Objective : Group B Streptococcus (GBS) and Escherichia coli (E. coli) are the leading causes of early-onset neonatal sepsis. Intrapartum antibiotic prophylaxis decreases vertical transmission of GBS, but testing women during labor is limited by the late availability of results. Antenatal culture screening for GBS has been shown to have relatively low sensitivity in some settings. To allow testing women during labor, we developed a real-time PCR assay for the simultaneous screening of GBS and neonatal invasive strains of E. coli.

Study design : Specific DNA targets for GBS are publicly available (cfb). For invasive E. coli, we identified candidate DNA targets by DNA hybridization on microarrays using invasive strains isolated from neonatal E. coli sepsis or meningitis (K1 and not K1 serotypes). Specificity of DNA probes was tested against a panel of bacteria and simulating clinical conditions (spiking vaginal samples from pregnant women). Then, the characteristics of the selected probes were evaluated in a pilot study including 200 women in labor.

Results : Prevalence of rectovaginal GBS and vaginal E. coli K1 colonization by culture were 16.0% and 3.5%, respectively. Prevalence of rectovaginal GBS, and vaginal invasive E. coli and E. coli K1 colonization by PCR were 27%, 20% and 10%, respectively. The turn-around time needed for PCR analysis was 2 hours. Compared to the culture, considered as the gold standard, the sensitivities of the PCRs for the GBS and E. coli were 97% and 71%, respectively. Specificities were 86% and 92%, respectively. Specificity is difficult to interpret, as a “false” positive PCR result may actually be a false negative result of the culture.

Conclusion : The rapid PCR assay for GBS screening is reliable. There is a need for further development of the PCR test for E coli. Whether bacteria only detected by PCR are virulent needs further investigation.