Effective stakeholder engagement in research and implementation is important for improving the development and implementation of policies and programmes [
1‐
4]. We define stakeholders as individuals, groups or organisations who have the potential to influence or who may be influenced by particular actions or aims [
3,
5]. Stakeholders are not uniform, but vary in each context by their available resources, their position and their interests. Consequently, reasons for engaging them, and their engagement levels with a project, may differ. Arnstein [
6] proposed eight levels of stakeholder participation, wherein the first (manipulation) and second level (therapy) allow no participation at all, while the third (informing), fourth (consulting) and fifth (placation) allow forms of tokenism in which stakeholders are informed of issues and their views are sought (fourth and fifth), but decisions are still made by those who hold power. Finally, in the sixth (partnership), seventh (delegated control) and eighth (citizen control) levels, shared decision-making and increasing levels of control are given to the stakeholders.
Overall, the process of stakeholder engagement can be mutually beneficial. Stakeholders may choose to engage with researchers because the research project might directly affect individual stakeholder interests, the engagement process might have financial incentives or benefits, or the engagement may lead to outcomes or outputs that benefit the general population [
3]. Researchers and project implementers, on the other hand, may have slightly different reasons for engaging stakeholders, including to understand the power, interests, perspectives, values, behaviours and opinions of stakeholders, to understand how change happens in different contexts and among different individuals, to build the capacity of local stakeholders by creating a learning process and developing leaders and teams, to create a stimulus for change, to promote local ownership, and to assess the effect of a programme [
7‐
11].
According to Durham [
11], when choosing a method for engaging with stakeholders, it is important to consider the aim of the engagement, the resources available and the expectations of stakeholders. In practice, researchers have employed various tools to engage stakeholders. Provision of information to stakeholders has often been done through simple stakeholder workshops or meetings. Alternatively, consultation of stakeholders about their interests, needs, relationships, perceived benefits of a project, or about drivers of change has been performed through a range of methods that include most significant change, participatory evaluation, positive deviance approach and beneficiary assessment [
2,
4,
7,
8,
10,
12]. For higher levels of engagement, participatory mapping and/or participatory social network analysis (PSNA) can be used to facilitate stakeholder involvement. Finally, tools such as participatory impact pathways analysis (PIPA) and approaches such as participatory action research are used by researchers to develop active partnerships and stakeholder engagement in project decision-making.
Participatory approaches are increasingly being advocated for because they give stakeholders a voice and allow them to table their concerns, as well as improving the identification of local problems and suggestions of feasible solutions and promoting the uptake of local solutions [
13‐
15]. However, participatory approaches differ in the extent to which they involve the community in decision-making and hence in the extent to which they empower the community to address problems [
14]. Approaches that are simply used to inform the community and stakeholders about what will be done, or that are used to facilitate community involvement in predetermined activities without shared decision-making are examples of passive community participation that generally tend not to empower the community, while those that allow the community to identify what their problem is and to get involved in identifying solutions for these problems are examples of active community participation. The latter empower the community to deal with not only their current problems, but also their future problems [
13‐
16].
The use of many of these methods is still in its infancy, especially in low-income countries [
17,
18]. In this paper, we discuss two participatory methods for engaging with stakeholders – PSNA and PIPA, which we have adapted and used to engage stakeholders as part of our work in the Future Health Systems (FHS) project in India and Uganda, respectively. Based on our experience, we derive lessons about when and how to apply these tools. Our work adds to the existing literature that summarises practical experiences with the use of these tools, highlighting the applicability and limitations of using the methods in different contexts.
Social network analysis (SNA) has been defined as a tool that allows the mapping and measuring of relationships and flows between people, groups, organisations or other information/knowledge processing entities [
19,
20]. Furthermore, it provides an opportunity to compare formal and informal information flows. Such information can guide the planning and implementation of new interventions [
17].
According to Blanchet [
17], there are three main stages in SNA, namely (1) identification and description of the actors, (2) characterising the relationships between the actors, and (3) analysing the structure and pattern of the network. PSNA follows the three outlined stages, but also adds the use of participatory approaches that permit more interaction between the researchers and the participants and allows for feedback of results to stakeholders [
21,
22]. These results can then be used to identify issues that need to be resolved – by so doing it provides a catalyst for change [
21,
23]. However, for this to happen, there must be a level of trust between the researchers and the participants so as to allow free discussion [
21]. In addition, the participants need to have the willingness and ability to solve any issues that they feel warrant their attention [
23].
