The evidence‐based health care debate – 2006. Where are we now?

When an academic in the History of Medicine comes to write a comprehensive account of the nature, characteristics and scale of contribution to human progress of the evidence-based health care movement, there will indeed be a great deal of considerable interest to read. Rarely before, it seems, has there been such fierce and greatly protracted polarization of both scientific and clinical positions as in the last 16 years since the coining of that now severely tired, and almost defunct neologism, ‘evidence-based medicine’ (EBM) (Evidence-Based Medicine Working Group 1992; Polychronis et al. 1996a,b). The Journal of Evaluation in Clinical Practice, over this time, has achieved a pre-eminent reputation in directly ensuring that the necessary debates on the nature of ‘evidence’ took place and, indeed, that they continue to take place. If one accepts, as is now essentially universally accepted, that clinical medicine and health care more broadly, are to be developed and evaluated according to the extent to which they are in alignment with widely accepted concepts and norms of practice, themselves continually evolving through highly reasoned debate and careful observation, then it is the duty and ambition of any scholarly periodical with a definitive interest in the evaluation and development of that most human of activities, the care of the sick, to place itself at the centre of such debate. While disliking the term ‘mission’, given its inane and intellectually bankrupt modern usage, and preferring the classical understanding of the term, this is precisely the ‘mission’ we, at the Journal of Evaluation in Clinical Practice have adopted and, on the basis of which mission, we will continue to operate, within the environment of complete editorial independence that Blackwell Publishing have always rightly afforded to their Editors. We had thought that the practice of publishing an annual Thematic Edition charting the progress of the EBM debate had ‘worn thin’, as it were, and that an alternative form of presenting the many papers we receive on EBM was warranted – a recurring section, perhaps, in successive issues of the Journal. But the reaction to the Editor’s announcement that the annual Thematic Edition was to cease, met with no small protest and so the decision to continue the publication of the annual Thematic Edition has been taken. We are therefore gratified to commit to the international medical literature the present Part One of the 9th Thematic Edition (Volume 12, Number 3) on the progress of the EBM movement, which augments the previous eight (Miles et al. 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004) and, at the time of writing, we are completing the editing of Part Two of the present Thematic Edition, to be published as the next issue (Volume 12, Number 4). This very large opus having been achieved, we invite contributions to the 10th Thematic Edition for publication in 2007. Starting with an analysis of concepts, we have made the centrepiece of this edition Tonelli’s article: ‘Integrating evidence into clinical practice: an alternative to evidence-based approaches’ (Tonelli 2006). Tonelli (2006) notes that despite the lack of good evidence that teaching EBM improves the quality of medical education or the subsequent care of patients, the vast majority of internal medicine programmes in the USA now incorporate some aspect of EBM into residency training and a third have a free standing curriculum in EBM. The author additionally notes that while such a widespread adoption of EBM into graduate medical education promises to alter the way that the next generation of doctors practise clinical medicine, it remains quite unclear as to whether such a change will ultimately benefit patients. An extraordinary observation indeed. The Journal of Evaluation in Clinical Practice has documented the same such observation on countless occasions previously and continues to share Tonelli’s elegantly expressed concern that, although the bulk of the literature on EBM focuses on the practical issues related to the development, acquisition, interpretation and incorporation of the results of clinical research into clinical practice, EBM continues to rest on certain philosophical assumptions and arguments about the nature of medical knowledge that have still not adequately been elucidated (see, for example, and noting the substantial bibliography within: Miles et al. 