Effects of Community Participation on Improving Uptake of Skilled Care for Maternal and Newborn Health: A Systematic Review

Background Despite a broad consensus that communities should be actively involved in improving their own health, evidence for the effect of community participation on specific health outcomes is limited. We examine the effectiveness of community participation interventions in maternal and newborn health, asking: did participation improve outcomes? We also look at how the impact of community participation has been assessed, particularly through randomised controlled trials, and make recommendations for future research. We highlight the importance of qualitative investigation, suggesting key areas for qualitative data reporting alongside quantitative work. Methods and findings Systematic review of published and ‘grey’ literature from 1990. We searched 11 databases, and followed up secondary references. Main outcome measures were the use of skilled care before/during/after birth and maternal/newborn mortality/morbidity. We included qualitative and quantitative studies from any country, and used a community participation theoretical framework to analyse the data. We found 10 interventions. Community participation had largely positive impacts on maternal/newborn health as part of a package of interventions, although not necessarily on uptake of skilled care. Interventions improving mortality or use of skilled care raised awareness, encouraged dialogue and involved communities in designing solutions–but so did those showing no effect. Discussion There are few high-quality, quantitative studies. We also lack information about why participation interventions do/do not succeed – an area of obvious interest for programme designers. Qualitative investigation can help fill this information gap and should be at the heart of future quantitative research examining participation interventions – in maternal/newborn health, and more widely. This review illustrates the need for qualitative investigation alongside RCTs and other quantitative studies to understand complex interventions in context, describe predicted and unforeseen impacts, assess potential for generalisability, and capture the less easily measurable social/political effects of encouraging participation.

, p=0.000)). No significant differences in: ANC attendance, FP use, or ITN availability. Reports from intervention sites of building laboratories, maternity wings, toilets and water tanks, improved decision-making processes (p. 601).

Brief description of intervention/study
Aim is to build the capacities of communities to take control of mother and child health issues that affect them. In the first arm of the study trained local female facilitators established women's groups and used participatory rural appraisal tools to guide participants through a community action cycle to identify and implement solutions to MNCH problems (p.112). A second arm contained an intervention involving volunteer infant feeding and care counsellors to promote behaviour change. 48 clusters were divided equally into four arms: 1. Women's groups only 2. Counselling only 3. Both women's groups and counselling 4. None-control.
See Rosato et al. Emphasis is on evaluating the impact of the women's group arm only.

Type/level of participation
High level of community participation. Despite being conceptualized externally, within the intervention, women's group members are responsible for identifying local problems with MNCH, designing strategies to address them, mobilising local resources and/or capacities and/or capitalising on partnerships to implement strategies, and then evaluating the strategies.
See Rosato et al.
Examines effects of community participation? Yes, but results not reported in this paper. Yes Pre-existing context Not found in text.
Prior to the intervention, three levels of community sensitisation took place with key stakeholders, introducing them to the aims of the

