OBJECTIVE: Group B streptococcus (GBS, Streptococcus agalactiae) is an important cause of neonatal sepsis. Prevention is possible by intrapartum screening for maternal GBS carriership and antimicrobial treatment of colonized women with risk factors during labor. The conflicting results of diagnostic performance are reported both for the newly developed rapid GBS antigen tests and Gram's stain. METHODS: The value of Gram's stain in GBS screening was investigated prospectively in 1,020 women. Intrapartum Gram's stains of the cervix from these women and of the introitus from 510 of them were compared with cultures of the cervix, introitus, and anorectum in a semiquantitative way. RESULTS: The sensitivities of the cervical and introital Gram's stains were 25% and 31%, respectively, and the specificities 99% and 98%, respectively. Higher sensitivities (52% and 44%, respectively) were found in heavily colonized parturients. No significant influence of rupture of the membranes was detected. There was a poor correlation between the number of gram-positive cocci in the Gram's stain and the growth density. CONCLUSIONS: We do not recommend the routine use of the Gram's stain for intrapartum GBS detection because of both the limited sensitivity and positive predictive value.
OBJECTIVE:Group B streptococcus (GBS, Streptococcus agalactiae) is an important cause of neonatal sepsis. Prevention is possible by intrapartum screening for maternal GBS carriership and antimicrobial treatment of colonized women with risk factors during labor. The conflicting results of diagnostic performance are reported both for the newly developed rapid GBS antigen tests and Gram's stain. METHODS: The value of Gram's stain in GBS screening was investigated prospectively in 1,020 women. Intrapartum Gram's stains of the cervix from these women and of the introitus from 510 of them were compared with cultures of the cervix, introitus, and anorectum in a semiquantitative way. RESULTS: The sensitivities of the cervical and introital Gram's stains were 25% and 31%, respectively, and the specificities 99% and 98%, respectively. Higher sensitivities (52% and 44%, respectively) were found in heavily colonized parturients. No significant influence of rupture of the membranes was detected. There was a poor correlation between the number of gram-positive cocci in the Gram's stain and the growth density. CONCLUSIONS: We do not recommend the routine use of the Gram's stain for intrapartum GBS detection because of both the limited sensitivity and positive predictive value.