Explaining the impact of a women's group led community mobilisation intervention on maternal and newborn health outcomes: the Ekjut trial process evaluation

摘要: Few large and rigorous evaluations of participatory interventions systematically describe their context implementation, or attempt to explain the mechanisms behind impact. This study reports process evaluation data from Ekjut cluster-randomised controlled trial a learning action cycle with women's groups improve maternal newborn health outcomes in Jharkhand Orissa, eastern India (2005-2008). The demonstrated 45% reduction neonatal mortality last two years intervention, largely driven by improvements safe practices for home deliveries. A 244 was implemented 18 intervention clusters covering an estimated population 114 141. We context, content, implementation this identify potential its impact, report challenges experienced field. Methods included review documents, qualitative structured discussions group members non-group members, meeting observations, as well descriptive statistical analysis on attendance, activities, characteristics attendees. Six broad, interrelated factors influenced intervention's impact: (1) acceptability; (2) approach development knowledge, skills 'critical consciousness'; (3) community involvement beyond groups; (4) focus marginalized communities; (5) active recruitment newly pregnant women into (6) high coverage. hypothesize that these were responsible increase delivery care led trial. Participatory can influence child if key are preserved tailored local contexts. Scaling-up such requires detailed understanding way which affects acceptability intervention; planned but flexible replication content features; strong support methods implementing agencies.