0947 Failure of chlorhexidine prophylaxis to prevent neonatal sepsis

An excellent example of how the scientific process should be followed is contained in a study published by Cutland et al in a recent issue of the Lancet. Given the current global focus on the Millenium Development Goals and recent evidence that South Africa has, if anything, regressed in terms of achieving targets in the areas of infant and child mortality, it is entirely appropriate that researchers investigate ways in which neonatal mortality might be reduced. Globally it is estimated that some 900000 sepsis-associated neonatal deaths occur annually in developing countries. Whereas in developed countries strategies such as targeted prophylaxis with intrapartum antibiotics coincided with a 70% reduction in early-onset group B streptococcal disease, such interventions are unaffordable in developing countries, and the focus has therefore been on ‘lower-tech’ strategies. Along these lines there have been two non-randomised Africa studies (in Egypt and Malawi) using chlorhexidine wipes to clean the birth canal and baby, with apparently-positive results (50-75% reduction in neonatal and maternal sepsis-associated morbidity and mortality), however conclusions drawn from reviews of the studies were that a randomized controlled study should be done in a developing country before proposing the strategy as policy. Such a study was carried out at the Chris Hani Baragwanath Hospital by the NRF/MRC unit. 8011 mothers were assigned to chlorhexidine vaginal wipes or external genitalia sterile water wipes, and their 8129 infants were assigned to full-body or foot-only chlorhexidine wipes. A background rate of early-onset sepsis of 30/1000 livebirths was estimated during a pre-study audit. Results at the conclusion of the study showed 289 neonates with sepsis, 141 (3%) in the study group vs 148 (4%) controls. In addition, 5826 women were assigned to a group B streptococcus colonization sub-study (i.e. to assess maternal carriage and neonatal colonization). In this sub-study there was similar exposure to chlorhexidine vs control prophylaxis. Colonisation of mothers was found in 20 and 21.8% in the two groups, with 54% and 54.5% transmission to neonates. These results suggest that the chlorhexidine strategy is ineffective and should not be promoted as policy, however an editorial in the same journal issue discounts the study and urges further studies, essentially on the grounds that the environment was not third-world enough in comparison to the other African studies.

Read more:
Lancet 2009; 374: 1909-16 and pp1873-5
Pediatr Infect Dis J 2006; 676-79
Cochrane Database Syst Rev 2004; 3: CD003520