Evaluation of the sensitivity of a rapid polymerase chain reaction for detection of group B streptococcus

The objective of this study was to compare detection of group B streptococcal (GBS) carriage using 'real-time' polymerase chain reaction (PCR) and microbiological standard culture. The study design was a test accuracy study comparing a novel molecular technique against the standard microbiological cultural technique in normal pregnant women. The setting and population consisted of 143 pregnant women with pre-labour rupture of the membranes, recruited from two large teaching hospitals in the UK. The study examined the efficacy of a polymerase chain reaction (PCR) assay for screening pregnant women who presented with term rupture of the membranes. Low vaginal specimens were obtained from the women. The specimens were tested for GBS by conventional culture and with a GBS-specific real-time PCR assay. The main outcome measure was the sensitivity and specificity of the PCR assay with 95% confidence intervals (CI) compared with the standard culture. The length of time to obtain a result was also reported for both methods. Among the 143 women, the results of the culture were positive (at least one colony) for GBS in 20 women (14%). The PCR assay detected GBS carriage in 10 women (7%). As compared with the culture method, the sensitivity and specificity of the PCR assay were 45% and 99%, respectively. The positive and negative predictive values of the PCR assay were 90% and 92%, respectively. The length of time required to obtain results for the majority of women (94%) was <2.5 h for the PCR assay and at least 24 h for culture. While a rapid result (within 3 h) of carriage of GBS can be obtained by the PCR assay, at present, it cannot replace conventional culture without further optimisation of the DNA extraction method. The sensitivity may further be improved by testing both low vaginal and rectal specimens.