Developing a collaborative care model for people with severe mental illness: intervention theory development

Problem

Ongoing care for individuals with severe mental illness (SMI) is often split between specialist mental health services and general practices. The PARTNERS2 (P2) programme aims to develop and evaluate a model of collaborative care (CC) to support adults with SMI have improved quality of life and care by situating specialist mental health workers (care partners) within general practices. This presentation describes how we developed a coherent, practical intervention in preparation for a RCT.

Approach

A realist synthesis drew from a) systematic review; b) narrative review of recovery; c) telephone interviews with international leaders, to create a prototype intervention theory. We extracted causal micro-hypotheses embedding them in the CC model. This was tested in practice in two sites and the intervention theory further developed with a formative evaluation incorporating interviews, tape assisted recall and activity logs.

Findings

The first stages of the synthesis process produced agreement for most key issues about an optimum model of CC for SMI:• Physical, mental health and social outcomes are all important • Individualised goals are prioritised • Monitoring & support for self-management should be provided through coaching • Care Partners (the case managers) are the central intervention resource, liaising with GPs, practice nurses, psychiatrists to ensure joined up care• GPs need to understand role of Care Partner in self-management enhancement• Supervisors need to have experience of supervision and working with individuals with SMI. • There is a need for shared records so that goals, plans and responsibilities are understood by allThe formative evaluation endorsed the potential value of all these components. Coaching was valued by individuals but came more naturally to some care partners. Most individuals likes the individualised approach, with some commenting that they were able to take positive risks. Some found identifying future oriented goals difficult. Care partners took time to feel at home in practices and sometimes did not link proactively with practice teams. The manual and initial training was valued by care partners. Supporting researchers were particularly valued by care partners in their roles of practical support and advice about intervention delivery.

Consequences

The systematic review revealed heterogeneity of CC models and anticipated outcomes. The narrative review emphasised the need to build on individuals’ strengths and address personal goals. The expert/leader interviews provided well-reasoned practical solutions and emphasised the need for strong supervision. Consensus meetings involving PPI members with open challenging discussions were able to reach agreement about the uncertainties. The formative evaluation provided support for the intervention theory with minor adaptations but demonstrated the need for a stronger ‘delivery platform’ . For the trial we will include follow up training, strong regular supervision, a formalisation of support provided by the research team and the use of tape assisted recall as part of supervision to support coaching.