Abstract
INTRODUCTION: Co-production with communities is increasingly seen as best practice that can improve the quality, relevance and effectiveness of research and service delivery. Despite this promising position, there remains uncertainty around definitions of co-production and how to operationalize it. The current paper describes the development of a co-production strategy to guide the work of the ActEarly multistakeholder preventative research programme to improve children's health in Bradford and Tower Hamlets, UK. METHODS: The strategy used Appreciative Inquiry (AI), an approach following a five-step iterative process: to define (Step 1) scope and guide progress; to discover (Step 2) key issues through seven focus groups (N = 36) and eight in-depth interviews with key stakeholders representing community groups, and the voluntary and statutory sectors; to dream (Step 3) best practice through two workshops with AI participants to review findings; to design (Step 4) a co-production strategy building on AI findings and to deliver (Step 5) the practical guidance in the strategy. RESULTS: Nine principles for how to do co-production well were identified: power should be shared; embrace a wide range of perspectives and skills; respect and value the lived experience; benefits should be for all involved parties; go to communities and do not expect them to come to you; work flexibly; avoid jargon and ensure availability of the right information; relationships should be built for the long-term; co-production activities should be adequately resourced. These principles were based on three underlying values of equality, reciprocity and agency. CONCLUSION: The empirical insights of the paper highlight the crucial importance of adequate resources and infrastructure to deliver effective co-production. This documentation of one approach to operationalizing co-production serves to avert any misappropriations of the term 'co-production' by listening to service users, stakeholders and other relevant groups, to develop trust and long-term relationships, and build on the learning that already exists amongst such groups. PATIENT OR PUBLIC CONTRIBUTION: The work was overseen by a steering group (N = 17) of individuals, both professional and members of the public with experience in undertaking co-production, and/or with some knowledge of the context of the two ActEarly field sites, who provided regular oversight and feedback on the AI process.