Ten years ago, Gabbay and le May published a prominent article in the British Medical Journal in which they challenged the ‘over-rationalist model implicit in evidence based health care’ [1,2]. In an ethnographic study in UK general practice, they showed that clinicians only rarely accessed research findings, clinical guidelines and other types of formal knowledge directly. Rather, they preferred to rely on what they called ‘mindlines’, defined as ‘collectively reinforced, internalised tacit guidelines, which were informed by brief reading, but mainly by their interactions with each other and with opinion leaders, patients, and pharmaceutical representatives and by other sources of largely tacit knowledge that built on their early training and their own and their colleagues’ experience’ [1].

Much subsequent research in mainstream health services research pursued and confirmed the negative finding of this study—that doctors rarely consult written guidelines when making clinical decisions. Studies sought to identify ‘barriers’ to guideline implementation on the assumption that more assiduous following of guidelines by individuals would lead to more evidence-based care—see for example [3-5]. A somewhat smaller literature (reviewed below) aligned with Gabbay and le May’s positive finding—that doctors follow mindlines—and sought to characterise, explore and occasionally critique the concept of collectively embodied tacit knowledge and how it links to the goal of evidence-based practice.

The early evidence-based medicine movement (more commonly known as EBM, but also referred to as evidence-based healthcare or practice) explicitly set out to ‘de-emphasise intuition, unsystematic clinical experience, and pathophysiologic rationale as sufficient ground for clinical decision making’, whilst stressing the instrumental role of evidence from research (especially the ‘gold standard’ design, the randomised controlled trial) in clinical decision-making [6]. Although Sackett et al. later softened this stance by writing that ‘evidence based medicine is not restricted to randomised trials and meta-analyses’ and should be seen as ‘the best external evidence with which to answer our clinical questions’ [7], at a philosophical level, EBM seems to rest on a Cartesian view of knowledge as ‘facts’ that are stored in the heads of individual practitioners and/or in formal knowledge repositories, separate from the physical body, independently verifiable and distinct from values. Notwithstanding recent calls by the ‘real EBM’ campaign for a broadening of the parameters of EBM [8], such a conceptualisation would see mindlines (at best) as a lesser form of knowledge and (at worst) as not really existing.

In this paper, we want to discuss whether and how mindlines have influenced, or should influence, the EBM movement. To explore the impact of mindlines on EBM, we sought to document how the concept of mindlines was picked up and applied by researchers and practitioners following Gabbay and le May’s seminal paper. We also sought to further advance how mindlines have challenged EBM by drawing together findings from these papers as well as revisiting Gabbay and le May’s detailed explanation of mindlines in their 2011 book and bringing in relevant literature from a wider range of disciplinary traditions, including philosophy of science. In this review, we argue that to study mindlines to their full potential, we need to break out of the constraining notions of ‘dissemination’ and ‘translation’ (both of which imply a Cartesian view of knowledge) and focus more on the embodied nature of tacit knowledge and the interactive processes of knowledge creation.

We started by searching for any article with a reference to the word mindline or mindlines using the search query mindline* in the PubMed, Web of Science and OvidSP reference search engines. In PubMed, we selected all NCBI databases up to 2014 and in Ovid the Embase, 1996 to 2014 Week 50, HMIC Health Management Information Consortium 1979 to November 2014, Journals@Ovid Full Text December 16, 2014 and Books@Ovid December 11, 2014 databases. We then searched and included articles that cited Gabbay and le May’s 2004 paper on Web of Science and screened Google scholar and books for any further relevant references. We merged the results and excluded hits in which the term ‘Mindline’ was an author’s name.

Searching for mindline(s), we found one book, 139 articles in PubMed Central (PMC), 11 articles in PubMed, 14 results in Web of Science and 69 results in Ovid. When merged, 196 results remained. We found 213 results citing Gabbay and le May’s paper and another 24 publications by screening Google scholar and books. After merging these results, removal of duplicates and excluding 12 articles by or referring to authors named Mindline, 340 references remained. The complete reference list is in Additional file 1.

There were no publications before 2004 in the remaining dataset; Gabbay and le May appear to be the first researchers to use the word ‘mindline’ [1]. They subsequently elaborated on the concept in a book [12]. Although the paradigm-shifting potential of the minxsgrdlines concept was the subject of an early BMJ editorial [2], uptake of the term by the wider academic community was slow. A study from 2006 was unable to detect ‘mindline’ as a term in a large sample of articles from 12 major medical journals [13]. Most papers referred to Gabbay and le May’s work on mindlines in an implicit way (for example, in 195 publications, we were able to confirm that the term ‘mindline(s)’ was not used at all).

We believe this recommendation reflects a naïve conceptualisation of the ontology of knowledge (seen as facts to be transmitted, simple ones being more transmissible than complicated ones). As we argue below, to understand mindlines requires that we go beyond such a conceptualisation.

