Intravaginal Washes with Chlorhexidine May Not Reduce Neonatal Sepsis in Low-Resource Settings

Lisa Melhado, Guttmacher Institute

First published online:

The use of chlorhexidine wipes during labor does not prevent early-onset neonatal sepsis or vertical transmission of group B streptococcus, according to a randomized, controlled trial conducted in South Africa.1 No differences in the rates of neonatal sepsis (3–4%) or mother-to-child transmission of group B streptococcus colonization (54– 55%) were observed between the intervention and control groups. These results contrast with those of earlier African studies in which the use of chlorhexidine was associatedwith significant reductions in neonatal and maternal sepsis and neonatal mortality. In a comment on the study, Mullany and Biggar suggest that contextual differences between the current study and the earlier onesmay account for the conflicting results, and that this simple, low-cost intervention should not be abandoned because of the new findings.2

Intravaginal washes with chlorhexidine have been thought to reduce sepsis by preventing newborns from acquiring vaginal bacteria during childbirth. Although nonrandomized studies have found intravaginal chlorhexidine washes efficacious in the reduction of sepsis-associated morbidity and mortality, the lack of demonstrated results from a randomized, controlled trial has hindered worldwide acceptance of the approach. If effective, such a cheap and accessible intervention would be a particular boon to developing nations, where the intrapartum anti- biotic prophylaxis that has greatly reduced early-onset group B streptococcal disease in developed nations is typically unavailable.

Between 2004 and 2007, pregnant women presenting at the Chris Hani-Baragwanath Hospital, an urban public facility in Soweto that provides free maternal and pediatric care to low- to middle-income patients, were screened for eligibility to participate in the trial. Women were excluded if they planned to have a caesarean section, had genital warts or ulcers, had had antepartum hemorrhage or an intrauterine death, had a known allergy to chlorhexidine, were fully dilated, were younger than 15, or were carrying a fetus that had a severe congenital malformation or was presenting face first. In all, 8,011 women (and 8,129 newborns) were randomly assigned to either the chlorhexidine group or the control group.

During labor, the cervix, vaginal walls and external genitalia of women in the intervention group were wiped with cotton pads soaked in 0.5% chlorhexidine gluconate solution; for women in the control group, external genitalia were wiped with autoclaved tap water. After they had been bathed, newborns assigned to the intervention were wiped with the chlorhexidine solution from head to toe (excluding the face and ears); newborns in the control group received a chlorhexidine foot wipe. In addition, for a subset of both groups, swabs were taken from mothers and newborns to determine if vertical transmission of group B streptococcus had occurred. Vertical transmission was defined as the isolation of the same bacteria from both the mother and child, and early-onset sepsis was defined as sepsis occurring within the first three days of life. Analysis was by intention to treat; in the intervention group, all of the mothers and 99% of infants were wiped with chlorhexidine.

Participant characteristics were similar across study groups: The women had a mean age of 26 years, 10% received intrapartum antibiotics and 26% were HIV-positive; the median gestational age of the infants was 39 weeks. Early-onset sepsis occurred in 3% of newborns in the intervention group and 4% of those in the control group. Among infants born to women with group B streptococcus, vertical transmission was found in 54% of those in the chlorhexidine group and 55% of those in the control group. Neither difference was statistically significant. The neonatal mortality rate was lower in the intervention group than in the control group (8.3 per 1,000 births vs. 12.8 per 1,000); however, most of the difference was due to deaths that occurred soon after delivery and hence were probably not related to sepsis.

On the basis of their results, the authors conclude that "chlorhexidine is unlikely to prevent vertically acquired neonatal infections in any setting or population." The lack of an effect on vertical transmission, they add, provides further evidence that the difference between groups in the infant mortality rate was unrelated to the intervention.

In their comment, Mullany and Biggar suggest that the authors of earlier chlorhexidine studies in Malawi and Egypt may have been mistaken when they attributed the reductions in sepsis in those studies to the prevention of vertical transmission; instead, chlorhexidine may have reduced morbidity and mortality by preventing environmental or hospital-acquired infections. The authors of the South African study seem to support this view, noting that the "results of a review of pathogens associated with neonatal sepsis in developing countries suggest that many neonatal infections might be acquired environmentally." Unlike the earlier studies, the current study took place in a hospital "in which professional care and facilities were superior, antibiotic use was extensive, aseptic conditions were maintained and overall rates of mortality and sepsis were low"; Mullany and Biggar note that the lack of such conditions in previous studies "might explain why chlorhexidine antisepsis was reported as efficacious." Given this possibility, "further studies of the role of chlorhexidine use in low-resource delivery rooms and community settings" continue to be warranted.

—L. Melhado


1. Cutland CL et al., Chlorhexidine maternal-vaginal and neonate body wipes in sepsis and vertical transmission of pathogenic bacteria in South Africa: a randomized, controlled trial, Lancet, 2009, 374(9705): 1909–1916.

2. Mullany LC and Biggar RJ, Vaginal and neonatal skin cleansing with chlorhexidine, Lancet, 2009, 374(9705): 1873–1875.