Use of a DNA probe for the rapid detection of group B streptococci in obstetric patients

To determine the accuracy of a DNA probe as a rapid diagnostic test for detecting colonization of the female genital tract by group B streptococci during pregnancy.

Two rayon-tipped applicators were used to collect secretions from the posterior vaginal wall of 440 pregnant women. One of the applicators was inoculated into selective Todd-Hewitt broth and used as the reference standard for identification of group B streptococci. The other applicator was used for analysis with the DNA probe, preceded by either 2.5 hours of incubation for the initial 75 patients, or 3.5 hours' incubation for the remaining women. Following hybridization with an acridinium-labeled probe, chemiluminescence was measured with a luminometer.

The prevalence of positive cultures was 20%. For the initial 75 patients whose cultures were amplified by incubation for 2.5 hours, the DNA probe had a sensitivity of 44%, specificity 94%, positive predictive value 79%, and negative predictive value 77%. For the cultures that were incubated for 3.5 hours, respective values were 71, 90, 61, and 94%. All vaginal specimens that had an average initial cell count of 1.5 x 10(3) cells/mL were accurately detected by the probe after 3.5 hours' growth amplification. False-positive results occurred primarily when the specimens were grossly contaminated with blood (26 of 39). The mean time required to perform the assay, including 3.5 hours of growth amplification, was 4.3 hours.

The DNA probe demonstrated good overall sensitivity and gave no false-negative results when group B streptococci were present in concentrations of 1 x 10(4) cells/mL or greater. Sensitivity improved significantly with 3.5 hours' growth amplification as compared with 2.5 hours (P < .05), reflecting better identification of lightly colonized patients.