PSNA was applied in the Indian Sundarbans – the world’s largest mangrove delta – as part of a knowledge intervention aimed at engaging different stakeholders through knowledge creation, dissemination and effective up-take of knowledge regarding child health in the Sundarbans to inform and influence existing health policies in the region. The Sundarbans region is characterised by poverty, with frequent climatic events, which often lead to massive destruction of the already poor infrastructure, leaving behind displaced families with insufficient food and low productivity of the land for cultivation and ponds for fishing. This situation has led to migration of males in search of alternative livelihood, creating significant numbers of women-headed households. Furthermore, the child health status is poor, with chronic malnutrition and a high burden of communicable diseases [
24]. Public health service delivery options are either absent or non-functional. Although non-governmental organisations (NGOs) provide some services, they cover only a limited area. Consequently, the gaps in health service delivery are filled by numerous Informal Healthcare Providers (IHPs), who practice modern medicine without any formal training or authorisation, locally referred to as village doctors or quacks.
PIPA is a relatively new planning, monitoring and evaluation tool designed to help the people involved in a project, programme or organisation work out how they will achieve their goals and impact [
18,
25,
26]. PIPA analyses project impact through the use of problem trees and network pathways. The problem trees utilise linear logic that shows how the problems solved by the project eventually contribute to solving other related problems, achieving the programme goal. On the other hand, the network pathways show how the actions and interrelationships between different actors contribute to creating an enabling environment to solve the problems identified [
18]. PIPA involves five distinct steps that include construction of problem trees, visioning, developing network perspectives, and defining an outcome logic model and an impact model. PIPA is usually implemented through 2- to 3-day workshops. The sessions are conducted through group meetings that comprise 4–6 stakeholders with a total of 3–6 groups. The workshops may be done at the beginning, middle and end of a project. However, different implementers have used it at different time points in their study. Alternatively, smaller reflection meetings can also be held to monitor progress, for example, every 6 months. These meetings provide an opportunity for learning and hence can provide a springboard for action research. In addition, for follow-up reflection meetings, linking the PIPA meeting to other technical or administrative meetings seemed to work better [
18].
Some of the benefits that have been attributed to the use of PIPA include providing mutual understanding about intervention logic and the potential for achieving impact, an opportunity for ex ante impact assessment and a hypothesis for post ante impact assessment, in addition to providing a framework and design that enhances implementation that is aligned to the project/programme plans with room for learning during the monitoring and evaluation process. It can also promote collaboration between different programmes by making existing opportunities explicit [
18].
The PIPA tool was implemented in three rural districts in Uganda (Kamuli, Kibuku and Pallisa), as part of a project that aimed to increase the number of births attended by skilled attendants. These districts have a high maternal and neonatal mortality rate comparable to that of the rest of the country (maternal mortality rate of 438 per 100,000 live births, neonatal mortality rate 27 per 1000 live births) [
27]. In Uganda, the uptake of cost effective interventions that can reduce this maternal and neonatal mortality has been limited by factors such as poor maternal and newborn care practices, poor healthcare seeking behaviour, lack of financial means, inadequate infrastructure, and the existence of few overworked and poorly motivated health workers [
28‐
31]. The Ugandan FHS project, MANIFEST (Maternal and Neonatal Implementation for Equitable Systems), focused on addressing problems related to inadequate knowledge about maternal and neonatal healthcare (MNH) practices, birth preparedness, poor access to emergency and routine transport, and poor quality of care at health facilities. Community mobilisation strategies supported locally organised, financed and monitored transport systems. Linkages between the community and the health facility were improved by using community health workers, who in Uganda are called Village Health Teams (VHTs). Quality of care improvements were stimulated using only non-financial incentives, which included training of health workers, mentorship, supportive supervision and recognition awards. The project was implemented using a participatory action research approach. PIPA was therefore seen as a method that would allow participatory monitoring of impact not only through the eyes of the researchers, but also through those of the community, who were both participants and implementers in this project.