A mapping and synthesis of tools for stakeholder and community engagement in quality improvement initiatives for reproductive, maternal, newborn, child and adolescent health

Abstract Background Stakeholder and community engagement promotes collaboration and gives service users an opportunity to actively participate in the care they receive. Recognizing this potential, The Network for Improving Quality of Care for maternal, newborn and child health aimed to identify tools and operational guidance to integrate stakeholder and community engagement into quality improvement (QI) implementation. Methods A mapping, consisting of a literature review and an open call through email and listservers, for implementation tools was conducted. Materials were included if they provided guidance on stakeholder and community engagement aligned to the Network's QI framework comprising seven phases. Screening of tools was done by two reviewers. Results The literature search and the call for tools returned 197 documents with 70 tools included after screening. Most included tools (70%) were published after 2010. International organizations were the most frequently cited authors of tools. Only 15 tools covered all seven phases of the QI framework; few tools covered the more ‘technical’ phase of the QI framework: adapting standards and refining strategies. Conclusion The quantity of tools and their varied characteristics including types of stakeholder and community engagement processes across the QI framework confirms that engagement cannot be captured in a ‘one‐size‐fits‐all’ formula. Many tools were designed with a generic focus to allow for adaption and use in different settings and sectors. Country programmes looking to strengthen engagement approaches can take advantage of available tools through an online portal on the WHO website and adapt them to meet their specific needs and context. Public Involvement Programme implementers provided tools and resources during data collection.


| INTRODUC TI ON
Stakeholder and community engagement is widely advocated for within health programming and is of paramount importance to achieving universal health coverage 1,2 and primary health care for all. 3,4 Defined as a process of involving communities in decisionmaking and in the planning, design, governance and delivery of service, 5 stakeholder and community engagement promotes collaborative relationships between stakeholders (i.e. community members, patients, health personnel, local authorities, policymakers and non-governmental organizations) and gives service users an opportunity to be active participants in the care they receive.
Although stakeholders and community members are frequently consulted in health programming, their engagement has been underdeveloped in efforts to improve the quality of health services provided (herein referred to as quality improvement (QI) initiatives), as historically this has been the responsibility of health professionals, managers and planners. Non-health stakeholders, particularly community representatives, are rarely seen as knowledgeable enough to participate in QI initiatives, and when included, their engagement is often limited to providing input into problem identification or serving as (unpaid) human resources for health promotion and programme implementation. 6,7 Principles of stakeholder and community engagement are entrenched in Alma Ata's call for full participation of individuals and communities in health care 8 and are aligned to the fundamental right of individuals to be engaged in processes surrounding their health needs and decisions. 9 Stakeholder and community engagement interventions have proven effective in improving health behaviours and outcomes, 4,5,10,11 providing more acceptable and people-centred services 3,12 and empowering communities, resulting in transformative change beyond the health sphere. 6 Benefits of stakeholder and community engagement for quality of care has been noted across various stages of QI initiatives including design and planning, 13 implementation and delivery, 14 monitoring and accountability 15 and for building relationships between providers and clients. 16 For these reasons, stakeholder and community engagement has been identified as an integral component of improving the quality of care for women, newborns and children, as participatory approaches encourage relevant, effective, context-specific interventions [17][18][19] and can hold service providers and policymakers accountable for quality and implementation. 15,20 Numerous frameworks and reviews exist to provide theoretical and conceptual considerations in guiding the design and implementation of stakeholder and community engagement programmes (Box 1). They highlight the increasing demand for stakeholder and community engagement to be embedded into health-care strategies and QI initiatives.
However, stakeholder and community engagement is not universally embraced. A recent World Health Organization (WHO) policy survey across all six geographic regions found that only 60% of countries have implemented mechanisms at the facility level to solicit feedback on quality and access from community members, users and families. The response varied by region, ranging from a high of 72.7% in countries in the South-East Asian region, 69.2% in the European region, 62.5% in the Eastern Mediterranean region, 58.5% in the African region, 53.3% in the American region and 47.1% in the Western Pacific region. 21 The importance of quality of care, and thus an emphasized need for stakeholder and community engagement to improve quality, is especially pertinent within the field of reproductive, maternal, Newborn, Child and Adolescent Health (RMNCAH). While access to care has increased, reductions in maternal and child mortality have not followed suite with poor quality of care being highlighted as a main driver of such inequities. 22 Poor quality of services for women and children is well documented, especially from low-and middle-income countries 23 and has been found to influence utilization 24 and outcomes 25 of health care.
To this end, The Network for Improving Quality of Care for maternal, newborn and child health (the Network; Box 2) envisions quality of care in terms of the provision of care, the way it is delivered by health workers, and experience of care by women, newborns, children and families. 19,26 The Network has included stakeholder and community engagement within its implementation guidance as a core system required to support QI implementation ( Figure 1). 27 Feedback from countries in the Network indicated interest in receiving support on how to integrate stakeholder and community engagement into QI implementation. In order to respond to this need and to build on existing efforts, we conducted a mapping exercise to identify and compile existing tools on stakeholder and community engagement and link them to the seven different phases of the QI implementation framework (Figure 2). The results of the mapping informed the development of a module integrated into the Network's implementation guidance. This paper presents the methods and results of the mapping exercise and outlines gaps and opportunities for future tool development and support.