Malawi (MaiMwana) women's groups
Relational aspects (e.g. building community support, transferring leadership to participants etc.) Intervention has a strong advocacy component, whereby groups are linked to appropriate organisations, locally, regionally, or nationally (p.115).
Intervention encourages the engagement of both women and men, although with the inclusion of women being more prominent.
Men were encouraged to participate in the women's groups, sometimes taking on prominent roles like being the secretary or chairperson. The inclusion of men may have influenced the way women participated, although men are described to have been particularly useful in implementing chosen strategies. When men were present there was an observed tendency for women to talk less and adopt a more 'traditional deferential role ' (p.231) in relation to men.
Symbolic aspects (e.g. Addressing women's status, types of social processes mobilised) Zonal facilitators (who are recruited locally from within their cluster) are trained extensively in community mobilisation techniques, advocacy, and the liaising of groups that can offer appropriate support for strategies. Such skills are transferrable across many issues within and beyond health.
See Rosato et al.
Material aspects (e.g. Efforts to achieve sustained funding/symbolic support, enhancing access to material resources, seeking ways for participants to put their new skills into practice elsewhere) By nature of the intervention, the strategies designed in community mobilisation action cycles to address MNCH were those that could feasibly address the priority problems identified by making best use of local resources (p.114). The project does not provide additional resources to implement strategies; groups must draw on community mobilisation, local capacities, or relationships with government and non-governmental organizations for support (p.115
The findings of the study are principally negative, i.e. the intervention does not have a measured effect on the outcomes of interest. There was no decline in maternal or neonatal mortality associated with the intervention. Odds of having any antenatal care in health facility were higher in the intervention group but of marginal statistical significance(Adj OR 1.44, 95%CI 1.00-2.13). Odds were lower of having a TBA attended birth (Adj OR 0.71, 95%0.51-0.97) but as there is no corresponding increase in births with a skilled provider, it is hard to assess what this might mean in terms of improved maternal safety. (All  statistics from Table 5.13, p.181) When the four arms were analysed separately, there was some evidence that the women's group only arm was associated with a reduction in perinatal mortality in years 2-3 (Adj OR 0.64 (95%CI 0.44-0.94) compared with the control arm) (Table 5.15, p.187) Again, while positive, this is difficult to interpret because there is no effect for all three years combined, and when analysed together with the arm containing women's group and infant feeding combined, there is no significant effect. Assesses the impact of a mix of participatory methodologies and techniques in improving youth reproductive health. The intervention was linked to other programmes, which had been developed and prioritised in a participatory way too.
Yes, particularly with regard to its effect on changing social norms surrounding reproductive health and the attitudes of healthcare providers towards young people interested in reproductive health services.
Pre-existing context Not found in text.
In the study area, women are often made to stay at home and so are unable to access information and health services. Major life and health decisions for women are typically made by husbands and mothers-in-law. The project was a collaboration of local Nepali NGOs and two international organisations (research and service delivery organisations). The intervention in the study site sought to address various barriers to reproductive health (structural, normative, and systemic) linking youth reproductive health programmes with others, and shape the everyday environments of young people (adult education, activities to address social norms, economic livelihood interventions, etc.). The control sites only focused on addressing immediate risk factors such as sexually transmitted infections and unwanted pregnancies. The authors report that in the intervention site social networks for the exchange of information and counseling were strengthened (p. 230). The intervention actively sought to build decision-making structures and coalitions as well as imparting information. The participatory structures of the study site built community skills in: decision-making; consensus building; planning; organising; consulting; negotiating (e.g. with the village development committee); and demanding accountability and resources from various actors (e.g. government funds to continue project activities).
Within the intervention, efforts were made to design strategies to link 'youth reproductive health programs with other programs that were deemed to influence the environment youth lived in' (p.9). Community infrastructure incorporated to coordinate the project, e.g. adolescent coordination teams and parent advisory committees. Youth and adult community members worked together to identify and prioritize interventions. In intervention sites, local service providers were trained, including traditional healers and family and child health volunteers, to address issues of lack of mobility that young women may face when trying to address services outside of their community. Nepal young people Symbolic aspects (e.g. addressing women's status, types of social processes mobilised) In the intervention site there was a focus on 'changing fundamental norms and institutions', which was a key factor in increasing disadvantaged people's demand for services and information (p.231). The authors also report that the participatory approach of the intervention led to 'a new mindset in the communities marked by a deeper, more sophisticated understanding of youth reproductive health and its implications' (p.232) (e.g. an understanding of how family, gender, and social structures and norms limit healthier sexual and reproductive behaviours). Community involvement built via advisory and coordination teams and consultative committees to engage community members (young people and older adults) with an emphasis on disadvantaged groups. Project staff used strategies such as 'rotating representation' to ensure the active participation of 'disempowered groups' (e.g. the poor, women and ethnic minorities) (p. 215). Community task forces created to establish priorities and design interventions.
Behaviour and attitudes of older adults were targeted within the intervention, because of their influential role for young people. Strategies were designed to link young people to 'adult education programs, activities to address social norms, and economic [livelihood] interventions' (p.9). Within the intervention, 'reproductive health outcomes were conceptualized to exist within a context of young people's ideals and aspirations and the broader social institutions and norms that define, shape and constrain life outcomes and choices for young women and men' (p.7). Authors report that through involvement in community-level committees, young people and older adults were able to participate in project decision making. The intervention 'enabled young women to gain confidence and tools to better articulate their own reproductive health needs, positively influenced attitudes of key gate-keepers such as husbands and in-laws about the need for young pregnant women to seek care, and changed provider attitudes and approachability for youth ' (p.21). In order to challenge pre-exisiting social beliefs, men were targeted through separate mobile seminars on sexual and reproductive health.
Material aspects (e.g . efforts to achieve sustained funding/symbolic support, enhancing access to material resources, seeking ways for participants to put their new skills into practice elsewhere) Not found in text. Not found in text. The authors report that the design and implementation of the intervention in the study sites were more 'inclusive, and interactive, with a great deal of attention to building community ownership and involvement at every step.' (p.214). Action planning conducted at the intervention design stage was participatory. The results of the needs assessments exercise were 'shared and analyzed with the community, and community task forces were created to set priorities and design feasible interventions' (p.215). In the study sites, adults and young people were allowed to increase their 'authority and decision-making power in the project' through community-level committees (p. 215). The other programmes linked to intervention in the study sites were developed and prioritized by community members.

