Progress towards the MDGs has been highly uneven. Inequalities in maternal and child health and health care are huge. To make things worse, effective interventions are known, but rarely reach those who need them most. Little is known about what works, and why, to reach lower socio-economic and otherwise vulnerable groups with health interventions. We examined the extent to which our maternal and child health focused intervention has affected women from different socioeconomic and socio-demographic positions. Specifically, we examined attendance at the women’s group meetings, paying particular attention to attendance by poor, less educated and young primigravida women. We also explored barriers and facilitators for attendance among different social strata.

We used population surveillance data from two cluster randomised control trials conducted in Dhanusha and Makwanpur Districts. Trials had tested the impact of participatory women’s groups on neonatal mortality. We completed a site-specific analysis of attendance at women’s groups by socio-economic and socio-demographic position. On the basis of this analysis we purposively sampled women who attended groups, women who did not attend groups, community key informants, family decision makers, and facilitators of women’s groups, to help us explain our quantitative analysis.

From our quantitative analyses, we found that young primigravida women were less likely to attend the women’s groups across both sites. Attendance was similar across lower and higher socio-economic strata, and where there was a difference, we found that attendance was higher among women of mid to low socioeconomic status compared to the better-off women, in particular the socio-economic elite. Caste and ethnicity affected attendance in Dhanusha, with lower participation of the higher status groups. Our qualitative data collection shows that one of the main factors affecting group attendance was the extent of family support that women experienced. Social taboos about the movement of young newly married women outside the home affected their access to groups.

Although direct participation in women’s groups is not the only way to be affected by this community based intervention, it is important to ascertain the accessibility of interventions to the poorest, most marginalised groups. It appears that a participatory intervention with women’s groups is successful in including mid and low socioeconomic status groups in different countries, yet future interventions may be modified to encourage the participation of younger women.