R D McNair1, S R MacDonald, S L Dooley, L R Peterson. 1. Departments of Obstetrics and Gynecology and Microbiology, Northwestern University Medical School and Northwestern Memorial Hospital, Chicago, IL, USA.
Abstract
OBJECTIVE: Our purpose was to compare the efficacy of the centrifuged and Gram-stained smear with the efficacy of both urinalysis and reagent strip testing for nitrites and leukocyte esterase in detecting asymptomatic bacteriuria in obstetric patients. STUDY DESIGN: A midstream urine specimen was evaluated in 528 patients either at the initial prenatal visit or at a visit because of possible preterm labor. Separate aliquots were tested by centrifugation (with a Cytospin Cytocentrifuge; Shandon, Inc, Pittsburgh, Pa) with Gram stain, by microscopic urinalysis for the presence of moderate to large numbers of bacteria or >10 leukocytes per high-power field, and by reagent strips for the presence of nitrites or leukocyte esterase activity. Results were compared with those of a quantitative urine culture obtained with blood and MacConkey agar plates. RESULTS: Thirty-six women (6.8%) had urine cultures showing 100,000 colony-forming units of a uropathogen per milliliter. The sensitivity and specificity of testing by centrifugation and Gram stain were 100% and 7.7%, respectively. Urinalysis and dipstick testing offered a sensitivity of 80.6% and 47.2%, respectively, with a specificity of 71.5% and 80.3%. No combination of tests, in series or in parallel, offered improved specificity over urinalysis alone. CONCLUSIONS: Centrifugation with Gram stain of a urine specimen offers excellent sensitivity but very poor specificity compared with microscopic urinalysis for the detection of asymptomatic bacteriuria and is not an acceptable screening test in an obstetric population. The false-negative rates of urinalysis (19.4%) and reagent strip testing (52.8%) preclude these from being excellent screening tests for asymptomatic bacteriuria. Given the potential sequelae of undiagnosed asymptomatic bacteriuria in an obstetric population, we conclude that urine cultures should be used for all pregnant patients to detect asymptomatic bacteriuria.
OBJECTIVE: Our purpose was to compare the efficacy of the centrifuged and Gram-stained smear with the efficacy of both urinalysis and reagent strip testing for nitrites and leukocyte esterase in detecting asymptomatic bacteriuria in obstetric patients. STUDY DESIGN: A midstream urine specimen was evaluated in 528 patients either at the initial prenatal visit or at a visit because of possible preterm labor. Separate aliquots were tested by centrifugation (with a Cytospin Cytocentrifuge; Shandon, Inc, Pittsburgh, Pa) with Gram stain, by microscopic urinalysis for the presence of moderate to large numbers of bacteria or >10 leukocytes per high-power field, and by reagent strips for the presence of nitrites or leukocyte esterase activity. Results were compared with those of a quantitative urine culture obtained with blood and MacConkey agar plates. RESULTS: Thirty-six women (6.8%) had urine cultures showing 100,000 colony-forming units of a uropathogen per milliliter. The sensitivity and specificity of testing by centrifugation and Gram stain were 100% and 7.7%, respectively. Urinalysis and dipstick testing offered a sensitivity of 80.6% and 47.2%, respectively, with a specificity of 71.5% and 80.3%. No combination of tests, in series or in parallel, offered improved specificity over urinalysis alone. CONCLUSIONS: Centrifugation with Gram stain of a urine specimen offers excellent sensitivity but very poor specificity compared with microscopic urinalysis for the detection of asymptomatic bacteriuria and is not an acceptable screening test in an obstetric population. The false-negative rates of urinalysis (19.4%) and reagent strip testing (52.8%) preclude these from being excellent screening tests for asymptomatic bacteriuria. Given the potential sequelae of undiagnosed asymptomatic bacteriuria in an obstetric population, we conclude that urine cultures should be used for all pregnant patients to detect asymptomatic bacteriuria.
Authors: Ann E Stapleton; Marsha E Cox; Robert K DiNello; Mark Geisberg; April Abbott; Pacita L Roberts; Thomas M Hooton Journal: J Clin Microbiol Date: 2015-06-10 Impact factor: 5.948
Authors: Michael Davenport; Kathleen E Mach; Linda M Dairiki Shortliffe; Niaz Banaei; Tza-Huei Wang; Joseph C Liao Journal: Nat Rev Urol Date: 2017-03-01 Impact factor: 14.432