Characteristics
Youth participation was a key strategy for community mobilization [the study] engaged young people in needs assessment, program design, implementation, and assessment' (p.6). Local structures (including NGOs and community member-led organizations, etc.) were used to implement programme activities.
External support? Not reported in text.
External NGOs (EngenderHealth), academic institutions (the International Centre for Research on Women), and local NGOs (New ERA and BP Memorial) collaborated to implement the project. Funding sources not reported.
Monitoring and evaluation done in participatory way?
The authors report that 'participatory techniques' were employed in monitoring and evaluation through 'advisory and coordination teams and consultative committees that engaged youth and adult community members, especially those who were disadvantaged. ' (p.214-215) Young people were involved in assessment of the programmes as part of the intervention design. The overall evaluation of the programme's effectiveness in achieving outcomes was conducted by the authors. See Malholtra et al.

Sustainability
The authors report that the intervention led to enhanced understanding of how different factors (family, gender, social structures and norms) constrain healthier reproductive and sexual heath, which they argue is an indication of the sustainability of the demand for youth reproductive health services.
The reliance on locally-established organizations allowed for 'ownership and sustainability' (p.29). However, a threat to sustainability is that the existing health infrastructure will likely not be able to keep up with increased demand for youth reproductive health services (p.29). Authors claim that the study was aimed at examining whether participatory interventions can lead to increased empowerment of and accountability to poor and disadvantaged people, and improve health outcomes and service accessibility for this population. They also argue that participatory interventions are more likely to have a positive impact on young people's critical thinking and decision-making capacities (p.213).
Not reported in text.

Equity considered (PROGRESS)
Study targeted youth reproductive health because young people disadvantaged when tying to access reproductive services and information, particularly in Nepal. Authors report that the rural-urban difference in the selection of the intervention and control sites was aimed at taking into account structural and wealth differentials and claim that the intervention reduced inequities.
Not reported in text.
Cost considerations Not reported in text. Not reported in text.

Study design
Moderate risk of bias. Intervention in two study sites (one urban and one rural) compared with two control sites (urban and rural). 'Communities selected were randomly assigned to study or control' (p.214). Study site interventions aimed to address structural, normative, and systemic barriers to youth reproductive health and were developed and prioritized by community members. Control sites focused only on the most immediate risk factors (e.g. sexually transmitted infections and unwanted pregnancies). Control sites did not employ participatory or collaborative methods (were developed and implemented by project staff) and received reproductive health interventions through adolescent-friendly services, peer education and counselling, and teacher education. Data collected through household survey at baseline and endline and analysed as a cross-sectional sample, not a longitudinal sample, as participants were not followed throughout the study. Qualitative data also collected. For prenatal care and institutional childbirth information, young unmarried women of the study areas (14-21-year-olds at baseline and 18-25-year-olds at endline) were compared. Also see Mathur et al.
Quantitative data were collected through household baseline and endline surveys conducted in 1999 and 2003 respectively. At baseline, 373 young people were interviewed and at endline, 359 were interviewed, of whom 84 and 81, respectively, were married. Qualitative data were also collected from young married and unmarried men and women, adults, and health service providers through focus group discussions, in-depth interviews, key informant interviews, community mapping, making lifelines, body mapping, creating reproductive health problem trees, and reproductive health service matrices. Also see Malholtra et al.

Participants
14-21-year-olds at baseline, 14-25-year-olds at endline, living in the study sites: two rural sites in Nawalparasi district and Kawasoti district and two urban sites in middleclass suburbs of Kathmandu.
Targets of the intervention were youth aged 14 -21, but participants within programming included young married women, their partners, older men and women, and health service providers.

