By Philip R. Fischer, MD, DTM&H

Synopsis: Of 4,188 children aged 1 to 36 months screened for urinary tract infection (UTI) in emergency departments, 407 (9.7%) had a culture-positive UTI. Pyuria (by various means of assessment) was not present in 20% of febrile children with a UTI, raising questions about the validity of using pyuria as a necessary component or as a tool for the diagnosis of UTI in young children.

Source: Shaikh N, Campbell EA, Curry C, et al. Accuracy of screening tests for the diagnosis of urinary tract infections in young children. Pediatrics 2024;Nov 20:e2024066600. doi: 10.1542/peds.2024-066600. [Online ahead of print].

---

In young children, the symptoms of urinary tract infections (UTIs) are nonspecific and sometimes vague. The presence vs. absence of pyuria on various screening tests often is used to screen for possible UTI and sometimes is required to diagnose a UTI. Since it is important both to avoid missing UTIs and to avoid over-treating UTIs, Shaikh and colleagues evaluated the sensitivity and specificity of various screening tests of urine that often are used to prompt clinical management decisions for children in emergency departments.

Symptomatic children aged 1 to 36 months who underwent urine testing for possible UTI in three academic pediatric emergency departments were included in this study. Children were excluded from the study if testing or documentation was inadequate (not noted how urine was obtained, for instance), if the child had pneumonia or meningitis, if the child had known underlying urinary pathology, or if the child had an immunodeficiency. Of 7,598 children who had screening tests of urine done, 4,188 were eligible for inclusion in the study.

Of the 4,188 children studied, 72% were female, and 81% had a fever within 24 hours of presentation. A total of 407 (9.7%) had a confirmed UTI (catheterized urine sample with at least 50,000 colony-forming units [CFU] of a single pathogen per mL).

Shaikh and colleagues compared various measures of “pyuria” with culture positivity and, with “hpf” indicating high power field, they found similar results for each test (see Table).

Combining a positive test for pyuria with concomitant bacteriuria, sensitivities were approximately 0.90, but specificities were unacceptably low.

The authors commented that shifting from manual to automated testing has not necessarily improved the usefulness of pyuria to predict positive urine cultures. With further analysis, they also noted that increasing the cutoff from 5/hpf to 10/hpf using an automated digital imaging with particle recognition test (and then, as proposed by some, not doing cultures or providing antibiotic treatment for those children whose urine results were below that threshold) would have missed 35% of actual UTIs.

While the authors still see value in screening for pyuria as part of the data informing clinical decision-making, they see their data as not supporting a policy of using a test for pyuria to determine which children should not have cultures done (and should not have treatment). Based on other data about the prevalence of asymptomatic bacteriuria, the authors believe that a strong majority of their study subjects without pyuria but with positive culture results had true, clinically significant UTIs. Clinicians still are responsible to consider each patient’s history, exam, and initial findings to decide when, pending culture results, empiric antibiotic treatment is warranted.

Commentary

It seems nice to avoid the discomfort of bladder catheterization for children with a low risk of having a UTI, and it would be judicious to avoid unnecessary antibiotic treatment of children at low risk of having a UTI. Unfortunately, as well shown in this new study, withholding accurate microbiologic cultures and antibiotic treatment based on a negative screening test for pyuria would leave 20% of young children with true and clinically significant UTIs undiagnosed and untreated (with risk of progressing to urosepsis and/or renal injury).

Nonetheless, there are some who want to avoid invasive testing and reduce antibiotic treatment based on a screening test for pyuria — reporting “successful” reduction of catheterizations and culture tests by doing initial screening tests (nitrites and leukocyte esterase) on voided urine samples collected in bags on the perineal skin.¹ In one such emergency department study, there was no report of follow-up after the patient left the emergency department.¹ The newer data from Shaikh et al suggest that the bagged urine screening protocol likely prevented or delayed many patients from receiving appropriate testing and antibiotic therapy.

An emergency department in Singapore improved the sensitivity of “pyuria” in identifying UTIs in children (up to 18 years of age) by reducing the cutoff from 100 white blood cells (WBCs) per microliter to 10 WBCs per microliter and combining the pyuria test with a negative test for nitrite in the urine.² In that setting with those criteria, they would have “missed” only 2.2% of true UTIs.² Shaikh et al had similar sensitivity results with a 10 WBC/microliter cutoff when they used a volumetric analysis, but Shaikh et al did not incorporate negative nitrite testing into their evaluation.

Meanwhile, in some countries, clean-catch urine testing is successfully done for infants. Briefly, the genital area is cleaned as for a catheterization, and a parent keeps close watch with a sterile cup handy to obtain a clean-catch mid-stream urine sample right when the infant voids spontaneously. A nationwide study in Sweden revealed that more than 90% of infant UTIs now are diagnosed with these clean-catch samples.³

There also is potential to develop point-of-care urine testing that is more sensitive and more specific than current tests for pyuria and nitrites. In a recent study, urine was tested for acute phase reactants calprotectin and YKL-40 in 67 children aged 1 month to 2 years who had febrile UTIs; their results were compared to similar urine test results in febrile children without UTI and in healthy children.⁴ Urine calprotectin concentrations were approximately 40 times higher in children with UTIs than in control children, and urine YKL-40 concentrations were approximately 100 times higher in children with UTIs than in controls.⁴ Further investigation, perhaps modeling the methods of Shaikh et al, could evaluate the sensitivity and specificity of these urinary biomarkers as screening tools predicting true UTIs.

Philip R. Fischer, MD, DTM&H, is Professor of Pediatrics, Department of Pediatric and Adolescent Medicine, Mayo Clinic, Rochester, MN.

References

1. Paluck F, Kestenbom I, Test G, et al. Decreasing invasive urinary tract infection screening in a pediatric emergency department to improve quality of care. Pediatr Emerg Care 2024;40:
812-817.

2. Teo JN, Teo YT, Ganapathy S, et al. Investigating urinary characteristics and optimal urine white blood cell threshold in paediatric urinary tract infection: A prospective observational study. Ann Acad Med Singap 2024;53:539-550.

3. Lindén M, Rosenblad T, Rosenborg K, et al. Infant urinary tract infection in Sweden — A national study of current diagnostic procedures, imaging and treatment. Pediatr Nephrol 2024;39:3251-3262.

4. Przekora J, Synowiec A, Kubiak JZ, et al. Assessment of urine calprotectin and YKL-40 levels in urinary tract infection diagnosis in children under 2 years of age. Sci Rep 2024;14:28695.