Urinalysis is not reliable to detect a urinary tract infection in febrile infants presenting to the ED.

Objective: Urinary tract infections are a common source of serious bacterial infections in febrile infants younger than 2 years. Our objective was to compare urinalysis with urine culture in the emergency department evaluation of febrile infants.

Methods: A febrile infant registry was instituted at a tertiary care hospital treating an average of 55000 patients annually (27% children), from December 2002 to December 2003. Patients were eligible if they were younger than 3 months and had a temperature of at least 38 degrees C or if they were between 3 and 24 months of age and had a temperature of at least 39 degrees C. Data abstracted included age, sex, and temperature. Urinalysis (UA) and urine culture (UCx) results were obtained from electronic hospital archives.

Results: Nine hundred eighty-five patients were entered into the febrile infant registry. Male patients comprised 55%. The mean age of patients was 12.6 months; median was 12 months. Four hundred thirty-five (78% of eligible patients) had both a UA and UCx from the same specimen, and there were 45 (10.3%) positive UCx result. Females accounted for 33 (73%) of 45 positive results. The sensitivity of UA for predicting a positive UCx result was 64% (95% confidence interval [CI], 49%-78%), whereas the specificity was 91% (95% CI, 88%-94%). The positive predictive value was 46% (95% CI, 31%-53%), with a negative predictive value of 96% (95% CI, 93%-97%).

Conclusion: Urinalysis is not reliable for the detection of urinary tract infections in febrile infants when compared with urine cultures.