Outcomes
Prenatal care in first pregnancy of unmarried women (whether she visited a trained provider, i.e. doctor, nurse, trained clinician, for prenatal care in her first pregnancy at least once); and institutional childbirth (whether the birth, miscarriage or abortion resulting from first pregnancy was at a medical facility). Other outcome measures not relevant to this review also reported. Statistical tests for significance of these differences not presented. Authors claim intervention successful in reducing advantage-based differentials in youth reproductive health outcomes (e.g. urban-rural differences in pregnant women accessing prenatal care declined in the intervention groups pre-and post-study from OR: 16.4 (p=0.001) to 1.2 (p=0.644). In the control group the decline was much smaller, from OR: 3.7 (p=0.028) to 3.2 (p=0.021). Similarly, for institutional births, the urban-rural difference declined in the intervention group from OR: 15.6 to 4.6 (p=0.000, 0.002) but increased in control group OR: 13.5 to 21.3 (p=0.000)). The authors draw on qualitative data to explain some of the factors that shaped the success of the participatory approach: '(1) facilitating the co-production of services; (2) empowering youths and adults and increasing the accountability of service providers and policy makers to the communit; and (3) increasing community demand for information and services. ' (p.230) Overall, the participatory intervention was 'only marginally more effective in changing women's knowledge and awareness of maternal care issues, but was more effective in increasing awareness about where maternal care services could be found, and changing practices around use of services' (p.12). Some raw data are provided on outcome measures (e.g. Tables 3-5), but no statistical analysis is presented.

Characteristics
Manandhar, D. S., D. Osrin, et al. (2004). "Effect of a participatory intervention with women' groups on birth outcomes in Nepal: cluster-randomised controlled trial." Lancet 364 (9438) Brief description of intervention/study Intervention arm: Facilitated women's groups first discussed issues around childbirth and care behaviours in the community and then developed, implemented, and assessed strategies, which included community-generated funds for maternal or infant care, stretcher schemes, production and distribution of clean childbirth kits, home visits by group members to newly pregnant mothers, and awareness-raising with a locally made film to create a forum for discussion. Also health-service strengthening activities. Control arm: health-service strengthening activities alone.
Employed trained local female facilitators to organise, generate interest in, and facilitate women's group meetings in each intervention site. Supervisors (nationally advertised) were also employed (1 supervisor to support 3 facilitators). 111 women's groups were established. Initial meetings focused on 'problem identification' through facilitated discussion of the reasons (social and medical) for neonatal and maternal deaths in the community, leading to groups identifying three priority problems of neonates and/or pregnancy. Following problem prioritisation, groups progressed to 'planning together' strategies for responding to these problems, and later, through a community meeting, the wider community was invited to this planning process. Before the intervention, the authors held meetings with the Makwanpur District Development Committee, the Chief District Officer, and local stakeholders. The study clusters were village development committees. One of the reasons given: representatives functioned as key points of liaison. Community leaders helped identify potential female facilitators. Word of mouth within the community was also used. Support for the intervention from other community members was planned together and sought by the women's groups. Partnership with the District Public Health Office to train female community health volunteers, TBAs, and government health staff in newborn care. There were security problems (third year of study), which meant that some women's groups had to postpone meetings. Women's group strategies involved interaction outside the group, which the authors say increased awareness of perinatal health-related issues.
Linkages established with community leaders, nongovernmental organisations, and district health services. Authors report that 'close' alliances with community leaders and community health workers helped implementation (p.1). Before implementation, the authors sought but did not find local NGOs or community based organisations that 'routinely' worked in all the study areas; these organisations had 'different agendas' (p.3). Community support ('awareness and interest') for the intervention was sought by group facilitators and supervisors (p.3). Support from the community sought through community meetings planned and organised by women's groups to 'legitimize' their work (p.5), and community (including leaders) involved in planning strategies. Lack of continuity for some groups was reportedly: lack of support from local leaders, husbands, and health workers, and an unstable security environment. When local health personel and chairmen participated, community meetings to plan strategies together, discussions 'were livelier and planning more productive' (p.6). Local female community health volunteers (lowest cadre of government-appointed health staff) and TBAs were actively involved in the women's groups, allowing the volunteers to have contact with a user group and conduct their health education work. See also Manandhar et al. (2004). Use of action-learning cycles in women's groups to: prompt social dialogue about local knowledge and beliefs; identify and prioritise problems; and encourage women to formulate strategies. The intervention was not about 'teaching' and transferring health-related information.
Authors emphasise the idea that the facilitator should act as a catalyzing agent for change and 'broker of information' rather than as a teacher or instructor. Women attending the group had an active role.
The intervention sought a social process of 'learning together' about perinatal problems, which involved dialogue within the groups and with other community members outside the groups (women were encouraged to talk with neighbours and friends).
Not found in text. See Manandhar et al. 2004 andMorrison et al. 2005.
Material aspects (e.g. efforts to achieve sustained funding/symbolic support, enhancing access to material resources, seeking ways for participants to put their new skills into practice elsewhere) Women were not given money to participate in the women's group activities (sustainability reasons given by sponsors). Women developed strategies which capitalised on local resources (e.g. The development of a local emergency fund, stretcher schemes, production of clean childbirth kits, etc.).
Women's group female facilitators were paid a salary 'slightly higher than the government equivalent' (p.3). Women were encouraged to find strategies (to identified problems) which would use local resources. 23 months after the first health fund, women's groups had generated between 10.5 and 133.8 US dollars (731-9635 rupees). The profits made from the locally produced home childbirth kits went to the mother and child health fund.
See Manandhar et al. 2004and Morrison et al. 2005. Characteristics Manandhar, D. S., D. Osrin, et al. (2004. "Effect of a participatory intervention with women' groups on birth outcomes in Nepal: cluster-randomised controlled trial." Lancet 364 (9438) The intervention was introduced by agents external to the community, but target women identified, prioritised and implemented strategies.