| ME THODS
A mapping exercise was conducted to identify and synthesize relevant tools (i.e. formal documents, handbooks and training guides) that could inform the implementation of stakeholder and community engagement across the seven phases of the QI implementation framework ( Figure 2  newborn and child health. 29 Searches were conducted in English, no language restrictions were applied, and dates were restricted to 2010 to present. Identified resources were screened to find programmes or mentions of tools that could be relevant to the mapping.

| Stage 2: open call for tools
The second stage of the mapping exercise involved gathering tools to guide stakeholder and community engagement in QI initiatives, specifically for reproductive, maternal, newborn, child and adolescent health (RMNCAH). In December 2018, a request for tools was sent out via email to key informants from United Nations (UN) agencies, non-governmental organizations (NGOs), listservs, authors of tools or programmes identified in the literature review and other known experts. Persons contacted were informed of the purpose of the mapping and asked to send formal documents used within their organization, or that they were familiar with, that met the inclusion criteria outlined below.
Collaboration was established with a concurrent call for tools to support country efforts for Integrated, People-Centred Health Services, led by the WHO in collaboration with Johns Hopkins University. It was agreed that tools identified in the two exercises would be shared.

| Inclusion and exclusion criteria
Tools identified through the two stages noted above were included in this mapping if they met the following criteria: (1) formal documents (handbooks, training guides, etc) that guide integration of stakeholder and community engagement in one or more of the seven phases of the Network's implementation framework (see  What are the goals of the Network?
• Reduce maternal and newborn mortality in health facilities in target country districts by 50% over 5 years and to halve intra-partum stillbirths; • Reduce avoidable morbidity targeting a 50% reduction in severe post-partum haemorrhage and of neonatal sepsis; • Improve experience of care.

What will the Network do?
• Focus on national leadership by strengthening national and district governance quality of care structures and helping develop national plans and advocacy strategies for improving quality of care. We excluded any tools that (1) only describe, outline or recommend an intervention but do not provide operational guidance on the QI implementation process; (2) were not relevant to stakeholder and community engagement in QI; or (3) were published before 2000.

