According to its advocates, Evidence-Based Medicine (EBM) represents an attempt to democratize health care (EBM Working Group 1992; Sackett et al. 1996, see also Oliver and McDaid 2002) by providing both patients and clinicians with better quality, objective information about the performance of medical procedures, pharmaceuticals and treatments. It aims to furnish a systematic review of medicine: a "gold standard" in "best practice."
EBM has to date, had a spectacular history. Prior to 1992, the term was absent from the medical literature; yet in fifteen years the terminology and its emphasis has become embedded within Occidental medical education and research--it has taken the medical community by storm (Kristiansen and Mooney 2004). Its emergence and subsequent rapid adoption surpasses that even of the notion of social capital in the social sciences. Seldom in the history of science and scientific practice can an emergent approach have been adopted so readily; indeed, converts occasionally adopt an evangelical zeal.
Despite EBM's laudable objectives there are sound reasons why caution should be exercised regarding its methods, adoption, subsequent proliferation, and potential to invoke institutional change. Extensive criticisms from medical, methodological, ethical and even (mainstream) economic perspectives are emerging. Yet institutional economics has largely been absent from this discussion and analysis. The aims of this paper are two-fold: to highlight the lacuna in the institutionalist literature, and second, to note the potential for insidious institutional change.
At the outset of the arguments presented here two qualifying notes are made: First, the term "instrumentalism" is not employed in the manner defined by John Dewey; instead I adopt its more positivistic and utilitarian rendering in referring to actions and theories as possessing no intrinsic value--their value lying in their consequences. Second, clinical-medical care is the preferred narrative, since care of this nature is usually provided in a particular institution centering on the social relationship between clinician and patient. By contrast, health care broadly conceived involves a richer milieu of institutions not necessarily focusing on the clinician-patient relation or the clinical-medical encounter.
To my knowledge, Keaney and Lorimer (1999) offer the only institutionalist analysis of the impact of formalized clinical governance (something that EBM is closely associated with, certainly in the UK). Their valuable contribution analyzed the clinical-medical system in Scotland. This paper endeavors to present a more abstract and theoretical consideration of the potential nature of institutional change associated with EBM. It is explicitly recognized that EBM is not the only initiative or source of institutional change in clinical-medical care (for an example of other initiatives encouraging instrumentalism, see McMaster 2004), and that the adoption of EBM principles varies considerably, both within and between clinical-medical care systems. Moreover, the paper by no means is intended to provide a comprehensive analysis of the rise of EBM and its impact on clinical-medical care institutions and systems. Much speculation could be, and doubtless has been, exercised attempting to rationalize EBM's incredible trajectory. Plausible explanations include: EBM representing an innovative progression, due in large part to advances in computing technology, of established clinical practice; a consequence of the rise in consumerism and decline in trust in the clinical professions; an attempt to reduce medical errors in an increasingly litigious society; an aspect of wider fiscal pressures and attempts to contain costs, and a manifestation of wider social trends that promote scientism, or seeming objectivity, and greater quantification (see, for example, Aasland 2001; Birch and Gafni 2004; Kristiansen and Mooney 2004; Miles et al. 2004, and more wide-ranging commentaries on scientism and the culture of quantification, such as Dupre 2005; Porter 2004; and Shenhav 1999). Instead, the paper attempts to make some reasoned conjectures to highlight the emergence of an agenda requiring attention.
These caveats in mind, it is contested that EBM privileges a particular type of evidence--randomized control trials (RCTs) and meta-analysis--that acts to relegate other forms of medical evidence. EBM has the potential to structure change by changing the dialogical context of the clinical-medical encounter into a reductionist rule-based practice grounded in the bio-medical paradigm, which relegates the significance of social, historical and even individual contingencies of health. This carries the capacity to objectify the individual patient, thereby erasing the individual from consideration, as the general is applied to the particular. EBM, then, potentially erodes deontological considerations and contexts of practice. The focus is on outcomes with procedure adopting an instrumental character. Concisely, the social construction of knowledge and illness are at best attenuated in a narrative reflecting an overarching mechanistic frame. This is far from inevitable, as professional medical practice draws upon a plethora of repertoires predicated on culture, social context and space; yet the trajectory remains.
