rules. The similarities are clear. First, EBHC enables centralized control and coordination in the form of clinical guidelines and protocols. Second, it emphasizes the separation of planning and execution in the form of science ���push��� models of innovation, diffusion, and technology transfer. Third, it institutionalizes the closer supervision of individual work performance by attempting to limit clinicians��� traditional freedom to prescribe. This paper is based on empirical research, which extends from a continu- ing concern for connecting health care research information with those in daily clinical practice (Wood and Ferlie 2000). The work was undertaken by the authors, who received funding from the NHS Executive London, Research and Development Programme. The views expressed in the paper are those of the authors and not necessarily those of the NHS Executive or the Department of Health. Drawing on our research and ideas from conti- nental philosophy, the paper will analyse qualitative empirical data to contribute a broader understanding of the organization of health care research. This raises interrelated issues of knowledge production and its appropriation, especially where the interface between research and practice is revealed as problematic. The proposition is that ���research��� and ���practice��� are more than a series of distant points, positions, or contiguous elements, which correspond without penetrating one another. Although both are real, they no longer have abstract form and functions that can be simply located. There is no fixidity of being that dualisms imply, nor any connections between them. Instead, they out- line a mutually constitutive process of becoming, each into the other. This analysis is informed by Bergson���s (1911) conception of the ���continuity of movement��� and Deleuze and Guattari���s (1988) notion of the ���rhizome���. These phrases refer to particular modes of immanence, which scramble the purported communicational transparency of EBHC models. The organiz- ation of health care knowledge is understood less as the objective move- ment between stops, than as the arrestation of a fluid reality, constructed as a function or effect of the various attitudes, behaviours, values, abilities and beliefs of investigators, health care professionals and other stakeholders. The term ���investigator��� is used throughout the text to refer to the subjects of the research and not the authors of the text. The following section offers a general description of EBHC intentions, together with a critical assessment of its assumptions about the dissemina- tion of knowledge and changing practice. Section three reviews some contemporary social science literature on the organization of knowledge. It uses a processual framework in the belief that this can help broaden the dominant health care literature, which expresses the idea of research gen- eration and appropriation as the communication of messages between separate domains. In section four the research design and methodology will be introduced. As might be expected, the resulting tabulation greatly simpli- fies the complexities of the individual empirical settings and is intended as a loose fit of research approaches rather than a rigid taxonomy. In section five, the background and character of the four qualitative cases are intro- duced. A brief, factual description of each is presented and this provides a 48 Organization Studies 24(1) at GLASGOW CALEDONIAN UNIVERSITY on March 5, 2012 oss.sagepub.com Downloaded from

contextual framework for a more analytical discussion in the following section. The paper concludes by recognizing the continual interpenetration of both research and practice within each other rather than a juxtaposition of discrete phases or stages. An Introduction to Evidence Based Health Care The organization of knowledge in EBHC is heavily prescribed by scien- tific method, whose facts and artefacts are part of a powerful rationalistic epistemology. A central tenet is that clinical judgements should be based on an explicit logic of empirical evidence combined with consideration of a patient���s individual needs (Barton 2000). Haines (1996) recognizes it as an important means of tracking down best evidence and critically apprais- ing the validity and usefulness of its application to practice. Rosenberg and Donald (1995: 1122) define the approach as ���a process of turning clinical problems into questions and then systematically locating, appraising, and using contemporaneous research findings as the basis for clinical decisions���. Barton (2000) argues these considerations support a ���hierarchy of evidence��� with derivatives of quantitative methods ��� usually in the form of random- ized controlled trials ��� at the top and qualitative studies and ���opinion��� at the bottom. While the underlying case for EBHC is not new, having been forcefully argued for over a quarter