The second framing of mindlines, we gleaned from the literature was the ‘in practice’ view, in which the term was used to explain the empirical finding that clinicians are rarely observed to follow written guidelines (but appear to follow mindlines instead). We found 76 papers that primarily referred to mindlines from this point of view. In these papers, the term ‘mindlines’ often seemed to mean ‘consulting colleagues’. For example, one study showed that, like GPs, general surgeons most often turned to colleagues before using other sources of knowledge such as the internet, educational meetings or the library [17].

In a few studies, the ‘in practice’ view of mindlines closely reflected Gabbay and le May’s original theorisation. In a study on the role of tacit knowledge in how public healthcare groups planned initiatives, for example, the authors found that ‘study participants used collectively reinforced tacit guidelines based on experiences and interactions in fluid communities of practice rather than drawing on research findings or explicit practice guidelines’ [18]. In a study on the influences on prescribing in general practice, Grant et al. found that practitioners used ‘prescribing mindlines’. ‘These were personal formularies developed from and informed by their experience of medication (including patient’s experiences), specialist advice, discussions with their practice pharmacist and GP colleagues, and the practice’s macro prescribing policy (if present). GPs rarely looked up information about medicines and relied on these prescribing mindlines. […] GPs relied on personal experience and social networks to update their mindlines’ [19].

In a cross-national study on what physicians gain and lose with clinical experience with diabetes, Elstad et al. write ‘In tune with Gabbay & le May’s “mindlines” […], we found that physicians gained their experience in part through their professional interactions’. They then quote an experienced German clinician who comments on comparing intuitive concerns: ‘Happily we are two physicians working here… I say, “It’s funny, something is wrong with him”. And my colleague says, “You know, you are right. Something is wrong. He simulated, or lied or something is rotten”. We nearly always agree. You can’t have that in an office where you work alone, but it’s really good. We swap ideas on the patients’ [20].

Chandler et al.’s study is cited in 33 papers in our sample, but most of these do not actually refer to the concept of mindlines (they cite the paper to support the statement that there is overdiagnosis of malaria).

Third, mindlines were sometimes framed as a theoretical or philosophical concept in publications (with or without an empirical component) whose main purpose appeared to be the development of theory. We classified 57 papers as predominantly this framing.

Similarly, Crites et al. talk about the knowledge creation cycle and link this to the literature on learning organisations as described by Argyris and Schon [23]; they take from mindlines that external knowledge needs to be ‘validated through informal team discourse and modified for practical application’ by working teams [24]. Ranmuthugala includes Gabbay and le May’s ethnography in an elaborate systematic literature review on why and how communities of practice are established in healthcare [25]. Soubhi et al. offer a theorisation of the communities of practice literature, enhanced by the notion of mindlines, as it might be applied to multi-disciplinary care for complex multi-morbidity [26].

Several studies see mindlines as empirical support for the tenets of social network theory primarily by taking the notion from Gabbay and le May’s work that clinicians rely on their peers to acquire knowledge. They set out to explore what personal relationships between clinicians exist—for example, to study the propensity towards EBM of physicians in relation to their ‘coreness’ (that is, closeness to the centre) in their social network [27], to explore how the structure of patient-sharing relationships among physicians [28] is related to care patterns of high or low costs within hospitals [29] and what factors affect the influence of certain physicians in a network on the thinking and practice of other physicians [30]. An alternative argument is that the concept of social networks neither confirms nor refutes the kinds of knowledge exchange that are implicit in the concept of mindlines.

Other studies link mindlines to social influence. Lomas, for example, provides a useful overview of the literature on the cultural gap between research and policymaking—and the essentially social (not technical) nature of successful efforts to bridge this gap [31]. He uses the concept of mindlines to underline the social nature of collective influence in both the research and policymaking communities. Nine publications in our sample explicitly criticised the political, economic and ethical dimensions of how research knowledge is generated (for example the tainting of research funding through industry conflicts of interest). However, they usually did not link these critiques to mindlines directly.

In an article on clinical decision-making in Ghana, the authors used a novel theoretical framework of guidelines (explicit), clinicians’ mindlines (tacit) and ‘clientlines’. The last are a patient version of mindlines, consisting of ‘client influences related to the preferences and pressures of the client and the wider family and community, including social, religious and cultural values and beliefs’ [32].

In a number of papers, however, this deeper understanding of mindlines was missing, and authors simply equated mindlines to heuristics.

Walach et al. propose a ‘circular’ model for linking evidence and clinical decision-making [36,37]. Like Henry et al. [38], they reject the hierarchy of evidence (randomised trials at the top, ‘anecdotes’ at the bottom) as valid only in relation to simple decisions about the efficacy of drug therapies. Most clinical decisions, they argue, are complex; they involve ethical and human decisions as well as scientific ones and, hence, require the integration of multiple considerations and forms of evidence. The tacit knowledge of mindlines may be more appropriate for the complex and organic nature of real-world medicine than the ‘if…then’ structure of guidelines and decision support tools.