| Data extraction and synthesis
One author (JS) received and logged all tools in Excel. JS and AP reviewed and discussed all tools against inclusion and exclusion F I G U R E 1 Quality of Care Implementation guide components. Adapted from Ref. [27] F I G U R E 2 The Network's Implementation framework for improving quality of care for maternal, newborn and child health 28 criteria and reached agreement on any materials received that were not clear. A third reviewer (BG) reviewed 20% of all identified items and any differences were resolved by discussion between all three reviewers. JS extracted article information in line with a predefined data extraction template (Data S1), in which characteristics were coded such as level of application (global or country/contextspecific tool), health area (RMNCAH, other health or non-health) and phase of the QI process. AP reviewed the extraction tables.
Key informants were contacted via email when clarifications were needed.
In August 2019, once all tools had been categorized, all authors and/or organizations of the included tools were contacted to request evaluation materials related to the specific tool, including materials, checklists and protocols used to evaluate the intervention, as well as evaluation reports and publications. The results of any evaluation would be valuable information for any future user of the tool.

| RE SULTS
The mapping yielded 197 documents that were screened for eligibility against our inclusion and exclusion criteria. Figure 3 outlines the searching and screening process. In total, we identified 70 tools to be included. Of the 127 excluded documents, 53 were categorized as reference material, in that they were not tools or implementation guidance documents, but their content could provide insight, instruction and explanation on how to engage stakeholders and communities into different phases of the implementation cycle.   x Highlights need to include marginalized populations and some tools include guidance on how to support engagement with such group.
Of these, all but eight tools addressed more than one RMNCAH topic, with six tools addressing two topics, and one tool addressing all topics within RMNCAH. 30

| Tools for stakeholder and community engagement across the QI implementation framework
A main interest of this exercise was to map tools to the different phases in the Network QI implementation framework ( Figure 2). For all included tools, most provide guidance on establishing leadership groups (phase 1), situational analysis (phase 2) and continuous measurement of quality and outcomes (phase 6) and related initiatives, at 76%, 87% and 77%, respectively. Less guidance was identified on involving stakeholders and community members into adapting standards of care (phase 3), identifying QI interventions (phase 4) and implementation of QI interventions (phase 5). Few tools were identified on stakeholder and community engagement for refinement of strategies (phase 7); of the only 15 tools that did cover phase 7, 12 of these covered all of the phases within the QI framework. Phases 1, 2 and 6 were the only ones to occur independently, at two, three and two instances, respectively. Phase 2 and 6 frequently occurred together, with eight tools covering only these two domains. Data S1 maps all included tools to the phases.
There is a wide variation in how each of the included tools addresses the phases of QI implementation. Some documents, such as Save the Children's 'The Spider Tool' 37 for planning and selfassessment of child-led initiatives, or Community Scorecards 36 focus on a specific activity/resource that can be applied to one or more phases. Others include more implementation approaches for stakeholder and community engagement that can be undertaken to support QI across multiple phases, like facility management committees, 33 or Community Health Committees. 38 Others Establishing leadership group still tackle broad concepts within QI, such as partnerships building and/or maintaining, 39-41 data use or evaluation 34,42 and, more broadly, participation across the project cycle, 32,43,44 including numerous approaches that can be undertaken. The latter type typically covers more phases of QI implementation, compared to more discrete activities/resources.

| D ISCUSS I ON
This mapping indicates that many tools are already available on stakeholder and community engagement in aspects of QI and highlights the importance of collecting available resources within this field to build upon and avoid duplication. As the tools included in this exercise have not been assessed for quality, it is hard to conclude that new or more tools are not needed. However, the mapping demonstrates the importance for programmes to take a comprehensive approach to stakeholder and community engagement (addressing all seven phases), to adapt the tools and implementation to specific contexts and to include an evaluation of the effectiveness of the tools. Included tools are available through an online portal on the WHO website (available at: https://www. who.int/activ ities/ tools -to-suppo rt-the-integ ratio n-of-stake holde r-and-commu nity-engag ement -in-quali ty-of-care-initi ative s-formater nal-newbo rn-and-child -health). The portal allows users to identify tools through filters based on the characteristics outlined above.
There appears to be uneven guidance per activity across the QI implementation framework. Topics that relate to consultation with stakeholders and communities (such as phases 1 and 2) and monitoring (phase 6) were more frequently covered, but stakeholder and community involvement in intervention development, implementation and evaluation was less so. This differs from findings in a sys- coupled with longer-term social transformations (e.g. social cohesion and community empowerment) and effects on equity would also be important to capture. 46 Most tools were targeted for global purposes and were developed by international NGOs. A next step for the Network would be to work with partner organizations to understand processes for local adaptation and translation of the predominantly English materials. The country-specific tools can also be reviewed and adapted to other contexts; these tools may offer valuable insights for regional and country-specific issues and goals.
Similarly, many of the tools did not have a specific topic focus (i.e. We speculate that tools specific to every health topic may not be required; rather, generic principles and approaches, with examples and cases of specific sectors or health content, may be sufficient.

