Editor,—Recently dipsticks using nitrates and leucocyte esterase have become available as markers of urinary tract infection (UTI). Leucocyte esterase is an enzyme from neutrophils not normally found in urine and is a marker of pyuria. Nitrates are produced by the bacterial breakdown of dietary nitrates. Most urinary pathogens reduce nitrates to nitrites.1Dipsticks have been extensively tested in adults, but there are few reports on their use as a routine screening test for UTIs in children. In children, the method of urine collection is often variable, and UTIs have far reaching implications.2 This study was conducted to identify which dipstick tests are most accurate for detecting UTIs in routine paediatric practice.

A retrospective study was done of 500 consecutive patients admitted to the children's ward of Hartlepool General Hospital between January and June 1999. All the children admitted to the ward had a dipstick examination using Bayer reagent strips (Bayer, Berkshire, UK) read by an automated colorimeter. Culture of urine was undertaken if the dipstick examination was abnormal. Urine culture was also done routinely before starting antibiotic treatment, if there was any clinical suspicion of UTI, and in children with a history of UTI or renal anomalies, even if the dipstick examination was normal. Dipstick testing was considered abnormal if positive for protein, blood, leucocyte esterase or nitrates. Urine culture was done in 312 (62.4%) children. Of these, the indication for culture was an abnormal urine dipstick testing in 272 (87.2%) cases. In the remaining cases urine was sent for culture even if the dipstick testing was normal, because of the aforementioned criteria.

Urine was collected from pads in infants and young children. Midstream specimens were taken from older children. Urine was sent for culture immediately or stored in a refrigerator if immediate transport was not available. A pure growth of a pathogenic organism with a colony count of > 1 × 10⁹/litre was deemed positive. Mixed growth was immediately repeated and subsequent result was taken for analysis.

Nitrates were found to have a very high specificity (92.4%) for detecting UTI, which is similar to previously reported studies (table 1). However, sensitivity of nitrates was very low even when combined with leucocyte esterase (64.5%). If all the indices (blood, protein, leucocyte esterase, and nitrates) were combined, and the urine sent for culture if any of these were positive, the sensitivity increased to 97.7% and the chance of missing a UTI was very small. Specificity, however, then decreased to 15.4%. The sensitivity of urine microscopy in our study was very low (12.5%). This is probably because of the delay in transport of the urine sample to the laboratory.

Unlike the previously reported studies in adults,3 4 UTIs in children cannot be excluded by a negative dipstick nitrates and leucocyte esterase enzyme reaction. Similar results were observed by Lejeune et al.5 In a study of 243 infants, they reported 97.6% specificity and 16.2% sensitivity for nitrates for detecting UTI. However, when leucocyte esterase, nitrates, and proteins were combined, the sensitivity increased to 89.2% and specificity decreased to 71.8%.

Therefore, urine culture needs to be undertaken if any of the four indices (nitrates, blood, protein or leucocyte esterase) are abnormal, or if there is a clinical suspicion of UTI. If nitrates are positive, starting empirical treatment for UTI seems to be reasonable until cultures are reported. This method helped to reduce the workload of the laboratory on urine cultures by 35.8%. Urinary dipsticks are useful screening tests for detecting UTI, only if their limitations are fully understood.