Press release

Women's support groups improve neonatal survival rates

Women’s community groups have had a dramatic effect on reducing neonatal mortality rates in some of the poorest areas of India, according to a study published today in the ‘Lancet’. The groups provide a cost-effective intervention with added benefits such as reducing significantly maternal depression and improving decision-making among the women.

Every year, an estimated four million children worldwide die within the first month of their lives. Fewer than a quarter of the 68 countries targeted by the Millennium Development Goal 4 (reduction of mortality rate in children less than five years old by two-thirds by 2015) are on track to achieve that goal.

A previous study conducted in Nepal and published by the 'Lancet' in 2004 suggested that participatory women's groups could achieve a significant impact on neonatal health in poorer countries, far more than one-to-one contact with a health worker. To see if these findings could be applied in other countries, the researchers repeated the exercise in Jharkhand and Orissa, two of the poorest states in eastern India. Neonatal mortality rates in the two regions are 49 and 45 per 1,000 live births respectively, disproportionately higher than India's national estimates of 39 per 1,000. By comparison, in the UK the figures are four per 1,000.

Between 2005 and 2008, a team of researchers led by Professor Anthony Costello from the Institute of Child Health, University College London (UCL), and Dr Prasanta Tripathy, from the Indian voluntary organisation Ekjut, evaluated how women's groups affected neonatal mortality and maternal depression in intervention areas as compared to areas where no participatory groups were set up. The groups were evaluated using a cluster-randomised controlled trial, which was funded by the Health Foundation, the UK Department for International Development, the Wellcome Trust and the UK Big Lottery Fund's International funding programme.

The groups were facilitated by women recruited in the local area: non-healthcare professionals with some schooling who tended to be married, respected members of the community. The number of women taking part of in the groups increased from one in six women (17 per cent) of childbearing age in the first year to over a half (55 per cent) in the third year.

The women worked through a 'community action cycle' involving four stages:identifying the problems associated with pregnancy, childbirth and care of newborns; developing strategies to tackle these problems, such as improving hygiene, raising emergency funds and producing their own birthing kits; working with local community leaders, teachers, politicians and others to implement these strategies; and evaluating their success.

"It was crucial that the women were allowed to think thought through the issues and implement their own strategies to tackle them, rather than us telling them what to do," says Dr Nirmala Nair of Ekjut. "We believe that a trained facilitator who supports informed peer-learning is more effective for lasting behaviour change than a traditional instructor/learner approach."

The effects of the interventions were dramatic: by the second and third years of the trial, the neonatal mortality rate in the areas where the participatory women's groups existed had fallen by 45 per cent. These areas also saw a significant fall (57 per cent) in moderate depression among mothers by the third year of the trial.

"What we were seeing was a change in behaviour towards better hygiene practices and improved care for newborns," explains Professor Costello. "There was a move away from harmful practices such as giving birth in unclean environments and delaying breastfeeding. We saw significant improvements in areas such as basic hygiene by birth attendants, clean cord care and women responding earlier to care needs."

The researchers believe that improved social capital - the access the group gave women to a wider support network of peers - was potentially the most valuable aspect of the groups and contributed towards the improved childbirth and childcare practices and the reduction in maternal depression. It may also explain why such groups have had much greater success than direct - even one-to-one - interventions with healthcare workers.

"Many of the women in these groups would have been relatively young, living in arranged marriages with only their mother-in-law or a very limited network of friends for support," explains Dr Audrey Prost from UCL. "The groups empower the women to take preventive measures and to deal with problems more effectively when they arise.If you've been to a group and a problem arises, you've got a ready-made network that you can go to for help and support."

The researchers estimate that the additional cost of introducing support to these groups per newborn life saved was around US$910. However, this raises questions over who would pay for supporting these groups: federal or state government, non-governmental organisations, or a combination of the two.

The 'Lancet' today also publishes a second study carried out by Professor Costello and colleagues using the same approach - women's groups - in Bangladesh. The study failed to reproduce the benefits of the India trial. However, in this case, the researchers believe that there were a number of issues that may have affected the outcome, including a failure to achieve the same coverage of women's groups and recruitment of pregnant women as the India trial.

"Improving maternal and child health is amongst the greatest challenges facing global health workers. Every year, an estimated four million children worldwide die within the first month of their lives and progress in reducing this terrible number is desperately slow," says Sir Mark Walport, Director of the Wellcome Trust, which helped fund the trials. "Women's community groups in India and Nepal set up as part of this study provided many benefits: there was a very significant improvement in neonatal survival and in addition the groups provided women with improved social and support networks. However a parallel study in Bangladesh did not show the same benefits and there is much still to learn."