A febrile infant with no other findings always raises the possibility of a urinary tract infection (UTI) — bacteria inside the urinary tract together with a response from the body. The question that decides what happens next is not whether the infection is possible but how likely it is. That estimate, made before any test has been done, is the pre-test probability, and it determines whether a urine specimen is obtained at all. It varies several-fold with age, sex and circumcision status, so it is worth knowing the figures the estimate is built from.
How often it happens
Two kinds of study supply these figures, and they answer different questions. Population cohorts follow children over years and record every first infection, so they show how many children are ever affected. Cross-sectional series test every child in a defined group, such as febrile infants of a given age, so they show how often a UTI turns out to be the explanation for the illness, which is the figure that matters when a febrile child is in front of you.
- By age 6, an estimated 6.6% of girls and 1.8% of boys have had at least one symptomatic UTI, in a population-based study of 299 children under 6 years. In that study the annual incidence in girls was 9–14 per 1,000 at risk in each of the six 1-year age intervals, while after the first year of life urinary infection of any kind was rare in boys, at 1–2 per 1,000 at risk.
- Fever without a source, up to 24 months — a fever for which history and examination find no explanation. Pooled prevalence of UTI is 7.0% (95% confidence interval [CI] 5.5–8.4), so roughly one febrile infant in fourteen in this age band has a urinary infection as the explanation.
- Uncircumcised boys under 3 months form the highest-risk group of all: 20.1% (95% CI 16.8–23.4) against 2.4% (95% CI 1.4–3.5) in circumcised boys of the same age.
- Among febrile girls, prevalence is highest in the first year of life and falls after 12 months.
- Older children with urinary symptoms. In children over 2 years who have symptoms pointing to the urinary tract, pooled prevalence is 7.8% (95% CI 6.6–8.9) — far below the roughly 50% found in adult women with genitourinary symptoms, which is why urinary symptoms alone cannot make the diagnosis in a child.
Girls are at risk across childhood because their urethra is short, which shortens the distance periurethral and vaginal flora must travel to reach the bladder. Boys have most of their infections, including the febrile ones, in the first year of life, and their risk then falls steeply; the prepuce (foreskin) offers a colonisable mucosal surface for uropathogens, which is why the uncircumcised infant boy sits at the top of the list.
From probability to a decision
Knowing these figures matters because the decision to test follows them. In a survey of pediatricians, most would not culture urine — send it to the laboratory to grow the organism — when the estimated probability of UTI was under 1%, and all would when it exceeded 5%: the probability is what tips the decision. Guidance follows the same logic from the other direction: test the urine of a child whose symptoms and signs increase the likelihood of UTI, do not test routinely when the illness points to another infection, and weigh several features together rather than allowing one symptom to decide.
The arithmetic is a starting point, not a verdict. A well child with a low probability can reasonably be left untested, whereas an unwell febrile infant is sampled and treated regardless of how the numbers fall, because the cost of missing kidney infection and the cost of an unnecessary specimen are not the same. Between those extremes, what moves an individual child’s probability up or down is what the child shows and can report, and that depends heavily on age.
At a glance
- By age 6, about 6.6% of girls and 1.8% of boys have had a symptomatic UTI.
- Fever without a source up to 24 months: 7.0% (95% CI 5.5–8.4) of cases are a UTI; uncircumcised boys under 3 months: 20.1% (95% CI 16.8–23.4).
- Girls carry risk across childhood; boys carry it in the first year of life.
- Urinary symptoms predict UTI far less often in a child than in an adult woman — 7.8% versus roughly 50%.
- Pre-test probability decides whether a specimen is obtained: most surveyed pediatricians would not culture below a probability of 1%, and all would above 5%.