J. Gigante et al., UNIVERSAL SCREENING FOR GROUP-B STREPTOCOCCUS - RECOMMENDATIONS AND OBSTETRICIANS PRACTICE DECISIONS, Obstetrics and gynecology, 85(3), 1995, pp. 440-443
Objective: To determine how obstetricians' opinions regarding universa
l screening of pregnant women for group B streptococcus (GBS) and thei
r responses to positive culture results vary from American Academy of
Pediatrics recommendations, and to determine the physician characteris
tics that predict divergent opinions. Methods: One hundred ninety-four
practicing obstetricians in the middle Tennessee region were queried
by a mail survey. They were asked if they agreed with universal screen
ing for GBS and to indicate whether they would prescribe antibiotics f
or women in labor, represented by six scenarios that differed with res
pect to presence or absence of preterm labor, premature rupture of mem
branes (ROM), prolonged ROM, and a positive GBS cervical culture. They
were also asked to describe their practice and personality characteri
stics. Results: Completed surveys were returned by 135 of 194 obstetri
cians (70%). Although only 28% of the respondents agreed with routine
prenatal screening for GBS, most (74%) said they would treat a patient
on the basis of a positive culture alone. Other risk factors, when ad
ded to a positive culture, slightly increased the decision to treat (f
rom 74 to 88%). Multiple logistic regression, used to assess the relat
ive effect of clinical and physician characteristics on treatment deci
sions, revealed that chemoprophylaxis for GBS was predicted most stron
gly by a positive culture at 28 weeks' gestation followed by prolonged
ROM and preterm labor. Practicing in an urban location and seeing few
er than 20 patients per day also influenced the decision to treat. Con
clusion: Although most obstetricians in the middle Tennessee region do
not believe in universal screening, most will prescribe intrapartum a
ntibiotics on the basis of a positive screening culture. However, othe
r clinical risk factors and physician characteristics significantly an
d independently affect the decision to treat as well.