of a century (Cochrane 1972 Maynard and Chalmers 1997), it is over the past decade in particular that EBHC, as a specific movement, has pursued the rigorous implementation of scientific research evidence into clinical practice (Sackett et al. 1996). In other words, it attempts to ���close the gap��� between the worlds of research and practice. The classical diffusion of innovations model (Rogers 1995) is an enduring focus of interest in this endeavour. Rogers (1995) identifies five stages in the stylized adoption decision: knowledge persuasion decision imple- mentation and coordination. Various criticisms have been made of such unitary, fixed and stage-like perspectives, particularly the difficulties emerg- ing over the power of such models to effect change within real-world health care settings (Ferlie et al. 2000 Fitzgerald et al. 1999). These deficiencies are important because the classic diffusion of innovations model continues to be used as a conceptual framework for modelling EBHC implementa- tion. Stocking���s (1992) review of change methodologies in health care explicitly draws on Rogers���s framework. Nutley and Davies (2000) draw on a number of authors in their review of the literature on the diffusion of innovations, but again Rogers is a major influence. More sophisticated models of the innovation process have been developed. Kimberly (1981), for example, argues that the career of managerial inno- vations is strongly shaped by the internal change capacity of the receiving organization and its external context. One neglected aspect of the EBHC project, however, remains the nature of the ���evidence��� itself. In diffusion of innovations models, what ���knowledge��� is appears unproblematic. Wood & Ferlie: Journeying from Hippocrates 49 at GLASGOW CALEDONIAN UNIVERSITY on March 5, 2012 oss.sagepub.com Downloaded from

Empirical work by Latour (1987) and Callon et al. (1992), however, stresses the translational and enrolment strategies performed by scientists in estab- lishing strong networks of association to support ever more powerful sci- entific theories. Knowledge is thus not seen as a necessarily independent source but rather the outcome of social, technical and political forces within clinical practice (Wood et al. 1998). Contemporary Theories on the Organization of Health Care Knowledge Truth as Correspondence As an epistemological approach to the organization of health care knowl- edge, EBHC relies on familiar dualisms: science/nature, research/practice, thought/action. With an attitude reminiscent of Plato���s doctrine of ideal forms, EBHC maintains the idea that research knowledge and practical activity are separate categories. Moreover, the sensible particulars of clin- ical practice are thought a reflection of universal science and truth. The diffusion of health care knowledge is classically treated as the flow of infal- lible messages to particular situations in a simply additive, or, at best, a two-way, process. Like Plato, present-day commentators consider this reflection to be an imperfect one. They routinely draw attention to a large and problematic ���gap��� between clinical practice and the findings of research (Haines and Donald 1998 Sheldon et al. 1998 Straus and Sackett 1998). One reason for this gap is the ���essence��� of the dialectical relation of correspondence between research and practice. In a world thought to be fundamentally static and immobile, achieving evidence based, clinical behaviour change is viewed as an often irksome, transitory phase between these two essen- tially separate domains. Informed by this immobility, EBHC is critically concerned with access to research information ��� witness the current interest in securing better, usually electronic, sources of evidence (for example, the Medline search engine and the Cochrane Reviews Database). An apparently important challenge for investigators and practitioners working within this premise is to stem the apparent degradation suffered in moving productive research along the ���knowledge chain��� to the con- sumptive contexts of practice. In other words, health care knowledge is organized around a truth-relation between research evidence and the real- ity of clinical practice that is often described using verbs such as ���agree���, or ���correspond���. The frustration is with the need to develop a more sophisticated under- standing of the relation between the two terms ���research��� and ���practice���. In organization and social studies of science, simple input���output matrices of knowledge flows have become increasingly questioned and as a result the traditional role of knowledge producers has been thrown into relief (Gibbons et al. 1994 Nonaka 1994 Sch��n 1995 Tranfield and Starkey 1998 Whitley 1984). This attention stems from a growing belief that 50 Organization Studies 24(1) at GLASGOW CALEDONIAN UNIVERSITY on March 5, 2012 oss.sagepub.com Downloaded from