Community Participation in Health Systems Research: A Systematic Review Assessing the State of Research, the Nature of Interventions Involved and the Features of Engagement with Communities

"Community participation can be instrumental as working with communities can help make interventions more relevant to local needs, informed by local knowledge and priorities, and therefore more effective. More fundamentally, depending on the social processes involved, it can also be transformative, helping to empower and emancipate marginalized communities. At the same time, community mobilization without attention to power relations can distort participation from its developmental aims, exacerbate existing patterns of exclusion and further entrench inequities."

Community participation is a major principle of people-centered health systems, with considerable research highlighting its intrinsic value and strategic importance. Existing reviews largely focus on the effectiveness of community participation, with less attention to how community participation is supported in health systems intervention research. Thus, this article intends to explore the extent, nature, and quality of community participation in health systems intervention research in low- and middle-income countries (LMICs). It is not the intention of the review to provide a comprehensive catalogue of the literature on community participation; rather, the aim is to review how published health systems research, as one aspect of the health systems policy and research community, is engaging with community participation. The findings are one input towards further understanding and supporting community participation as a part of strengthening health systems research and interventions at community level.

The researchers searched for peer-reviewed, English language literature published between January 2000 and May 2012 through 4 electronic databases. Search terms combined the concepts of community, capability/participation, health systems research, and low- and middle-income countries. The initial search yielded 3,092 articles, of which 260 articles with more than nominal community participation were identified and included. They further excluded 104 articles due to lower levels of community participation across the research cycle and poor description of the process of community participation. Out of the remaining 160 articles with rich community participation, they further examined 64 articles focused on service delivery and governance within health systems research.

Most articles were led by authors in high-income countries. (Despite the focus on LMIC countries, more than half of first authors were based in high-income countries (58%, 150/260). Among LMIC articles, almost a half focused on sub-Saharan Africa (45%, 117/260), even though just under a fifth were authored by those based in sub-Saharan Africa (19%, 50/260).) While there was a mix of purely qualitative, quantitative or mixed methods studies in our review, most were explanatory in nature, with very few using probability or experimental designs to assess health effects. The quality of studies, irrespective of their study designs, was variable. The extent of community participation in health systems research interventions was assessed across 5 different elements, depending on whether communities were involved in: (i) identifying and defining the problems addressed; (ii) identifying and defining the interventions developed to address those problems; (iii) implementing interventions; (iv) managing resources for the interventions; and/or (v) monitoring and evaluating interventions. To be included in this review, articles needed to have community participation in at least one of these five elements. Of those articles that had some degree of community participation in the health system intervention under investigation, almost all detailed community participation in implementing interventions (95%, 247/260). Very few were involved in the strategic decisions that framed the research by identifying and defining the problems that needed to be addressed (18%, 46/260), although just over half were involved in identifying and defining interventions (50%, 131/260). Fewer articles detailed community participation in terms of managing resources (31%, 80/260) or monitoring and evaluating (24%, 63/260). Only a minority involved communities in 3 steps (55/260, 21%) or in 4 steps (12%, 31/260), with only 4 involving communities in all 5 steps. Articles were most likely to describe community participation in health promotion interventions (78%, 202/260), even though they were less participatory than other health systems areas. Community involvement in governance and supply chain management was less common (12%, 30/260 and 9%, 24/260, respectively), but more participatory. Information systems was the health system domain with the fewest articles 3% (7/260) and also the least participatory 29% (2/7). These interventions included studies on community-based surveillance systems for specific health conditions or treatment programmes, community engagement in social audits or quality improvement, or the challenges of engaging community participation in information systems in post-disaster situations. In most of these articles, coordination, management and action on data was undertaken by local government bodies or non-governmental organisations (NGOs) and not directly by community members themselves.

Many articles did not discuss who in communities participated, with just over a half of the articles disaggregating any information by sex. Almost a third, 32% (50/156), of articles reached out to women or men or focused on sex-specific health conditions. A large number of these articles focused on sexual, reproductive, maternal and child health issues, primarily focusing on women as beneficiaries. Only one article reached out to men in participatory way by supporting fathers' clubs to promote child health. Of the 106 articles with rich community participation that did not reach out to women or men or their sex-specific health issues, 28% (30/106) did not discuss gender in any way, 35% (37/106) mentioned gender in passing (one or two sentences), and only 28% (30/106) discussed gender issues substantively.

Articles cut across all health conditions and varied by scale and duration, with those that were implemented at national scale or over more than 5 years being mainstreamed by government. Of the articles with rich community participation, the authors did further analysis on the 64 that involved service delivery and governance in health systems research. These findings are explored in detail. In general, most articles detailed improvements in service availability, accessibility, and acceptability, with fewer efforts focused on quality, and few designs able to measure impact on health outcomes. Articles were largely under-theorised, with only 16 of the 64 articles related to governance and service delivery providing a definition or framework of community participation. Only 5 of the 64 mentioned "power" or "control", which is notable because, per the researchers, "a defining element in assessing community participation is the level of control or power that communities command in an initiative". Only 4 mentioned Freire, despite his seminal role in advocating for community empowerment and critical consciousness.

The majority of the articles (57/64) described community participation processes as being collaborative, with fewer describing either community mobilisation or community empowerment. About half of the articles that described community mobilisation in service delivery and governance interventions focused on raising community awareness around a specific health issue through drama presentations, video clips, education sessions, picture cards etc. In addition, a few interventions trained community volunteers to raise awareness and conduct health promotion through household visits or counseling. Project stakeholders also conducted community level meetings to sensitise communities and enlist their support for the intervention. Yet, only a couple of articles described feedback sessions with communities, mostly about the results of baseline surveys stressing the importance of an impending health problem. Overall, many articles did not detail who was in charge of the participatory interventions or who set the agenda-defining interventions. In almost half of the articles that we classified as describing community empowerment, communities became skilled in: identifying and prioritising problems; devising action plans; and implementing, monitoring, and evaluating the plans. This was a skill common to women's group interventions, where women learnt problem solving techniques through participatory learning and action cycles.

Facilitating and challenging aspects of community participation were documented by all the articles to some degree. At the individual level, community members are motivated to engage for a variety of intrinsic reasons, although financial support cannot be ignored for long-term continuity. Community-level support and trust was indispensable and facilitated when communities' perceived interventions to fit with their needs and adapted to their context. External linkages, particularly to the health care system, were seen as important and at times inhibited by broader factors. Institutional processes that facilitated trust, transparency, and communication were central. Contextual factors supportive of community participation included legal reforms or policies leading to decentralisation inclusive of communities, as well as specific political histories and engagement with social movements. Others have also emphasised multiple layers of factors that influence community participation, most particularly the importance of trust, acceptance, and long-term partnerships. Challenges included lack of training, interest, and information, as well as weak financial sustainability.

The researchers conclude that, despite positive examples, community participation in health systems interventions was variable, with few being truly community directed. "Whether this is appropriate or not depends on the context, and most importantly what communities themselves want with regards to their role in health systems interventions. Maximum participation without delegation of resources or democratization of power, may marginalize those communities and members that can least afford to participate." Future research should more thoroughly engage with community participation theory, recognise the power relations inherent in community participation, and be more realistic as to how much communities can participate and cognisant of who decides that. "Understanding, negotiating and contesting power remains a foundation to be laid with health systems researchers, health systems interventions and societies."