The slow death of evidence based medicine (EBM)
I wrote quite a few articles (EBM) many years ago on the inadequacies of the EBM sect and their dogmas. My take is that EBM is now slowly dying as a serious attempt to understand the natural world or clinical practice except at an institutional level. EBM remains popular with bureaucracies such as the NHS or NICE because they can use it to control medical care much as organisations like ATOS seek to control disability. A recent loose phrase in Archives of Dermatology prompted me to fire off a letter. My intro was:
I enjoyed greatly the article in the Archives by Gilchrest and Martin1 about some of the pitfalls of evidence-based medicine. The article was insightful and nuanced, so much so that I cannot but believe that their first sentence was written with a degree of irony. The authors state that “No one can argue about the merit of evidence-based medicine.”1 (p528) Well, of course, the medical literature is full of cogently argued expressions of doubts about the epistemology of evidence-based medicine (for a recent review, see Goldenberg et al2 and all articles in that same volume of Perspectives in Biology and Medicine devoted to this subject).
I ended with a refrain of Bradford Hill’s views. It is ironic that medical statisticians in the golden age of medical discovery had a much more nuanced view of medicine than their successors.
Second, RCTs, as any clinician will tell you, are only approximations of everyday clinical practice. That is why they are called experiments —they abstract some of the components of everyday practice to test particular generalizations. This is both their great strength and their weakness. Look at the terminology that surrounds them: we talk about testing or rejecting a hypothesis —what we do not (and should not) say is that a trial is about whether to treat any particular patient in a particular way. Randomized controlled trials are not facsimiles of everyday practice but are rather experiments that, if done well, provide guidance for our clinical practice. But all experimentation is, if you will, a hazardous attempt to trick nature into revealing her secrets to us—just how hazardous has only recently become apparent from the work of John Ioannadis4 and others who have argued (admittedly with some excess enthusiasm) that most findings from RCTs are likely insecure knowledge or just plain false.The interested reader may consult the references cited herein, but it is worth remembering also the words of Sir Austin Bradford Hill (who many consider the father of the RCT) in his 1965 Heberden oration: “Any belief that the controlled trial is the only way [to measure therapeutic efficacy] . . . would mean not that the pendulum had swung too far but that it had come right off its hook.”5 (p108)
John Hampton, who IMHO, made major contributions to clinical science, in an earlier piece stressed the importance of RCTs, and wrote ‘Clinical freedom is dead, and no one need regret its passing’ (Hampton JR. The end of clinical freedom. Lancet 1983; 287 pp1237-8). More recently he wrote:
So we seem to have the perfect storm, where a meeting of evidence-based (which we ought to call opinion-based) proscriptive guidelines, mechanistic doctors and financial control have come together to contribute to the demise of the responsibility that doctors used to have for individual patients. We need to change medical culture in such a way that doctors can use their opinions about published evidence to select the best treatment for each individual patient. We need a return to clinical freedom.
The two groups of corporations that dominate medicine— pharma and the large health care corporations such as the NHS— could not have wished for a better vehicle than EBM to corporatise control over the sick.