System-level mechanisms and contexts for health and social care coordination through Multi-Specialty Community Providers in England: a Realist evidence synthesis

Introduction : Current National Health Service (NHS) policy towards care coordination in England, involving the introduction of Multi-Specialty Community Providers (MCPs), assumes that repeated unplanned admissions of older people with multiple morbidity make disproportionately heavy use of hospital bed-days, that a substantial number of these admissions are preventable, and that reducing these admissions would substantially reduce cost and access pressures on hospital services. 'Integrated' (better-coordinated) care, delivered by MCPs, is intended to reduce these admissions by partly replacing hospital care with non-hospital care, hence raising the quality and reducing the cost of care. Our study synthesised existing evidence about how to integrate care at the system-level, including management of complexity and whole system patient flow. Theory/Methods : Stage 1. From grey literature, MCPs’ published logic models, and a think tank (policy makers, health-workers, patients/public), we elicited policy-assumptions ('programme theory') about what structures, working practices and services ('mechanisms') the MCPs will contain, and how these mechanisms are expected to produce the outcomes above. Stage 2. Realist review: Evidence from other ‘integrated’ care projects which related to the programme theory identified in stage 1 was synthesised to produce a revised, more strongly evidence-based explanation of how complex, MCP-like ‘integrated’ care systems function in different contexts. 3. 'Critical Analysis': The policy-makers' initial programme theory (from stage 1) was compared with the evidence from stage 2. In that way the policy-makers’ initial programme theory was elaborated, qualified and revised, better to inform the development of health and social care co-ordination in England. Results : We present 1) the initial programme theory, built from NHS England policy-makers' assumptions as to what outcomes the MCP models of care integration will produce in England, and by what means, and 2) some main findingsof how different models of care co-ordination might contribute to achieving these outcomes and under what conditions. Discussions : We discuss 1) how realist review methods were used to elaborate, qualify or challenge policy-makers'  assumptions about how care co-ordination by means of MCPs would achieve the desired outcomes, and 2) how these findings can inform better care co-ordination. Conclusions : We describe the main empirical strengths and weaknesses in the assumptions underlying the models of care and care coordination which inform current NHS policy, and indicate where existing evidence is weak or lacking. Lessons learned : We discuss how we met the challenge of focusing the scope of a project that encompasses complex system-level issues across health and social care, and how far the findings and lessons might apply to other, differently-structured health systems. Limitations : Policy-makers assumptions were complex and incompletely articulated, making our findings depend, at points, upon our interpretation of what policy makers specifically intended. The complexity constrained us to focus our review on the policy makers’ most central assumptions. Future research : 1) Understanding how the logic models informed practice by evaluating the usefulness, and any future policy application, of our findings by policy makers in England. 2) Testing our findings empirically through primary research (Realist evaluation) of MCPs in practice.