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2006; Norman 1999, 2003). Indeed, and as Tonelli (2006) points out, although now at last acknowledging the need to integrate various kinds of what they continue to regard as medical knowledge with what they continue to regard as ‘non-medical knowledge’ in the making of clinical decisions, the protagonists of EBM have said little about how such integration should actually take place and it is precisely this process whereby ‘evidence’ is applied to decision making that is most under-described by EBM-influenced educational curricula. Thus, Tonelli (2006) argues that it remains unclear by what process clinicians are to balance what may be described as conflicting ‘warrants for action’, deriving not only from published clinical research, but also from personal clinical experience, pathophysiological understanding of disease, the preferences of individual patients and other such sources which have typically been regarded by the protagonists of EBM as being ‘non-evidentiary’ in nature. Tonelli (2006) proceeds to argue that the very few attempts by the protagonists of EBM to describe a method for integrating various kinds of medical knowledge and reasoning into clinical decision making have all, without real exception, proved unsatisfactory. The failure of the EBM approach centres, this author feels, on its attempts to treat different potential warrants for medical decision making, such as empiric evidence, clinical experience and pathophysiological rationale, as different in degree, rather than different in kind. Tonelli (2006) argues that, under such a rationale, one form of medical knowledge, specifically that derived from clinical research, is judged (wrongly in our view and in his) as superior to the others and he argues that this approach is philosophically untenable and that various potential warrants for medical decision making differ in kind from one another. It is this understanding of the nature of medical knowledge, Tonelli (2006), contends, that requires an alternative method for integrating various warrants into a particular medical decision, a method that for him closely resembles the casuistic or case-based approach to medical ethics advanced by Jonsen et al. (1992). For Tonelli (2006), this understanding of medical epistemology can provide for an explicit defence of a more ‘common sense’ practice of medicine that has been called for time and time again (Porta 2004). We regarded Tonelli’s thesis to be of such importance that we took the unusual step of commissioning not one, but rather 11 commentaries on this author’s reasoning, from a wide range of scholarly sources, spanning a large range of disciplines, both within, and applied to, Medicine, and it is this most important set of articles which follows Tonelli’s piece. The commentaries by Djulbegovic (2006), Miettinen (2006), Porta (2006), Lipman (2006), Tanenbaum (2006), Couto (2006), Upshur (2006), Loughlin (2006a), Malterud (2006), Gupta (2006) and Geanellos & Wilson (2006), make fascinating reading and to exercise editorial comment here on these editorially commissioned commentaries would be superfluous. Instead, we are content to advance these articles as essential reading in conjunction with Tonelli’s work (Tonelli 2006) and we remain excited by the way in which they have substantially developed the primary insights which Tonelli records, each commentary providing considerable scholarship in its own right. In previous Thematic Editions on EBM, the Journal of Evaluation in Clinical Practice has studied contentions relating to the so-called ‘ethics of EBM’, publishing papers by colleagues who have argued for the existence of an ethical imperative to implement the tenets and methodologies of EBM into routine patient care. Most notable among these have been the articles contributed by Gupta (2003, 2004) learned commentaries on the same (see Couto 2003; Miettinen 2003; Norman 2003; Shahar 2003) and an incisive Essay Review by the distinguished health philosopher Michael Loughlin (Loughlin 2003). We turn now to the article in the present issue by Borry et al. (2006) and its related commentary ‘A platitude too far’. (Loughlin 2006b). The central arguments advanced by Borry et al. (2006) relate to the ‘need’ to base clinical decision making on the latest and ‘best available’ medical research findings, with the clear implication that should such clinical decision making not be based on such data, it would, ipso facto, be less than ethical, if not directly unethical. For Borry et al., the use of the latest research will automatically result in decision making that is more ethical, because it will therefore be ‘better informed’ and ‘better justified’. For his part, Loughlin (2006b) is able to agree with the authors that one should not fully support the call for a misleading extension of the term ‘evidence based’ to the ‘field’ of ‘ethics’. In fact, he is clear that one should not support it at all and he shows the authors’ overall thesis to be fundamentally defective because it fails (or refuses) to acknowledge the fundamental fallacy embodied by the use of the terms ‘evidence’, ‘evidence-based’ and ‘evidence-based medicine’ in the context of medical epistemology. In the analysis which follows, Loughlin (2006b) goes on to study, in no small depth, both of the two central assumptions of ‘evidence-based ethics’ put forward by the authors. We find ourselves in agreement with this commentator when he argues that the last thing we currently need is the creation of a new ‘field’ of so-called ‘evidence-based ethics’, ‘with training days to explain to us why we are all unethical because we fail to conduct our practices in terms of whatever meaning the powers that be care to attribute to the term “evidence- based” on the day’. Indeed, as Loughlin (2006b) points out, such developments can harm proper professional practice ‘as competent professionals (are) labelled immoral for refusing to “genuflect” to a variety of academic altars in the absence of