Each year almost 4 million children die within the first four weeks of life, many from preventable or treatable causes. Much programmatic aid is now devoted to devising ways to ensure that simple effective health practices, such as ensuring a more sterile birth environment, are adopted on a wide scale. A number of recent evaluations from South Asia suggest that the active involvement of local women’s groups in problem solving can be among the most cost-effective interventions to prevent deaths.
One of the earliest studies to investigate the effectiveness of women’s groups , published in 2004 by Dharma Manandhar and co-authors, first created a maternal and child health surveillance system blanketing one district in rural Nepal. Then, through pair-matching, randomly assigned half of the study communities to receive a participatory intervention in the form of women’s groups focused on infant and maternal survival. In each community, a group was formed under the guidance of a trained facilitator which then held a cycle of meetings to discuss issues around child birth and care behaviors.
Apparently, a key aspect of these group meetings was the active roles played by the constituent members where problems and roadblocks to effective maternal health were identified and discussed. The groups proposed solutions, such as community generated funds for maternal care or the local production and distribution of clean delivery kits, which were then put into practice. Interestingly, not everybody participated in these groups – only 8% of women of reproductive age participated although 37% of newly pregnant women joined. From such participation rates, were observable gains possible?
After two years of these meetings, the results from the surveillance data system were striking. Neonatal mortality, at 26 deaths per 1000 births, was 30% lower than the same rate in control communities. Maternal mortality was also reduced by a factor of 5. Corresponding to these survival gains were important gains in healthy behaviors, especially for home deliveries which were still the norm in the treatment communities. For example, women in labor were 4.6 times as likely to use a clean home delivery kit and 3.5 times as likely to use a boiled blade to cut the delivery cord.
Although promising, the Nepal evaluation didn’t assess baseline information on health outcomes. Technically, it was possible that the random assignment, which resulted in slightly richer and more educated communities receiving the intervention, may have in part determined the results observed at endline. However the evidence base continued to grow.
Several years after the Nepal trial, a study in eastern India instituted a similar surveillance system and participatory intervention which was evaluated through a stratified cluster randomized trial . This study, by Prasanta Tripathy and co-authors, collected baseline data to ensure balance between treatment and control areas. Dramatic gains from the women’s participatory groups were also observed. In the second and third years of the program the neonatal mortality rate in intervention areas stood at 37 deaths per 1000 births while in control areas the same rate was 62. Healthy home-care practices around delivery also substantially improved as did measures of maternal depression.
Why might community interventions be particularly adept at improving infant survival? The community groups served an informational role conveying simple and low-cost health practices through participatory learning and problem solving. However the women’s groups served as more than the mere conduit of information.
These community based interventions explicitly attempted to address various social causes of ill health  such as isolation, lack of social support, and low self-esteem along with any informational deficits. Through these groups, networks formed that could be called on during health crises, villages developed obstetric emergency plans, and communities were able engage health professionals in ways that individual patients could not.
So what made these interventions so effective? Was it the information, the empowerment through communal linkages, or the combination of the two? Understanding the causal channels can help devise even more effective interventions as well as help translate these interventions to other settings.
One promising study-in-progress  by Joao Montalvao and co-authors begins to shed light on this question. An extension of the India study, presented at the recent IHEA conference in Toronto , leveraged micro-finance groups to serve as vehicles of health education. Certain study villages already contained existing micro-finance groups while in the remaining intervention villages the groups were newly formed as part of the intervention.
The education activities significantly reduced neonatal mortality in areas that did not have pre-existing microfinance groups. Interestingly these gains occurred largely to babies born to women deemed at baseline to be unempowered in the sense of not having within-household decision making power. At least in some contexts it appears that empowerment is a critical component.
Further unpacking the channels through which empowerment improves health remains an important direction of inquiry and hopefully further analysis in the Montalvao et al. study will shed additional light. Revisiting the Tripathy et al. study to analyze possible heterogeneity in impacts by community characteristics may also be an informative exercise. Other key questions directly relevant to policy adoption include:
- How can these interventions be successfully scaled-up? High-capacity NGOs do not operate in every geographical area of need. Can governments assume the role that has heretofore been played by dedicated maternal and child health NGOs?
- What is the sustainability of these interventions over time? Does there need to always be an outside monitor or are the community groups self-sustaining in the medium and long run?
- Are the lessons from such interventions generalizable beyond the South Asian social and cultural context? One of the first promising pilots took place in the Bolivian highlands  but this small scale pilot was not evaluated with a counterfactual group. There is now an ongoing study in Malawi. 
- More and more of the world’s poor are concentrated in cities. Can these lessons be extended to urban areas where communal ties are typically less strong and formal health care is usually more accessible?