Background: Voided urinalysis to test for urinary tract infection (UTI) is prone to false-positive results for a number of reasons. Specimens are often collected at triage from women with any abdominal complaint, creating a low UTI prevalence population. Improper collection technique by the patient may affect the result. At least four indices, if positive, can indicate UTI.
Objective: We examine the impact of voided specimen collection technique on urinalysis indicators of UTI and on urine culture contamination in disease-free women.
Methods: In this crossover design, 40 menstrual-age female emergency department staff without UTI symptoms collected urine two ways: directly in a cup ("non-clean") and midstream clean catch ("ideal"). Samples underwent standard automated urinalysis and culture. Urinalysis indices and culture contamination were compared.
Results: The proportion of abnormal results from samples collected by "non-clean" vs. "ideal" technique, respectively, were: leukocyte esterase (>trace) 50%, 35% (95% confidence interval for difference -6% to 36%); nitrites (any) 2.5%, 2.5% (difference -2.5 to 2.5%); white blood cells (>5/high-powered field [HPF]) 50%, 27.5% (difference 4 to 41%); bacteria (any/HPF) 77.5%, 62.5%, (difference -7 to 37%); epithelial cells (>few) 65%, 30% (difference 13 to 56%); culture contamination (>1000 colony-forming units of commensal or >2 species) 77%, 63% (difference -5 to 35%). No urinalysis index was positively correlated with culture contamination.
Conclusion: Contemporary automated urinalysis indices were often abnormal in a disease-free population of women, even using ideal collection technique. In clinical practice, such false-positive results could lead to false-positive UTI diagnosis. Only urine nitrite showed a high specificity. Culture contamination was common regardless of collection technique and was not predicted by urinalysis results.
Keywords: urinalysis; urinary tract infection.
Copyright © 2015 Elsevier Inc. All rights reserved.