Individual interventions, collective lessons: Developing mid-range theory on women's groups to improve health.

Background: Interventions with women's groups have been widely implemented to improve health outcomes in low- and middle-income settings, particularly India. While there is a large evidence base on the effectiveness of single interventions, it is challenging to predict whether a women's group intervention delivered in one setting can be expected to work in another. Methods: We applied realist principles to develop and refine a mid-range theory on the effectiveness of women's groups interventions, summarised key lessons for implementation, and reflected on the process. We synthesised primary data from several interventions in India, a systematic review, and an analysis of behaviour change techniques. We developed mid-range theories across three areas: maternal and newborn health, nutrition, and violence against women, as well as an overarching mid-range theory on how women's groups can improve health. Results: Our overarching mid-range theory suggested that effective interventions should: build group or community capabilities; focus on health outcomes relevant to group members; and approach health issues modifiable through women's individual or collective actions. We identified four key lessons for future interventions with women's groups, including the importance of skilled and remunerated facilitation, sufficient intensity, supply-side strengthening, and the need to adapt delivery during scale up while maintaining fidelity to intervention theory. Conclusions: Our experience demonstrated the feasibility of developing mid-range theory from a combination of evidence and insights from practice. It also underscored the importance of community engagement and ongoing research to 'thicken' mid-range theories to design effective and scalable women's groups interventions in India and similar settings.

Feasibility, acceptability and equity of a mobile intervention for Upscaling Participatory Action and Videos for Agriculture and Nutrition (m-UPAVAN) in rural Odisha, India.

Addressing undernutrition requires strategies that remove barriers to health for all. We adapted an intervention from the 'UPAVAN' trial to a mobile intervention (m-UPAVAN) during the COVID-19 pandemic in rural Odisha, India. In UPAVAN, women's groups viewed and discussed participatory videos on nutrition-specific and nutrition-sensitive agricultural (NSA) topics. In m-UPAVAN, weekly videos and audios on the same topics were disseminated via WhatsApp and an interactive voice response system. We assessed feasibility, acceptability, and equity of m-UPAVAN using a convergent parallel mixed-methods design. m-UPAVAN ran from Mar-Sept 2021 in 133 UPAVAN villages. In Feb-Mar 2021, we invited 1000 mothers of children aged 0-23 months to participate in a sociodemographic phone survey. Of those, we randomly sampled 200 mothers each month for five months for phone surveys to monitor progress against targets. Feasibility targets were met if >70% received videos/audios and >50% watched/listened at least once. Acceptability targets were met if >75% of those watching/listening liked the videos/audios and <20% opted out of the intervention. We investigated mothers' experiences of the intervention, including preferences for m-UPAVAN versus UPAVAN, using in-person, semi-structured interviews (n = 38). Of the 810 mothers we reached, 666 provided monitoring data at least once. Among these mothers, feasibility and acceptability targets were achieved. m-UPAVAN engaged whole families, which facilitated family-level discussions around promoted practices. Women valued the ability to access m-UPAVAN content on demand. This advantage did not apply to many mothers with limited phone access. Mothers highlighted that the UPAVAN interventions' in-person participatory approaches and longer videos were more conducive to learning and inclusive, and that mobile approaches provide important complementarity. We conclude that mobile NSA interventions are feasible and acceptable, can engage families, and reinforce learning. However, in-person participatory approaches remain essential for improving equity of NSA interventions. Investments are needed in developing and testing hybrid NSA interventions.

'That is because we are alone': A relational qualitative study of socio-spatial inequities in maternal and newborn health programme coverage in rural Uttar Pradesh, India.

This qualitative study was conducted in Uttar Pradesh state, India to explore how interrelated socio-economic position and spatial characteristics of four diverse villages may have influenced equity in coverage of community-based maternal and newborn health (MNH) services. We conducted social mapping and three focus group discussions in each village, among women of lower and higher socio-economic position who recently gave birth, and with community health workers (n = 134). Data were analysed in NVivo 11.0 using thematic framework analysis. The extent of socio-economic hierarchies and spatial disparateness within the village, combined with distance to larger centers, together shaped villages' level of socio-spatial remoteness. Disadvantaged socio-economic groups expressed being more often spatially isolated, with less access to infrastructure, resources or services, which was heightened if the village was physically distant from larger centers. In more socio-spatially remote villages, inequities in coverage of MNH services that disadvantaged lower socio-economic position groups were compounded as these groups more often experienced ASHA vacancies, as well as greater distance to and poorer perceived quality of health services nearest the village. The results inform a conceptual framework of 'socio-spatial remoteness' that can guide public health research and programmes to more comprehensively address health inequities within India and beyond.