Article Text

Download PDFPDF
G425(P) Reliability of urine dipstick and microscopy for diagnosing urinary tract infections (UTIs) in children – an audit
  1. EU Erinaugha1,2,
  2. G Nyamugunduru1
  1. 1Department of Paediatrics, University Hospital of North Durham, Durham, UK
  2. 2Department of Neonatology, Leeds Teaching Hospitals, Leeds, UK


Aim To determine the reliability of urine dipstick and microscopy in diagnosing UTIs in children.

Methods Urine samples were obtained from children admitted with a suspicion of UTI over a period of 30 days. As per local protocol, samples were analysed on the ward by dipstick & phase-contrast microscopy. For the purposes of this audit, all samples were sent to the laboratory for automated cell count and culture. The culture results were compared with automated cell count, dipstick and microscopy results. The data was analysed using SPSS version 15.0 statistical package.

This audit was registered with and approved by the audit department.

Results Ninety-one samples were obtained but 63 contained full results and were analysed. Forty (64%) were from children <5 years old. Seventeen (27%) had pure, significant bacterial growth on culture. Of these, 8 (47%) and 4 (24%) had tested positive for white cells and nitrite on dipstick, respectively. Only 7 (41%) had tested positive for bacteria on ward microscopy and only 1(6%) had >100 white cells on laboratory automated cell count.

Dipstick for white cells had the highest sensitivity of 53% while automated white cell count had the least (5%). Conversely, dipstick for nitrites had the highest specificity of 95% while dipstick for white cells had the least (40%).

Dipstick for nitrites had the highest positive and negative predictive values of 80% and 65%, respectively. Automated white cell count had the lowest positive predictive value of 33% while dipstick for white cells had the least negative predictive value of 53%.

Discussion Urine analysis and microscopy are cheaper but less reliable alternatives to urine culture. For instance, 16(94%) or 10(59%) of the culture positive samples would not have been cultured because they had <100 white cells on automated cell count or no bacteria on ward microscopy, respectively and the diagnosis would have been missed. We conclude that these rapid tests cannot detect all cases of UTI. Therefore, all samples should be sent for cultures. Treatment should be started for all children <1 year with nitrite positive samples and/or symptoms/signs of UTI.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.