Comparing Urine Specimen Collection Methods for Urinalysis and Culture Agreement: A Paired Sample Design.

Abstract

The primary aim of this study was to determine if results from clean catch urine specimens agree with results from catheterized specimens in a urogynecology patient population. The secondary aim was to identify clinical scenarios in which catheterized specimens are preferred over clean catch specimens. Both a midstream clean catch and a catheterized specimen were obtained for each participant. Dipstick urinalysis was performed. If either specimen was positive for nitrites, leukocyte esterase, or blood then both were sent for urine culture.Kappa statistics were calculated to measure agreement between the paired specimen data for the total sample and for stratified samples. We agreed to accept clean catch results as preferable to catheterized results if the κ statistic was 0.7 or greater. Three hundred forty-two participants were enrolled. For all participants, the agreement between the paired samples was strong for nitrite (κ = 0.884), moderate for blood and colony count (both κ = 0.656), weak for culture species (κ = 0.566), and minimal for leukocyte esterase (κ = 0.382). When data were stratified for menopause, vaginal estrogen use, body mass index, and prolapse, there were no clinical scenarios in which the κ values were consistently greater than our accepted value of 0.7. Our data indicate that catheterized urine specimens should be used in the evaluation of urinary tract infection or microscopic hematuria in the typical patient presenting to a urogynecology office who is often menopausal, overweight, and may have prolapse.