Predicting urine culture results by dipstick testing and phase contrast microscopy.
AIMS: Urine is the most frequently received clinical specimen for bacterial culture. To determine whether dipstick or microscopy results reliably predicted the presence or absence of a reportable urinary pathogen we performed dipstick testing and phase contrast microscopy on unspun urine from 500 specimens. We also investigated the relationship between the presence of squamous epithelial cells (SECs) and mixed growth on culture. METHODS: For each specimen, the presence of leukocyte esterase and urinary nitrite was recorded as well as the quantity of leukocytes (WBC), SECs, and erythrocytes. All cultures had the number and type of each organism recorded. Pyuria was defined as > or =10 WBC/mm(3). RESULTS: Specimens with <10 SEC/mm(3) had fewer isolates (0.9 isolate per culture) than specimens with > or =10 SEC/mm(3) (two isolates per culture); therefore, SEC contamination was defined as > or =10 SEC/mm(3). Of 500 specimens, 266 (53%) had no growth and 136 (27%) had mixed growth. A total of 288 (58%) specimens had negative dipstick results and completely normal microscopy. Of these, 11 (4%) had a pure growth of a urinary pathogen but the pathogen was present in only three (1%) at > or =10(5) CFU/mL. Of 413 urine specimens without SEC contamination, 323 (78%) had either no detectable growth or pure growth compared with only 41 of 87 (47%) with SEC contamination (P<0.001). Of the 413 urine specimens without SEC contamination, 90 (22%) had > or =2 organisms compared with 46 of 87 (53%) with contamination (P<0.01). Pyuria was present in only 21 of 266 (8%) urine specimens without growth but was present in 60 of 95 (64%) specimens containing a reportable pathogen (P<0.01). CONCLUSIONS: SEC contamination does not reliably predict cultures with mixed growth, urine specimens with negative dipstick results and microscopy rarely contain a reportable urinary pathogen, and screening algorithms are warranted and justified.
Smith, P; Morris, A; Reller, LB
Volume / Issue
Start / End Page
Pubmed Central ID
International Standard Serial Number (ISSN)