Comparative Evaluation of the AccuProbe Group B Streptococcus Culture Test,the BD GeneOhm Strep B Assay,and Culture for Detection of Group B Streptococci in Pregnant Women

We compared a rigorous culture method with the Gen-Probe AccuProbe Group B Streptococcus Culture Test (APGB) and the BD GeneOhm StrepB assay (GOSB) for the detection of group B streptococci (GBS) from an 18- to 24-h LIM broth. Culture (95.3%) and GOSB (95.3%) were more sensitive than APGB (86.5%) for the detection of GBS.Prenatal screening for group B streptococci (GBS) is routinely performed on women at 35 to 37 weeks of gestation (10, 11, 12). We and others have previously presented data supporting the use of the AccuProbe Group B Streptococcus Culture Test (APGB; Gen-Probe, San Diego, CA) for the detection of GBS from LIM broths (3, 13).The advent of PCR-based technology for the detection of GBS offers the potential for increased sensitivity and a shorter turnaround time than culture (2, 4, 11). APGB is a nonamplified nucleic acid hybridization assay, while the BD GeneOhm StrepB assay (GOSB; BD Diagnostics, Ste-Foy, Quebec, Canada) is a real-time PCR nucleic acid amplification assay. Moreover, since the time when we performed our initial evaluation of APGB, improved culture media for the detection of GBS have been developed (5, 6). The purpose of this study was to compare the performance of APGB with a commercially available PCR assay, GOSB, and a rigorous culture method.Vaginal and vaginal/rectal specimens submitted to the Geisinger Medical Center microbiology laboratory for prenatal testing for GBS colonization were included in this study. All of the specimens were collected with Copan Venturi Transystem double swabs and liquid Stuart''s medium (Copan Diagnostics, Marietta, CA).Swabs were first used to inoculate a neomycin-nalidixic acid blood agar plate (NNA; Becton-Dickinson Microbiology, Cockeysville, MD) and then placed into a 3-ml tube of LIM broth (Remel, Lenexa, KS). Following 18 to 24 h of incubation, LIM broths were vortexed and subcultured to an NNA plate. Primary plates and LIM broth subculture plates were incubated at 35°C in a 5 to 10% CO₂ incubator and examined after overnight incubation and a second time 24 h later if the first reading of the plates yielded no GBS. Organisms were identified by the routine bacteriology procedures used by the Geisinger Medical Center microbiology laboratory, including Lancefield identification of GBS with the PathoDX kit (Remel).After 18 to 24 h of incubation, a 50-μl aliquot of each LIM broth was tested with APGB. The directions in the product insert were followed, with one exception. The product insert specifies that when an enrichment broth such as LIM broth is used, the swab should be dipped into the LIM broth and then discarded. We followed the standard procedure that is used in our laboratory for APGB and did not remove the swab from the LIM broth after inoculation.The product insert for GOSB provides instructions for direct testing of swab specimens for GBS. We utilized a protocol that tested a sample from a LIM broth rather than a swab (9). Briefly, 0.2 ml (200 μl) of the LIM broth growth was transferred to sample buffer tubes (blue cap; contains 1 ml of sample buffer) with extended, filtered pipettes. The sample buffer tube stood for 5 min (in a cold block) and was then vortexed at high speed for 15 s. Fifty microliters of each cell suspension was transferred to a corresponding lysis tube (yellow cap), and the cap was closed tightly. The standard GOSB protocol was followed to complete the performance of the assay.For this study, a true positive result was defined as one of the following: GBS from the primary plate culture, GBS from the LIM broth subculture, or GBS from both the primary culture plate and the LIM broth subculture; GBS culture negative but GOSB and APGB positive; GOSB positive, primary culture and APGB negative, but culture positive for GBS following additional culture testing; or APGB positive, primary culture and GOSB negative, but culture positive for GBS following additional culture testing. For any specimen that did not yield the same result with APGB, GOSB, and culture, the result was classified as discrepant. For any specimen with discrepant results, the following protocol was followed: APGB was repeated from the LIM broth; the GOSB was repeated from the reserved lysis tube, as well as from the LIM broth; the LIM broth was subcultured to NNA, and plates were examined for GBS as with the primary culture; and primary plates and initial LIM broth subcultures were reexamined for GBS. Importantly, none of the results were changed based upon the use of the discrepant protocol. The discrepant analysis protocol was used only to correctly classify the initial result. For example, two isolates were initially GOSB positive and culture and APGB negative. Repeat LIM broth subcultures yielded GBS, so these two specimens were classified as GOSB true positive and culture false negative. Statistical analysis was performed as described by Ilstrup by using McNemar''s asymptomatic chi-square test (7).A total of 498 specimens were tested in this study. One hundred twenty-three specimens yielded positive results by all of the methods, APGB, GOSB, primary culture, and/or LIM broth subculture. Three hundred forty-seven specimens were negative by all of the methods. Of the 28 specimens that yielded discrepant results, 25 were determined to be true positives by the above-described criteria. Among these 25 specimens, 3 were culture negative but positive by APGB and GOSB and thus were classified as true positive results. Three specimens were GOSB positive only and were considered false positives. The overall performance characteristics are summarized in Table