any demonstration of the true nature of whatever academic deities – be they labelled “quality”, “evidence” or even “ethics”– these altars supposedly represent’ (see Loughlin 2002, 2003). Following on, we move to four articles that focus specifically on the limitations of ‘evidence-based public health’, ‘evidence-based psychiatry’ and ‘evidence-based cardiovascular medicine’. In the first of these articles, Kemm (2006) examines how the concept of the ‘evidence-based’ approach has transferred from clinical medicine to public health and has been applied to health promotion and policy making. He notes that in policy making, evidence has always been interpreted broadly in order to cover all types of reasoned enquiry and after some debate the same is true, he observes, for health promotion. For Kemm (2006), taking communities rather than individuals as the unit of intervention, and given the importance of context, means that randomized controlled trials are frequently inappropriate for the study of public health interventions. This author goes on to suggest that the notion of a ‘best solution’ ignores the complexity of the decision-making process. Interestingly, and as we agree, Kemm (2006) is clear that the availability of evidence ‘enlightens’ policy makers in confronting their policy problems but does not in any way provide a direct answer to the given problem(s). He concludes, persuasively, that there are lessons from the way that evidence-based policy is being applied in public health that could usefully be taken back into clinical medicine. Kemm’s analysis of the limitations of the EBM approach in public health medicine leads us to Maier’s discussion of the incompleteness of the concept and practice of EBM in psychiatry (Maier 2006). While Maier is clear that ‘evidence-based psychiatry’ is an important concept and might prove with time and development to be an useful method, he is equally clear that its application currently covers only a limited range of the clinical problems that psychiatrists and indeed psychotherapists encounter within their daily work. By way of specific example, the author describes the limited validity of psychiatric diagnoses and notes how EBM, which is fundamentally based on correctness of diagnosis (a limitation we have discussed previously), has an almost automatically doubtful applicability in psychological medicine. Moreover, there is the issue of ‘complexity’, the effects of which are more obvious in psychiatric and psychotherapeutic cases than perhaps in other branches of Medicine. In this particular context, Maier (2006) argues that, owing to its conceptual bases, EBM cannot sufficiently consider such effects and is therefore of limited usefulness when clinicians address particular problems in fluctuating or symptomatically shifting diseases. Additionally, the predominant focus of EBM on decision making simply does not reflect the clinical reality of psychological medicine where diagnostic and therapeutic procedures are in fact more iterative, with decisions being constantly re-evaluated. This highly individual approach to therapy means that, ipso facto, EBM cannot assist clinicians in achieving anything which could be described as optimal psychiatric clinical care. This brief focus on EBM and psychiatry continues in Dewhurst’s short account of its limitations, but from the particular perspective of personal experience of illness – valuable insight indeed for any clinician who may treat the disease or condition from which he himself has suffered (Dewhurst 2006). The fourth and final paper of this section is concerned to examine the relationship between evidence-based treatment of heart failure and the resulting quality of life of the patient. Here, Dobre et al. (2006) report the results of their study which explored whether prescription of evidence-based drug therapy is associated with better quality of life in patients with heart failure. The authors were able to observe that quality of life measurements did not, in fact, differ significantly between evidence-based and under-treated patients, unadjusted or after adjustment for significant patient imbalances. On this basis they conclude that conventional step-up medication approaches in heart failure may well have a positive impact on survival and morbidity but do not necessarily translate into improvements in the patient’s quality of life – a noteworthy observation indeed. Wherever there is a consensus on the definition of what exactly constitutes evidence for a given clinical practice – and such a consensus is usually, and entirely properly, reached only after sufficient time has elapsed to enable adequate debate by experts in the given field and for adequate experience of the operational use of an intervention in ‘hands on’ clinical practice to have accumulated – then there is a clear imperative to move to of Indeed, that is the basis on which clinical practice evaluation and development and place. In the of this section of the & (2006) describe how experience of knowledge in care can useful and to implement evidence-based practice in clinical as they a process for the and application of knowledge in both knowledge the authors as and explicit knowledge the authors describe as which they as within of