The following section briefly outlines the properties of EBM. The third section reviews the nature of the medical and standard health economic critiques, while the following draws on the approaches of Dewey, Charles Sanders Peirce, and Douglas Walton in particular as bases for questioning the claims of EBM advocates and for conceptualizing the nature of institutional change that may be prompted by EBM. The penultimate section sets out a range of theoretical conjectures on the institutional impact of the unfettered adoption of EBM.
EBM as the Chosen Path? A Brief Resume
Canadian-based clinical epidemiologists coined the term EBM in 1992 in their vigorous and successful promotion of a "new paradigm for clinical medicine" (EBM Working Group 1992). As one of its leading proponents, David Sackett, contends, the principal aim of EBM is to equip clinical practitioners with the means to cope with a fluid technical environment, and to expedite the dissemination of information, thereby offering the prospect of a "gold standard" in health care. Hence, EBM, its proponents argue, offers the ultimate source of medical knowledge. It provides the norms or guidelines, via quantitative techniques, of best practice, and thereby democratizes provision by furnishing both patients and clinicians with objective evidence about the performance of medical pharmaceuticals, procedures, treatments, etc.
EBM has been remarkable in its rapid adoption by the medical profession: in less than fifteen years it has initiated new journals, and is a core element of the undergraduate medical curriculum in North America, the UK, and other parts of Western Europe (Kristensen and Mooney 2004; Mykhaiovskiy and Weir 2004), and is claimed to offer medical students enhanced skills in using evidence (see for instance, Dorsch, Aiyer and Meyer 2004; Fritsche et al. 2002; and Rajashekher et al. 2002). EBM is also a central constituent element of clinical audit, an example of systematic review promoted by recent reforms in the UK's National Health Service (NHS).
EBM and systematic review may be traced to the Cochrane Collaboration (Donaldson, Mugford and Vale 2002). This attempted to facilitate the translation of new evidence on clinical procedures, technologies and pharmaceuticals to clinicians on an international basis. It was, and is, an attempt to feed through new evidence in a manner that will change clinicians' behavior in order to enhance patients' health (Donaldson, Mugford and Vale 2002). The Cochrane Collaboration is the result of the work in the late 1960s and early 70s of the eminent medical researcher, Archie Cochrane, who advocated the increased scrutiny of the medical profession through the incorporation of empirical evidence (Donaldson, Mugford and Vale 2002; Kristensen and Mooney 2004; Whynes 1996). For Cochrane, the primary data source to increase the flow of reliable information for clinicians was the RCT. In an RCT a group of patients is separated on a random basis into control and experimental groups. These groups are followed up for the outcomes (of treatment, procedures, etc.) and variables of interest.
Through the dissemination of the results of RCTs, Cochrane argued that the evidence base of clinical procedures would be greatly enhanced, and, moreover, would enable greater efficiency as well as effectiveness (Donaldson, Mugford and Vale 2002), (1) as well as harness clinical knowledge more effectively by addressing the information overload relating to technological advances and on-going accumulation of evidence in all areas of medicine that threatens to overwhelm clinicians. As with Cochrane, RCTs are the principle means of collecting evidence in EBM, although this can be supplemented by meta-analysis of all published RCTs, i.e., where small group sizes are pooled to make larger ones. EBM is explicit in its invocation of a hierarchy of evidence and information, with RCTs (and meta-analysis) ranked ahead of other information and evidential sources, such as observational data sources derived from case control and cohort studies, and experiential evidence (see for instance, Miles et al. 2004; Gupta 2003; Sackett et al. 1996; Villanueva-Russell 2005. See also the Appendix for a range of definitions and gradations of evidence, and subsequent levels of recommendations issued by various health bodies). Following this if the evidence for a particular procedure is favorable then a further function of EBM is to inform policy (Oliver and McDaid 2002), and recommendations--resembling if x, then y--for medical procedures (see Appendix). Indeed, the prominent health economist Alan...