Navigating the complexity of a collaborative, system-wide public health programme: learning from a longitudinal qualitative evaluation of the ActEarly City Collaboratory.

Background: Addressing the upstream social determinants of health (e.g. built environment, education) can reduce the burden of non-communicable diseases. To do so effectively often requires system-wide collaboration. However, collaborating across multiple sectors, organizations and disciplines within a complex system can be challenging. ActEarly was a public health research consortium that aimed to improve child health by building an interdisciplinary, cross-city partnership to develop and/or evaluate upstream interventions, increase research capacity and improve collaboration between researchers, local authorities and communities. This paper explores ActEarly's experiences of navigating complexity to identify mechanisms that supported its implementation and proposes recommendations for future intersectoral and interdisciplinary population health research collaborations.

Methods: We conducted a longitudinal qualitative study of ActEarly, integrating findings from inductive documentary analysis of internal documents (mainly meetings minutes and reports) (n = 114) and interviews (n = 70) with 45 consortium members at three different timepoints (2018, 2021, 2023). Participants worked across different organizations, cities, roles and levels of seniority in the consortium.

Findings: Clarity, Unity, Flexibility and Feasibility were seen as the key mechanisms required to support ActEarly's implementation. Clear aims, governance structures and communication were necessary to manage the uncertainty of the complex system. A unified approach, characterized by strong relationships, having a shared vision and communal access to resources supported effective collaboration. Flexibility was required to adjust to different ways of working, respond to wider system events and manage the consortium. Establishing feasible aims that responded to the limitations of the system, the available resources and research infrastructure was required for teams to deliver the work.

Conclusions: Implementing multi-faceted programmes in a complex system can be challenging. We recommend that future whole-systems consortia seeking to improve population health build Clarity, Unity, Flexibility and Feasibility into their programmes, noting the complex interrelationships between these factors. Iterative reflections from all parties should support delivery amidst the uncertainty that comes with running a population health research collaboration, and strong leadership and governance should play a key role in ensuring that these are built into foundations the programme.