Trial intervention introduced by external agents (MIRA).
Group facilitators selected through interviews by MIRA employees. Supervisors were also recruited through formal interviews but not clear from the text whether MIRA employees were also involved. Although female facilitators were trained to use a meeting facilitation manual adapted from the Warmi project in Bolivia, they were also 'allowed scope for their own input' (p.3). Both women's groups and community members participated in the planning together stage of finding strategies for identified problems. The authors report that 'it is highly likely that the facilitation team's attempts to adequately support the facilitators may have led to less participatory processes taking place, especially in the case of strategy development.' (p.9). The manual was developed as a 'reference guide' but was regarded by the facilitation team as 'an essential resource' and 'an instruction booklet'.  See Manandhar et al. 2004. Paper reports a sub-group analysis of women who experienced pregnancy before the intervention and who had another pregnancy during the study period.

Data collection
Baseline data collected from each potential member of the cohort from March -July 2001. Baseline service audit also completed. Menstrual status of individual participants recorded monthly by local female enumerator and structured interviews with women determined to be pregnant were undertaken at 7 months gestation and at 1 month postpartum.

Participant observation by technical advisors (anthropologists)
were reported in their monthly reports, which were discussed with and added to by the facilitation team (facilitators, supervisors and senior facilitation manager).
Baseline data collected via structured questionnaire interview of all eligible participants (women who had been pregnant) to identify perinatal practices. At endline, participants who had another pregnancy during study period interviewed with same questionnaire to identify changes in perinatal care practices. 5,400 women eligible for inclusion.
Participants 14,844 married women of reproductive age in the intervention clusters. The 111 women's groups that emerged in the intervention.
5,400 women who experienced pregnancy before the intervention and who had another pregnancy during the study period.