RMNCAH
Generic guidance should support the ability to identify strategic and practical objectives of stakeholder and community engagement at different levels of the health system, to conduct stakeholder mapping and analysis for specific QI goals and to learn to adapt tools and processes to the local context. This may help to avoid a rigid, mechanistic or ad hoc approach to applying stakeholder and community engagement, even when considering developing at a larger scale, for instance in country programmes.
Ultimately, the success of stakeholder and community engagement initiatives will depend on broader programme and contextual factors in which initiatives are occurring. 48 Tools need to be considered within a wider process of QI and health system development, and the many factors that influence stakeholder and community engagement will require consideration. This is echoed by authors advocating for improving capability of stakeholders, communities and health actors ensuring linkages with health system, supportive policy and funding environments 7,49 and understanding of populationlevel motivation and limitations. 48 It was beyond the scope of this mapping to assess whether included tools were informed by underlying approaches, such as Human Rights-Based Approach, or integrated interculturality, intersectionality and/or gender considerations in their design and/ or implementation. However, to support adaptation, countries or organizations are encouraged to consider such aspects of a tool's development and focus, ensuring alignment with the programme's goals to support more contextually informed implementation.
Regardless of the QI phase in which stakeholder and community engagement is initiated, it seems to be particularly important to engage in early interaction and conversation with stakeholders and communities 50 to build support and trust and align with broader community needs. 6

| Limitations
While we utilized several methods to search and identify relevant tools, our strategy may have resulted in missed items. This limitation may have been reduced by conducting a parallel structured systematic review, which supported the identification of additional tools or snowballing sources. We were reliant on email and listserv responses and may not have reached all relevant individuals and organizations. While we employed multiple levels of screening and had two screeners, the complexity of these tools makes straightforward inclusion/exclusion difficult; for example, some relevant resources may have been excluded if they did not align to the seven phases.
There were very few tools that were developed by national or local organizations as these resources may have been harder to identify through our searching strategy. An important next step for countries within the Network will be to expand this inventory by gathering the tools within their own contexts.

| CON CLUS ION
Given the considerable number of tools available, developing more impacts is essential. The variation within tools and the overlapping guidance found within the included resources confirm that stakeholder and community engagement cannot be captured in a 'onesize-fits-all' formula. Countries and programmes will need to explore available tools and adapt these to their relevant needs and specific context. However, this field may benefit more from understanding such initiatives and supporting the tailoring and adaptations of existing materials, as opposed to developing more stand-alone resources.

D I SCL A I M ER
The authors alone are responsible for the views expressed in this article and they do not necessarily represent the views, decisions or policies of the institutions with which they are affiliated.

ACK N OWLED G EM ENTS
We would like to thank all the organizations and persons who responded to our request and sent their tools and information. We University for sharing resources from the efforts they led to collect tools for Integrated, People-Centred Health Services. Elizabeth Katwan from WHO Department of maternal, Newborn, Child, and Ageing provided the raw data files from the SRMNCAH Policy Survey.

CO N FLI C T O F I NTE R E S T
None declared.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available in the supplementary material of this article.