Realist process evaluation of a co-designed intervention to address the unmet information needs of people with dementia and their family carers.

Information needs among people with dementia and their family carers remain often unmet, despite policy emphasis on information as the key enabler of choice, agency, and empowerment. An intervention to address these unmet information needs was co-designed and implemented in the North-East of England. This study is a realist process evaluation of the intervention's implementation through the local community mental health services for older people. The realist process evaluation drew on three strands of qualitative data: longitudinal interviews with the service manager of the implementing services, longitudinal group interviews with practitioners, and a focus group with the co-design working group. Thematic analysis used a deductive approach to identify implementation outcomes (Reach, Dose, Fidelity & Adaptation, Acceptability, Adoption, Appropriateness, Penetration, Feasibility, Sustainability), followed by a retroductive approach to interpret data through the realist framework of Context, Mechanisms, and Outcomes. Participants showed different strategies for offering the leaflet in clinical practice (Reach), but all appeared to offer it multiple times and at every opportunity (Dose). Some adaptation in implementation was identified. The intervention was successfully integrated into practice (Penetration, Feasibility) and well received by implementers (Acceptability, Adoption). The intervention was perceived as addressing a genuine need and as compatible with the implementation setting (Appropriateness), also in the longer term (Sustainability). Nine levels of contextual factors and eight mechanisms appeared to contribute to the observed outcomes. The study offers an in-depth understanding of how information-giving occurs at the level of the individual practitioner and identifies some variation in information-giving practice. Some participants' approach to information-giving was consistent with a person-centred, relational, and situated approach to care, while others embodied a rational approach to information which did not account for the circumstances of the information recipient. Policy and practice should acknowledge and address this variation to ensure that critical stances on information-giving are widely adopted.