Beyond EBM

Introduction The evidence-based medicine (EBM) movement has claimed to have induced and improved the production of trustworthy clinical guidelines. However, ethnographic research showed that clinicians rarely use explicit evidence from guidelines directly. Instead they draw heavily on socially shared knowledge and patterns of behaviour, called ‘mindlines’. To understand how EBM reaches its limits in producing meaningful guidance and finding what may lie beyond, this thesis explores the tensions between guidelines and mindlines from a philosophy of science perspective. Methods The approach I take to uncover these tensions is to view them as anomalies of a mature Kuhnian paradigm. Unlike most previous scholars, I chose not to study the use of guidelines in clinical practice but the guideline development process as a central means of the EBM paradigm to create decision making support for individual patients instead. The project draws on theoretical findings from a literature review of mindlines and philosophical analyses of perceptions in EBM about knowledge, reasoning and truth, which are juxtaposed and synthesised with empirical data from a digital ethnography of mindlines in virtual social networks of clinicians and an ethnography of guideline panels in the United Kingdom, Norway and the Netherlands. Results I show how mindlines have been researched and acknowledged in multiple contexts and evolved conceptually as a container for many theories of knowledge. In a critique I consider how mindlines challenge the fundamentals of EBM. Taking a Latour inspired perspective, I elicit how a changing EBM is trying to cope with the downsides of what he calls “purification”: a scientific practice of categorising and quantifying reality. I then provide insights into the challenges of EBM for guideline developers concerning ways to make inferences and integrate evidence. It highlights important issues in guideline development regarding its purpose, use and integration of different knowledges and suggests alternatives for the dominant reasoning with RCTs. Next, drawing on work by Isabelle Stengers, I consider bias and the ideal limit theorem as a theory of truth in EBM. I argue that debates about truth remain hidden behind debates on bias and contend that the latter is in a dual, complex, necessary, unproductive as well as a productive conjunction with truth. I present characteristics of the ideal limit theorem as a dominant theory of truth in EBM and look at several other theories of truth that operate in the clinical encounter. I then report the ethnographic findings on how all these issues played out in some real examples of informal knowledge creation and sharing in large groups of clinicians, and in formal guideline production panels. I provide an analysis of the characteristics of informal knowledge in virtual social networks of clinicians in the United Kingdom, Norway and the Netherlands, which sees it as produced collectively by clinicians to base their decisions on; a concept for explicit as well as tacit knowing shared among social groups and reinforcing norms of good practice in a fluid, dynamic and constantly evolving way. Analysis of the findings of guideline panels in the UK, Norway and the Netherlands indicate that there are several tensions regarding mindlines and guidelines, most notably originating from two linked yet contrasting epistemologies within the guideline community: a purist understanding of EBM, adhering to frequency type reasoning, the ideal limit theorem and strict explicit purification, and a more pluralist understanding of EBM embracing multiple kinds of knowledge, reasoning and truth. Conclusion My empirical and theoretical work was aimed at providing new insights in mindlines and the production of guidelines using an epistemological perspective on EBM as a mature and changing paradigm, with a focus on knowledge, reasoning, truth and purification. The implications for guideline production, teaching EBM and future research include a call to rethink the nature of knowledge and knowledge translation, to develop and examine new guideline tools including the use of virtual social networks, to discuss and critically appraise hidden assumptions regarding key concepts such as bias and pluralism, as well as to further theorise inductive inference in individual patients with the help of the philosophy of science as a proper academic field in healthcare.