The appliance of science?

Evidence-based medicine (EBM) has become the Big Idea of the 1990s NHS, and is beginning to make an impact on continuing medical education, policy making and the health services research agenda. Since no clinician can object to the idea that medical treatment should be based on the best available evidence, the ideal of EBM should not be contentious, and has provoked such enthusiasm, as well as resistance, that it is becoming the biggest breakthrough in medicine since...well, lunchtime probably. Why should EBM, the by-product of a North American approach to medical education, suddenly become such an attractive idea?

EBM has become powerful because it functions in three ways. It is a substitution technology, replacing one method of learning with a cleaner, truer method. EBM offers clinicians an educational mechanism based on a structured and systematised scientific method of analysing knowledge itself, and leads away from educational approaches that have been commercially biased (by the pharmaceutical industry), dominated by specialist knowledge (in the teaching hospitals and the traditional research community), or shallow (based on anecdote and case series). The progressives in the NHS like the slant away from vested interests and a corrupt power structure, while traditionalists welcome the fact that EBM keeps scientific developments within the scientific community and excludes inconvenient interest groups like patients or local communities.

It is a replacement technology because it extends the market for useable scientific knowledge to two important groups in healthcare: public health physicians and managers. Public health physicians have, through EBM, found a policy tool for guiding decisions about purchasing clinical care, at a time when cost containment pressures demand the rationing. Similarly, health service managers have in EBM a tool for controlling clinical activity, something they have long wanted but never had.

Finally, EBM is a placement technology that creates an entirely new market for ways of understanding knowledge itself rather than of applying scientific techniques. Through EBM clinicians will all learn to how to ask questions and how to answer them, becoming practitioners in the philosophy of science. To do this they need teachers and training, books and courses, journals and seminars. All these needs will be met by new kinds of educators and scientists, whose expertise will suddenly become essential to the NHS. A new sub-profession is in the making, with a high priesthood of super-scientists trained in the seminaries of Oxford and York.

So EBM should triumph. But will it? EBM has three major limitations: it is a vehicle for reductionist thinking; it allows decision making to remain a professionally dominated; and it substitutes epidemiological perspectives for a broader understanding of scientific method.

First, the temptation to reduce the complexity of clinical practice and consider, for example, only the treatment of a ‘primary diagnosis’ neglects the problems of caring for patients with multiple pathology, or without diagnoses at all. It also leads to considering only a those treatment outcomes for patients that are easily quantified.

Second, categorising interventions by their effectiveness makes implicit value judgements. Who decides on the outcome? Does lack of evidence mean lack of value? Wise clinicians know that some patients opt for treatments known to be less efficacious than others, which purchasers may fund, because patient choice may reflect a judgement of benefits and risks more subtle than that of researchers.

Third, the dependence of EBM on epidemiological methods may lead to the devaluation of the unquanitifiable in clinical practice, but may also distort understanding of the quantifiable. EBM’s reliance on the results of randomised controlled trials leads its advocates into the classic ecological fallacy - that group averages can tell clinicians what they need to know about causal processes in individuals.

EBM is flawed, but it allows purchasers to appear as informed consumers, and could become an instrument for the regulation of public provision of medical care. That is why it will become embedded in NHS management as well as in academia, long before evidence of its effectiveness emerges - if it ever does.