Aims Urinary tract infections account for approximately 7% of emergency department attendances. Untreated infections can have serious consequences. Point of care testing has advantages over laboratory urine testing in that manpower requirements, cost, and patient time spent in the emergency department are all lower. However, urinalysis is not as sensitive or specific as microscopy for UTI. We sought to investigate whether point of care microscopy of uncentrifuged urine is a useful adjunct to urinalysis in accurately diagnosing UTI in children in the emergency department.
Method The study was conducted in the emergency department of an Irish, tertiary paediatric hospital during October and November 2013. The prospective arm of the study involved recording the Multistix urinalysis, and microscopy results, for all clean catch and mid-stream urines examined over a four day period. Microscopy results were correlated with corresponding laboratory results where available. The corresponding retrospective arm compared the point of care urinalysis and microscopy findings with laboratory results for all children diagnosed with UTI, based on point of care (POC) results, over the preceding month.
Results Prospective arm: over a seven day period 102 urine samples were examined. 23 (22%) were then sent for laboratory analysis. Of these, 12 (52%) of POC results over-estimated the laboratory white cell count, 30% underestimated, and the 8% were the same. In the nine samples sent to the laboratory, where only one of nitrate or leucocyte esterase (LE) were positive, only 3/9 (33%) of POC microscopy led to UTI treatment when culture results turned out to be negative. Retrospective arm: 100 children were diagnosed with UTI during the study period. Of the 52 complete datasets were available for analysis, 33 (63%) had significant bacterial growth. 81% (25/31) of nitrate positive, and 65% (31/48) of leucocyte esterase positive samples demonstrated significant bacterial growth in the laboratory, rising to 82% (23/28) if both were positive. There was a 0.64 correlation between ED and laboratory microscopy.
Conclusion Although this was a small study, it shows that point of care microscopy may be useful in predicting UTI, especially where Multistix leucocyte esterase and/or nitrite results are negative.