Outcomes
Primary: neonatal mortality rate. Secondary: stillbirths and maternal deaths, uptake of antenatal and childbirth services, home-care practices at childbirth and postpartum, infant morbidity, and health-care seeking Analysed and described the development of women's groups, considering factors characteristic of successful groups, the group meetings, and the strategies developed.
Primary Outcomes: attendance at antenatal care; use of a boiled blade to cut the cord; appropriate dressing of the cord; and not discarding colostrum.
Of the 111 women's groups, 77 developed and designed strategies to address identified problems and 100 continued to meet to discuss perinatal health. The most popular and successfully implemented strategies were: the mother and child health fund, locally produced clean home childbirth kits, management and production of stretchers, and awarenessraising through video shows. Gaps in local knowledge were identified during the planning stages which led to a process of participatory learning being initiated. Authors discuss factors influencing success of groups, reporting that they found 'no specific formula', however, support of husbands, local political organisations, and local health staff seemed important. Authors also discussed some of the challenges emerging from the participatory process, in particular the balance between being directive, particularly regarding health knowledge and education, vs participatory.
Improvements among women not following good practice at baseline in intervention group vs. control in all four indicators. Women not following good practice at baseline were more likely to show a positive improvement if they lived in an intervention area or received the intervention compared with the control. Also, evidence of a greater improvement in attending antenatal care and not discarding the colostrum for women attending groups compared with. Improved attendance at antenatal care: BETTER/WORSE ratio: 1.77 (95% CI: 1.30, 2.40). Improvement in not discarding the colostrum:. BETTER/WORSE ratio 1.03 (95% CI: 1.01, 1.06).
India: Ekjut. Jharkhand and Orissa, women's groups Relational aspects (e.g. building community support, transferring leadership to participants etc.) Authors met and sought permission for the intervention and surveillance from village councils, headmen, and representatives from panchayats (elected representatives for basic governance). Other community members who were not regular attendants were also encouraged to participate in the women's group discussions. Also, in all clusters (control and intervention), health committees formed to express opinions about the design and management of local health services.
Critical consciousness' among women's group members enabled them to spread awareness to the wider community, 'as evidenced in group members' support to local village health committees and their involvement of community health workers in discussions about entitlements to health services' (p.9). The presence of frontline government staff, ASHAs, Anganwadi workers, and auxillary nurse midwives at meetings meant that different community members, healthcare workers, and decision-makers have an increased awareness of maternal and neonatal health initiatives in their communities (p.9). The presence of auxillary nurse midwives and Anganwadi workers at meetings reportedly increased their accountabilty to community members.
Symbolic aspects (e.g. addressing women's status, types of social processes mobilised) At women's meetings participants discussed the difficulties mothers encounter in the community and devised strategies to collectively address them. The authors report that women improved problem-solving skills through participating in the meetings. The authors call for the need to evaluate the outcomes of delivering the intervention in partnership with government and NGOs.
The intervention actively targeted marginalized and pregnant women.
Material aspects (e.g. efforts to achieve sustained funding/symbolic support, enhancing access to material resources, seeking ways for participants to put their new skills into practice elsewhere) Facilitators received seven-day residential training and fortnightly support from district coordinators. Knowledge about clean childbirth practices and care-seeking enabled local women to address maternal and neonatal health.
Cluster-level meetings allowed women's groups to share problems and strategies with the broader community. As a result, community members, including men, offered to help implement the groups' strategies (p.9). A considerable percentage (over 37%) of home births are attended by family members. These births are likely managed 'using information and skills from the meetings' (p.10). Participation in groups and cluster meetings allowed for participants to become health advocates in their communities and homes, providing assistance to pregnant women (e.g. members recalled providing assistance to 3822 pregnant women during the study period) (p.10).
Rath, S., N. Nair, et al. (2010). "Explaining the impact of a women's group led community mobilisation intervention on maternal and newborn health outcomes: The Ekjut trial process evaluation." BMC International Health and Human Rights 10 (1) Women in intervention clusters were poorer and more disadvantaged than those in control clusters.
Ekujut targeted areas predominantly inhabited by tribal people with no or few land holdings, low literacy, and many living below the poverty line. Groups also succesfully targeted pregnant women. See also Tripathy et al.

Cost considerations
Authors report that the incremental cost of women's group intervention (2007 prices Characteristics Tripathy, P., N. Nair, et al. (2010). "Effect of a participatory intervention with women's groups on birth outcomes and maternal depression in Jharkhand and Orissa, India: a cluster-randomised controlled trial." The Lancet 375 (9721): 1182-1192.
Rath, S., N. Nair, et al. (2010). "Explaining the impact of a women's group led community mobilisation intervention on maternal and newborn health outcomes: The Ekjut trial process evaluation." BMC International Health and Human Rights 10(1): 1-13.