In a systematic review of knowledge exchange mechanisms, Contandriopoulos and colleagues consider knowledge in two essential forms: individual, that is, held in people’s heads and translated (or not) into action by human will and agency and collective, that is, socially shared and organizationally embedded—a form akin to Gabbay and le May’s mindlines [39].

A final framing of the ‘mindlines’ concept, sometimes offered as a conclusion in an empirical or theoretical paper, was solution-focused: some authors considered the question of how to actively promote and support the development of valid embodied/collective knowledge or ‘evidence-based mindlines’. In 28 sources, this seemed the main purpose of the article. For example, the challenge was expressed (somewhat obliquely) in a letter by Glasziou, in which he recognises the concept of mindlines in clinical practice, but is worried that they could ‘supply counterfeit evidence. […] a puzzle remains: how do we get valid memes into the mindlines while not driving out the wisdom of experience?’ [40].

Two studies describe efforts to set up (and influence the behaviour of) communities of practice among researchers [41] and doctors [42] by employing a facilitator and co-ordinator respectively, though each of these studies mentions mindlines only in passing. In a more theoretically informed paper, Soubhi et al.’s model of communities of practice in multi-morbidity care is also solution-focused, emphasising relationship-building and collaborative learning as the basis for developing mindlines. They hope that qualitative and quantitative research could ‘examine how primary care physicians develop mindlines and how they test them to eliminate harmful ones and standardize others into routine practice’.

Others call for knowledge brokers [31], transformational leaders [43], thought leaders [44] and individuals more generally [41] to alter, expand and embed new knowledge through social influence within (and indeed extending beyond) existing communities of practice. These proposals echo and flesh out Gabbay and le May’s original exhortation to make sure that the knowledge circulating within communities is based on sound research [1]. In their systematic review, Contandriopoulos et al. suggest that knowledge may become collectivised through a variety of mechanisms, including efforts to make it relevant (timely, salient, actionable), legitimate (credible, authoritative, reasonable), and accessible (available, understandable, assimilable) and to take account of the assumptions and priorities of a particular audience [34].

Reeve et al. draw parallels between mindlines and the (subtle and often overlooked) skills of expert generalism in GP practice [45]. These authors offer a four-phase approach to developing generalist expertise: sense-making (popularising the concept of generalist expertise and raising awareness of it across a community of practitioners), engagement (influencing practitioners to prioritise this issue), action (e.g. delivering education, promoting scholarship as part of professional practice) and monitoring (measuring the impact of this approach).

In an observational study to assess how ‘evidence-based’ GP consultations were, Zwolsman et al. observed that GPs were often unable to account for the source of their knowledge (personal experience or research evidence) and that rapid, intuitive decisions, suggesting a predominantly tacit form of knowledge, were the norm [46]. The authors felt that making all steps in the decision-making chain more explicit (for example, justifying the chain of reasoning to the patient) would surface gaps in knowledge, inform further learning and—in the long run—make practice more evidence based.

In contrast, Levine et al. suggest the opposite. They propose making guidelines more implicit by transforming them in to aphorisms which they define as ‘succinct sayings that offer advice’. These should function as ‘interface between intuitive approaches to make rapid decisions, and the implementation of specialty-specific clinical guidelines’ [47].

This systematic narrative review has revealed a relatively sparse literature on the important concept of mindlines, first introduced in 2004 by Gabbay and le May. Whilst we included 340 publications in our final dataset, some authors appeared to have (with the best of intentions) used this term incorrectly and naively. But others had recognised and explored mindlines’ fundamental philosophical challenge to EBM. The relatively limited contributions from the guideline development community suggest that the concept has been largely ignored by the guideline industry—perhaps because, as American physicist and philosopher of science Thomas Kuhn wrote, ‘the proponents of competing paradigms practice their trades in different worlds’ [48].

‘Knowledge translation’, one-way transfer of knowledge from producer (research) to user (clinician) [49,50], remains a (contested) policy challenge. Multiple research traditions have contributed to a vast body of literature on how evidence from research is disseminated [51,52]. Underlying this literature is a fundamental hope for the possibility of optimising the ‘intermediation’ of knowledge—that is, the managed processes by which practitioners can be supported to interact with knowledge [53]. But an alternative metaphor that goes beyond the concept of ‘spreading’ good ideas is needed [54].

In their book from 2011, Gabbay and le May explain in great detail the origins of mindlines, their implications and related theories [12]. Mindlines fit a view that knowledge is not so much a set of external facts that are waiting to be ‘translated’ or ‘disseminated’ but a more fluid and multi-directional phenomenon in which knowledge is ‘re-created’ in different contexts by different people again and again as previously postulated by Mol and Nonaka [55,56]. From this perspective, improving knowledge intermediation is more like maximising the opportunity to create knowledge. How this might be achieved differs from the EBM paradigm in a number of ways uncovered by Gabbay and le May throughout their book. We explore these philosophical questions further below.