A clinical prediction tool demonstrated strong diagnostic performance in identifying urinary tract infections (UTIs) in young children with fever, according to a study published in JAMA Network Open.


Researchers prospectively validated UTICalc version 3.0 among 2,561 children aged 2 to 24 months presenting with fever at two tertiary paediatric emergency departments in Canada between November 2022 and January 2025. Nearly two-thirds of participants were younger than 12 months, and 47% were female. Follow-up confirmed UTIs in 111 children, representing 4% of the study population.


Diagnosing UTIs in preverbal and pre-toilet-trained children presents particular challenges due to non-specific symptoms and difficulty obtaining clean urine samples. The UTICalc clinical model uses five risk factors: age under 12 months, sex and circumcision status, temperature above 39°C, absence of another fever source, and fever duration exceeding 48 hours. This model achieved an area under the receiver operating characteristic curve (AUROC) of 84.1%.


When dipstick results for leukocyte esterase and nitrite were added to the clinical factors, the model’s AUROC increased to 95.3%, with a sensitivity of 94.0% and a specificity of 86.9% at a 5% risk threshold.


Emergency department clinicians showed sensitivity of 98.2% and specificity of 57.3%. Whilst UTICalc did not outperform experienced clinicians, corresponding author Dr Ceilidh Kinlin of the Children’s Hospital of Eastern Ontario noted that it offers a useful, evidence-based adjunct for guiding urine testing decisions.


The researchers concluded that these findings support using UTICalc to guide UTI evaluation and management in young children, particularly as the prospective study design allows for generalisability of results.


Source: Physicians Weekly / JAMA Network Open (2026)