Realistic nurse-led policy implementation, optimization and evaluation: novel methodological exemplar

Authors


Abstract

Aim

To report the first large-scale realistic nurse-led implementation, optimization and evaluation of a complex children's continuing-care policy.

Background

Health policies are increasingly complex, involve multiple Government departments and frequently fail to translate into better patient outcomes. Realist methods have not yet been adapted for policy implementation.

Design

Research methodology – Evaluation using theory-based realist methods for policy implementation.

Methods

An expert group developed the policy and supporting tools. Implementation and evaluation design integrated diffusion of innovation theory with multiple case study and adapted realist principles. Practitioners in 12 English sites worked with Consultant Nurse implementers to manipulate the programme theory and logic of new decision-support tools and care pathway to optimize local implementation. Methods included key-stakeholder interviews, developing practical diffusion of innovation processes using key-opinion leaders and active facilitation strategies and a mini-community of practice. New and existing processes and outcomes were compared for 137 children during 2007–2008.

Results

Realist principles were successfully adapted to a shorter policy implementation and evaluation time frame. Important new implementation success factors included facilitated implementation that enabled ‘real-time’ manipulation of programme logic and local context to best-fit evolving theories of what worked; using local experiential opinion to change supporting tools to more realistically align with local context and what worked; and having sufficient existing local infrastructure to support implementation. Ten mechanisms explained implementation success and differences in outcomes between new and existing processes.

Conclusions

Realistic policy implementation methods have advantages over top-down approaches, especially where clinical expertise is low and unlikely to diffuse innovations ‘naturally’ without facilitated implementation and local optimization.

Ancillary