Ever since group B streptococcus (GBS) emerged as the commonest cause of perinatal sepsis in the late 1970s, there has been controversy about prevention strategies. A few hospitals in Australia were among the first in the world to introduce routine antenatal screening for GBS carriage and intrapartum antibiotic prophylaxis for carriers.1 This approach was later vindicated by randomised controlled trials in selected carriers2 and the demonstration of lower rates of sepsis after intrapartum prophylaxis compared with historical rates.3 However, problems remain. Group B streptococcus is a normal vaginal commensal in healthy women, but colonisation is often intermittent, and rates of colonisation can vary from 18% to 27%, depending on the detection method.4 Moreover, vaginal carriage is a very crude predictor of perinatal sepsis, with fewer than 1% of the infants of carriers affected (1–2/1000 overall) without intervention.1,5

In 1996, the Centers for Disease Control and Prevention (CDC) in the United States published consensus guidelines for selecting women for intrapartum antibiotic prophylaxis using either of two alternative strategies. One strategy was based on maternal GBS carriage, and the other on clinical risk factors — preterm labour (< 37 weeks' gestation), prolonged rupture of membranes (> 18 hours) or intrapartum fever (> 38oC).6 The rationale for the latter strategy was that, before widespread use of intrapartum antibiotics, one or more of these risk factors was found in up to 80% of mothers of infants with GBS sepsis.1,7 Gradual implementation of these consensus guidelines in the US was associated with a fall in the incidence of perinatal GBS sepsis from 1.7/1000 in 1992 to 0.5/1000 in 1999.8 In Australia, there was also a decrease in the incidence of perinatal GBS sepsis, from 1.2/1000 in 1991–1993, when three of nine neonatal units surveyed had prevention strategies in place, to 0.5/1000 in 1995–1997, when all 11 units surveyed had prevention strategies.5 A recent review concluded that there is evidence, albeit from relatively poor-quality trials, that intrapartum prophylaxis reduces the incidence of neonatal sepsis, but not deaths.9

Despite this evidence, concern continues about excessive use of intrapartum antibiotics. There have been several reports that their increasing use is associated with an increased proportion of cases of neonatal sepsis caused by penicillin-resistant bacteria.10,11 Although it is sometimes difficult to prove, there is considerable empirical evidence that increased antibiotic use generally leads to increasing bacterial resistance. It is plausible that exposure to antibiotics in utero might delay colonisation of the infant gut with penicillin-sensitive anaerobes and allow penicillin-resistant facultative bacteria — many of which are potential pathogens — to become established.

Recently, the CDC published revised guidelines, recommending a single strategy for prevention based on universal prenatal screening for vaginal or rectal GBS colonisation.12 The recommendation was based on a retrospective cohort study, which showed that perinatal GBS sepsis was significantly less frequent in infants of women given intrapartum antibiotics on the basis of documented GBS screening results (0.33/1000 births) than in infants of women managed on the basis of risk factors (0.59/1000 births; relative risk, 0.48; 95% CI, 0.37–0.63).13 This result is not surprising. A risk factor-based protocol cannot, by definition, prevent sepsis in infants whose mothers have no risk factors. On the other hand, the proportion of cases prevented by a protocol based on GBS colonisation depends on the sensitivity of the screening method, effectiveness of prophylaxis and compliance with the protocol.4,14

What was surprising in the CDC study was that the anticipated overall rate of intrapartum antibiotic use was similar for both prevention strategies (31% and 29%).13 In contrast, we showed that in Australia a strategy based on risk factors would lead to significantly less use of intrapartum antibiotics (18%–20% of women) than a strategy based on antenatal screening at 35–37 weeks' gestation (35%).4 This difference is apparently due to a higher incidence of risk factors in the US compared with Australia, and failure to account for women given intrapartum antibiotics during preterm labour before results of screening are available. They suggest that obstetricians in Australia should not immediately discard the option of a risk-based strategy.

Neither strategy is ideal, but either, if properly implemented, can reduce the incidence of perinatal GBS sepsis. The GBS screening strategy results in at least a third of healthy young women (and their infants) being given intravenous antibiotics during labour, at significant cost and with some risks, but can achieve a lower rate of perinatal GBS sepsis.13 In Australia, with a risk-based strategy, significantly fewer women and infants would receive intravenous antibiotics. Whichever strategy is chosen, the most important determinant of its effectiveness will be compliance.