Study design
Low risk of bias. Cluster-randomised controlled trial. 18 clusters (estimated population of 228 186) from a total of 36 clusters in Jharkhand and Orissa were assigned to the intervention or control arms using stratified randomisation. A total of 244 women's groups emerged in the intervention clusters. In intervention clusters, a facilitator met with 12 -14 groups every month to go through the participatory action cycles and participatory learning. Intent to treat analysis.
This study describes the context, content, and implementation of the Jharkhand and Orissa women's groups intervention. For intervention study design, see Tripathy et al.

Data collection
Surveillance by key informants in each of the three districts (from the Jharkhand and Orissa states where the study took place) identified all births and maternal and newborn deaths within their jurisdiction. Monthly, the key informant met with an interviewer who verified the births/deaths. Births were followed up approximately six weeks later with a structured questionnaire about sociodemographic characteristics, pregnancy, childbirth, and the postnatal period. Deaths were followed-up with verbal autopsies.
Data were collected through the review of intervention documents, qualitative structured discussions with women's group and non-group members, meeting observation, and statistical analysis of program records.

Participants
Women aged 15-49, residing in the project area, and having given birth during the study. See Tripathy et al.

Outcomes
Primary: Neonatal mortality rate and maternal depression scores. Secondary: stillbirths; maternal and perinatal deaths; uptake of antenatal care and childbirth services; home care practices during and after childbirth; and healthcare-seeking behaviour.
Qualitative analysis.
Rath, S., N. Nair, et al. (2010). "Explaining the impact of a women's group led community mobilisation intervention on maternal and newborn health outcomes: The Ekjut trial process evaluation." BMC International Health and Human Rights 10 (1) Six key factors influenced implementation and impact: acceptability among communities; use of a participatory approach to develop knowledge, skills, and a 'critical consciousness'; community involvement beyond the women's groups; a focus on marginalized communities; active recruitment of newly pregnant women; and high population coverage (p.1,12). The authors suggest that the participatory approach was key to its success. The use of stories by the women was entirely participatory and played an important role in educating participants and helping local facilitators familiarize themselves with preventative strategies for common problems (p.8). The authors also suggest community mobilisation may have been the catalyst for developing 'critical consciousness' among group members and the wider community (p.9) UNICEF (2008). Maternal and Perinatal Death Inquiry and Response: Empowering communities to avert maternal deaths in India. New Delhi, UNICEF.
Material aspects (e.g. efforts to achieve sustained funding/symbolic support, enhancing access to material resources, seeking ways for participants to put their new skills into practice elsewhere) Mobilisation of funds and restructuring of health systems to improve access to services. (E.g. In response to MAPEDIR maternal death audit data, where local leadership was directly engaged, local village council funds have been used for emergency referral/transport systems (p.41/42)). Authors report that in West Bengal, state action was effectively stimulated and the system strengthened, resulting in efforts being made to improve health infrastructure including "free for all" maternity beds and a cashless (voucher-based) referral and transport system (p.43).

Leadership, planning and management
Intervention was entirely led, planned, and managed externally, although the sharing of audit data findings was sometimes taken up by local NGOs. Design of local interventions, when it occurred, relied on community organizations for planning and implementation.
External support? Several institutions and groups involved: the Government of India, State Governments, District Administrations, medical faculties of Indian universities, the Johns Hopkins Bloomberg School of Public Health, WHO, UNFPA and UNICEF (p.5). Monitoring and evaluation done in participatory way?
No. Data compiled and analysed at the district-level with technical assistance from UNICEF.

Duration of intervention Varied in different locations. Sustainability
Efforts were made to involve local administration in the process to ensuring scale-up and sustainability (p.18) Theoretical Framework Used 'three delays' framework to structure enquiry. Equity considered (PROGRESS) MAPEDIR exposed inequities in maternal death by documenting causes related to poverty, low status of woman who died, etc.
Cost considerations Not found in text.

Study design
Narrative report of implementation of MAPEDIR in different locations: series of primarily qualitative case studies Data collection Data from the maternal death audits was collected and analysed continuously throughout the intervention. Collection was by trained local auxillary nurse midwives (ANMs), ANM supervisors, lady health visitors, NGO members, and Integrated Child Development Services supervisors (p.16).

Participants Communities in different areas of India
Outcomes Awareness of causes of maternal death that could be tackled at community level. 'The ultimate goal of MAPEDIR is to reduce maternal deaths through communities themselves taking remedial measures to address gaps at the local level, and officials taking corrective measures in the health system' (p.25).