One of the key criticisms of the EBM movement as it has taken shape is that the hierarchies the movement has embraced have placed clinical expertise and localized knowledge at their bottom. This was a reaction to one of EBM’s general aims, to move medicine away from the old habit of simply following authority and into a new era of using good evidence to guide medical decisions. But clinical expertise and blindly following or expressing authority are not the same thing, and I and others think it has been a mistake of EBM to take that assumption and run with it into the arms of the RCT.
Rather, the strategic thinking and use of tacit knowledge within the clinic, at its best, is an expression of learned skills in thinking about the patient context and in identifying what is the best evidence at hand.
This brings us to the question of how to make the best evidence available to the thinking clinician. Recent developments in the practice of EBM show that there is a growing awareness of the need for clinical expertise, and consequently a need to incorporate the role for such expertise within EBM approaches. For instance, GRADE recently called for a distinction to be made between the kind of evidence that can be rated according to a hierarchy, and the kind of evidence that should not be. Some recommendations made by guidelines panels, that is, ought to be written as ‘good practice statements’, because even though the evidence for them do not meet the usual GRADE criteria, they are nonetheless important guidance for clinical practice (Guyatt et al., 2015). Such recommendations arise out of accepted expertise, or indirect evidence, rather than out of the results of RCT studies, for instance, but we can still be certain that they will lead to better results if adopted. This can be taken as an acknowledgement that the best evidence can come from diverse types of knowledge about what counts as best clinical practice.
A similar situation can be found in the research realm. Many medical discoveries have and still occur as a result of unexpected observations or case reports that, in turn, inspire new research directions and present as evidence for recommendations for practice. These individual cases are still a kind of evidence, even before they are tested by large trials. A thoroughly evidence-based approach to medicine is able to incorporate such examples of good evidence despite their lower place on the traditional hierarchy of methods for producing evidence.
It can seem like the people who make unexpected observations and then follow them up to show their value as evidence are working on instinct or from intuition. But if we look more closely we can see that such intuitive responses to potentially good evidence are the result of many years of experience and the development of reasoning skills about what kinds of things to pay attention to and what kinds of things should be ignored. These skills develop over time through experiences of success and failure, whether the people are basic scientists, clinical researchers, practicing clinicians, and even patients.
Rather than focusing exclusively or primarily on how to make better trials and how to better disseminate the results of such trials, then, EBM proponents should also be paying attention to how the evidence from individual or small series of cases is recognized and distributed. While this is a case of tacit knowledge, the kind of knowledge that is difficult or even impossible to translate into methods or procedures that can be written down, it is not impossible to assess the strategies employed by the people who have such knowledge. Presumably, if GRADE has confidence that good practice statements can be distinguished from bad ones, they are not also suggesting that we merely rely on the authority of people who claim they are good, but rather that we can develop methods for sorting such things out as a group.
Therefore, any program that wants to ensure the best available evidence is used in practice ought to abandon approaches that assume clinical expertise is somehow mysterious, or that it only involves interpreting how trial results should be applied in the particular case. Rather, it should embrace clinical expertise as a source of good evidence. Taking a critical and analytical approach to understanding the kinds of reasoning processes and strategies employed by the people who recognize an instance of good evidence when it arises unexpectedly, in a single case, or as a pattern of observations, can broaden our understanding of what good evidence is. As well, it may introduce new and valuable approaches for EBM itself that will help it to work better in practice.