Uncovering the processes of knowledge transformation: the example of local evidence-informed policy-making in United Kingdom healthcare

Background: healthcare policy-makers are expected to develop 'evidence-based' policies. Yet, studies have consistently shown that, like clinical practitioners, they need to combine many varied kinds of evidence and information derived from divergent sources. Working in the complex environment of healthcare decision-making, they have to rely on forms of (practical, contextual) knowledge quite different from that produced by researchers. It is therefore important to understand how and why they transform research-based evidence into the knowledge they ultimately use. Methods: we purposively selected four healthcare-commissioning organisations working with external agencies that provided research-based evidence to assist with commissioning; we interviewed a total of 52 people involved in that work. This entailed 92 interviews in total, each lasting 20-60 minutes, including 47 with policy-making commissioners, 36 with staff of external agencies, and 9 with freelance specialists, lay representatives and local-authority professionals. We observed 25 meetings (14 within the commissioning organisations) and reviewed relevant documents. We analysed the data thematically using a constant comparison method with a coding framework and developed structured summaries consisting of 20-50 pages for each case-study site. We iteratively discussed and refined emerging findings, including cross-case analyses, in regular research team meetings with facilitated analysis. Further details of the study and other results have been described elsewhere. Results: the commissioners' role was to assess the available care provision options, develop justifiable arguments for the preferred alternatives, and navigate them through a tortuous decision-making system with often-conflicting internal and external opinion. In a multi-transactional environment characterised by interactive, pressurised, under-determined decisions, this required repeated, contested sensemaking through negotiation of many sources of evidence. Commissioners therefore had to subject research-based knowledge to multiple 'knowledge behaviours'/manipulations as they repeatedly re-interpreted and recrafted the available evidence while carrying out their many roles. Two key 'incorporative processes' underpinned these activities, namely contextualisation of evidence and engagement of stakeholders. We describe five Active Channels of Knowledge Transformation - Interpersonal Relationships, People Placement, Product Deployment, Copy, Adapt and Paste, and Governance and Procedure - that provided the organisational spaces and the mechanisms for commissioners to constantly reshape research-based knowledge while incorporating it into the eventual policies that configured local health services. Conclusions: our new insights into the ways in which policy-makers and practitioners inevitably transform research-based knowledge, rather than simply translate it, could foster more realistic and productive expectations for the conduct and evaluation of research-informed healthcare provision.