Engaging stakeholders in embedded newborn health services/systems research in Kenya; a continuous process involving multiple actors.

Background Engaging relevant stakeholders throughout the research cycle is increasingly recognised as critical to conducting quality health systems research. There are few descriptions and analyses of stakeholder engagement in practice for embedded health systems research especially those that must navigate multi-level decentralised health systems. We describe and reflect on the stakeholder engagement activities of an international multi-disciplinary programme of research focused on newborn care in hospitals in Kenya. Methodology Our experienced project stakeholder engagement group coordinated engagement activities across multiple stakeholders ranging from those close to the intervention to those further away. with differing interests in the research. We conducted a stakeholder mapping and analysis using an engagement matrix to include national and county-level policymakers, professional communities, associations and regulators, health managers, frontline healthcare workers, patients, families and patient representative groups. Our engagement group maintained a log of engagement activities and had regular programme feedback meetings. Our analysis of stakeholder engagement drew on Programme’s documents and meeting minutes, and on a conceptual framework which distinguishes between the moral, strategic and practical dimensions of stakeholder engagement. Results We engaged a wide range of stakeholders based on our understanding of their needs, interests and concerns. We drew on the International Association for Public Participation model on encompassing ‘inform’, ‘consult,’ ‘involve,’ ‘collaborate’, and ‘empower’ to inform strategies of engaging stakeholders and the need to balance moral, strategic and pragmatic components of engagements. Although we had significant prior engagement experience and relationships at the hospitals and the counties, introducing new staff into Newborn Units triggered complexities that required careful consultation along the bureaucracies at the counties. Despite the counties having similar hierarchical architectures, engagement processes varied and achieved different research approval, recruitment of additional workforce and outcomes across counties. There were also multiple officeholder transitions over the research period, occasioned by factors in our external environment, often necessitating engaging afresh. Conclusion Even with a carefully developed stakeholder engagement plan, an experienced team, and a landscape backed by long-term relationships and embeddedness, health research stakeholder engagement can be complex and unfold in unexpected ways and requires continuous effort, resources, and adaptability. Meeting the moral, strategic, and practical potential of engagement requires flexibility, responsiveness, and commitment, including adequate resources.