066600.