Low urinary bacterial counts: do they count?

In their article which appears in this issue of Pediatric Nephrology, Dr. S. Swerkersson and co-workers claim that as many as one in every five infants with a true episode of urinary tract infection (UTI) will be missed with the commonly used cut-off level of a parts per thousand yen10(5) colony forming units (CFU)/mL. This controversial finding is supported by the results of seven previous studies including a total of 1587 children. Dr. E.H. Kass, who in the 1950s suggested the presently used cut-off level, knew at the time that it excluded a number of patients with a true infection. Later studies in adult patients also showed that up to 46-49 % of women with a likely diagnosis of cystitis had low bacterial counts. These findings can, if true, improve our understanding of cases of unexplained post-infectious renal scarring. Children with low bacterial counts during an acute infectious episode have a significant risk of receiving delayed or even no antimicrobial treatment. The results of scientific studies are also confounded if 20 % of the subjects with a true infection are wrongly included in a control group diagnosed as having no UTI due to low bacterial counts. This problem cannot be easily solved by lowering the cut-off level and generally accepting that any bacterial count signifies a "true" infection as this approach will drastically reduce the specificity of the culture result. Instead, as many as possible urine samples for culture should be collected from babies, infants and small children with a suprapubic bladder puncture, or a catheterised sample. In such samples bacterial counts as low as 10(3) CFU/mL are generally regarded as significant. In many cases, however, the only possibility is a "clean catch" or bag sample. In these situations, the treating physician needs to take all relevant clinical and laboratory parameters into account and if clinically important data support the diagnosis of a UTI not disregard this diagnosis based only on low bacterial counts. C-reactive protein or procalcitonin can, in a febrile child, help the physician differentiate between a febrile bacterial UTI and a viral infection. A positive nitrite test provides, albeit with a low sensitivity, strong support for a UTI diagnosis.