A Cochrane review retained the three community
A 2015 Cochrane review retained the three community-based studies mentioned above and concluded that there was high-quality evidence of reduced risk of infection or omphalitis and death. The review also included two studies in which chlorhexidine was applied in a hospital setting: the above-described German study and a newer neonatal intensive care unit (NICU)-based study conducted in India. On the basis of the results from these studies, the review concluded that there was moderate-quality evidence that chlorhexidine reduces risk of omphalitis or infections in hospital settings (RR 0·48, 95% CI 0·28–0·84). The Indian trial enrolled only 140 participants, patients admitted to NICU of gestational age 32 weeks or older, who on admission were expected to require NICU care for at least 5 days. They received chlorhexidine 2·5% solution three times daily during their NICU stay. In addition to cord separation time, endpoints included umbilical colonisation, neonatal sepsis, culture-proven sepsis, meningitis, and death up to the time of discharge from NICU. The risk of culture-proven sepsis among these hospital births was higher in the dry cord care group than the chlorhexidine group (0·13, 0·01–0·40), although “incidence of probable sepsis and meningitis was … similar [across] groups”. There were also more deaths in the dry cord care group (four of 70 patients) than in the chlorhexidine group (no deaths in 70 patients, p=0·042).
The cited meta-analytic review included only trials conducted in melk pathway settings in developing countries (ie, the three large trials discussed above) and concluded that chlorhexidine reduced risk of omphalitis and newborn death. It did not include any trials in which application was done solely in a hospital setting. However, as mentioned, two of these three trials did include births at home and in facilities; the authors provided a sub-analysis focusing on about 3000 facility births and reported that mortality risk was about half among those facility births receiving chlorhexidine (RR 0·50, 95% CI 0·27–0·92).
Similar to the Nepal and Bangladesh trials, study participants in the two studies in included both home and facility births, and chlorhexidine application was done only in the home. Both studies documented that chlorhexidine reduced risk of omphalitis. In the Pemba trial, relative risk for omphalitis varied from 0·61 to 0·76, depending on level of severity, with p-values for all levels <0·0001. The Zambia trial found a similar effect size (RR 0·73), but with an overall rate of omphalitis of only 0·6% in the control arm there was a relatively wide confidence interval, which included 1·0 (95% CI 0·47–1·13). It is true, as Osrin and Colbourn stated, that these two trials did not show an effect of chlorhexidine in reduced risk of death among those born in hospital, just as no such effect was found for home births. Indeed, no difference was seen in effect sizes when comparing home versus facility births.
For risk of infection to the cord, we see the same picture across all sources of evidence cited by Osrin and Colbourn; whether community-based or hospital-based, home births, or facility births, low-income settings or high-income settings, chlorhexidine reduces risk of such infection. As Sazawal and colleagues conclude, the results of the Pemba study suggest that use of chlorhexidine is justified for its effect in reducing risk of cord infection. Further, in settings where the underlying mortality risk in the population is high, chlorhexidine cleansing reduced mortality regardless of whether babies were born in facilities or at home.
The studies of Russian-backbone live attenuated influenza vaccine (LAIV) in children from Senegal and Bangladesh in are welcome, in view of the high burden of influenza-associated morbidity and mortality in resource-limited settings. Although the lack of immunogenicity data is highlighted as a limitation by the authors, we feel that the importance of generating such data in future studies is underplayed. With the exception of one study including children from South Africa, to our knowledge, no paediatric immunogenicity data exist from sub-Saharan Africa for the Ann Arbor-based LAIV. This amounts to a significant gap in knowledge of how an intranasal LAIV performs immunologically in these settings. Parallels can be drawn with the well-described poorer performance of live oral vaccines in resource-limited settings.