The benefit of the approach that these authors is that it fundamental account of and they argue for more study of how such methodologies can be into modern health care and the of so Clinical practice when certain are met and their limitations fully useful for to clinical practice and the process of deriving and clinical practice has developed into a in its own so and has this field of study The Journal of Evaluation in Clinical Practice has published on this over the last 11 that the proper place of practice is one of as of an of clinical evaluated and within the context of the experience of the individual patient in the of the (Miles et al. The of the methodologies for practice development, in one in the of of the of practice of highly limited in clinical practice and in many the limitations of practice are their The paper in this section, in the process of the of for development in where the conduct of trials would be or unethical et al. and indeed usefully to on development and In the article of this final section of the Thematic Edition, & (2006) are concerned to the of ‘evidence-based’ medical For these the evidence base for most educational has been essentially of evidence with four to its The authors the adoption of a approach in the evaluation of educational that a comprehensive of under one and the range of applied of use in these which from the and of and that the research and a set of or and they describe the use of clinical decision support in the of EBM. In the paper which follows, et al. (2006) report the results of their study of the effects of an educational in on the of back and we conclude the issue with a contribution from & (2006), the knowledge and practice of ‘evidence-based medicine’ between doctors with their We have previously that in their so and with such the protagonists of EBM have a most important in quite what good clinical medicine is not about (Miles et al. 2004). In direct it is now clear that evidence-based clinical practice as by and good clinical medicine, cannot be and remain in our view fundamentally Thus, EBM is currently to be as a or of clinical practice, and in nature, with no evidence for a superior clinical or patient these quite central and fundamental and by an it extraordinary that so continues to take that so many medical education continue to change with to EBM and that so clinical practice has been – all on the basis of a concept by an lack of The EBM so set in as at the time, are now in such that it is currently to quite what colleagues who still describe themselves as ‘evidence-based actually now the concepts and approaches have into many in the of and clinical and as the understanding of the true nature of what constitutes evidence for clinical practice has It is indeed that in so on a definition of evidence, the of the directly a more fascinating new insights into a very In that they have made an even the process that they in the of their It perhaps, the that the et al. that their taking well for having developed medicine in a way that they did not quite or to the EBM was indeed from and where EBM continues to be its to remain based on some need for in the of clinical as has been argued in the Journal of Evaluation in Clinical Practice and is not to deal with the of at all, but on the a more to result in care at the and and at the than the exercise of what has been described as clinical medicine with its on experience and that one has to experience the use of the ‘evidence-based’ in to health and, indeed, even in the academic On the a large of it has to be colleagues continue to use the with a understanding of the proper use of that an observation all too made when they are with an to the latest is the of ‘evidence-based can explain such a the almost ‘evidence-based’ continue to highly but it is that some colleagues lack the necessary to in order to it for what it (Loughlin it EBM continues to for these two very different the almost ultimately for clinical and the almost some colleagues to experience when or the If these are then there is a clear need to continue the debate. Tonelli’s article in the present edition (Tonelli and the associated set of 11 learned directly to this debate. The wide range of made and within these articles individual study if the debate on EBM is to move forward from its We fully the protagonists of EBM to their for proper academic and to remain in the of the of considerable scholarship on the limitations of EBM contributed to the literature of academic medicine by the present If we are proved then we would indeed be and the to answer their The of scholarly on the various of the EBM debate to the Journal of Evaluation in Clinical Practice in the last has that we have been to in the present issue all articles of to the debate that we would have Thus, the articles will be published in the next issue (Volume 12, Number which will in Part of the Thematic and argue our here the content of that the may be to that in the edition which follows, we the of and also of that may or directly result in which to the concept and practice of EBM by learned in the we and while at the same time and debate. et al. 2006; Upshur 2006) to from the of into the of the and of the into a call for a process on which we in the next