Background: Current prevention of infection due to group B Streptococcus (GBS) involves giving intrapartum antibiotics to women on the basis of either antenatal culture colonization status or presence of risk factors.
Methods: We prospectively compared the performance characteristics of a rapid molecular diagnostic test (IDI-Strep B; Infectio Diagnostic) with culture for intrapartum GBS detection after 36 weeks' gestation in 5 North American centers during the period September 2001-May 2002. Antenatal GBS screening was done according to the usual practice of participating hospitals. Two combined vaginal/anal specimens were obtained from participants during labor by use of standard techniques and processed by the same laboratories that processed the antenatal specimens. Each swab sample was processed simultaneously by culture and with IDI-Strep B. The collected specimens were randomized for order of testing of the swab samples by culture or the rapid test.
Results: Of enrolled women, 803 (91.1%) were eligible for analysis. The overall intrapartum GBS colonization rate by culture was 18.6% (range, 9.1%-28.7%). Compared with intrapartum culture, the molecular test had a sensitivity of 94.0% (range, 90.1%-97.8%), specificity of 95.9% (range, 94.3%-97.4%), positive predictive value of 83.8% (range, 78.2%-89.4%), and negative predictive value of 98.6% (range, 97.7%-99.5%). The molecular test was superior to antenatal cultures (sensitivity, 94% vs. 54%; P<.0001) and prediction of intrapartum status on the basis of risk factors (sensitivity, 94% vs. 42%; P<.0001).
Conclusion: Use of this test for determination of GBS colonization during labor is highly sensitive and specific and may lead to a further reduction in rates of neonatal GBS disease.