Evidence-Based Medicine?

Related articles

[Editor's note: Paul Lee, in an article for SkepticReport.com, has argued that complementary and alternative medicine methods are by definition unproven and that we should prefer "evidence-based medicine" but Saul Green cautions that the term "evidence-based medicine" is often used not by responsible mainstream scientists but by CAM adherents who merely go through the motions of performing tests and gathering data without those tests and data producing reliable results in order to lend their work a veneer of legitimacy. TS]

Epistemology is the branch of philosophy that considers how we acquire knowledge and the limits of this knowing. Over the past decade, attitudes governing how we go about acquiring medical knowledge have changed. No longer does the phrase "the burden of proof" imply such criteria as plausibility and probability. No longer do conclusions from experimental trials have to be in accord with established, basic biological facts. By ignoring the criteria of plausibility and probability when designing clinical trials, investigators may promote the emergence of medical practices that are vacuous and unreliable. Having technically arisen from clinical trials, these practices can then be said to be "based" on experimental evidence and referred to as Evidence-Based Medicine (EBM) by their proponents.

Contrary to what medical consumers may believe, EBM does not have to embody some or any prior existing evidence. EBM's selective lack of accord with established facts from basic science, physiology, and clinical evidence has been noted in numerous papers in the Journal of the American Medical Association. Today's EBM may limit itself to results of a single clinical trial or to a small number of such trials. The flaws in such a system can be: l) lack of agreement among reviewers on the validity of the methods of treatment used, 2) lack of accord with standard values based on prior knowledge, 3) lack of means to discover errors, misrepresentations, selection of end points, ideological biases, and language distortion, 4) misuse of statistical analysis to determine the number of clinical trials necessary and sufficient to generate valid determinations of effectiveness.

As a result, today's EBM represents a shift away from achieving scientific validity through rational analysis and knowledge. It is a system that tolerates incomplete research results, reported as revelations. EBM is practiced today because its proponents and reviewers believe that the specific formats they use for reporting trial results assure the validity of their conclusions. Not surprisingly, today's EBM is at the heart and soul of many so-called "complementary and alternative" medicine practices.

Saul Green is science editor of the Scientific Review of Alternative Medicine and president of ZOL Consultants. He can be reached for more information at saul[at]ssr.com


June 18, 2003

To the editor:

I'm not exactly sure what you mean in your editor's note atop Saul Green's article. I certainly don't want anyone to think that I disagree with Saul, for whom I have the utmost respect. If there's a conflict, then I defer to him, but I don't believe there is. I think we're talking about two different things.

I am talking about true, well-done EBM, which I indeed "prefer" (your words, not mine), in contrast to altmed nonsense. He is talking about altmedders misusing the term and calling lousy research EBM. In other words, when they use the term for their own so-called research, we can't trust the term "evidence-based medicine" to mean what it should mean. I don't think we should abandon the term because it gets misused, but we should clearly point out when altmedders are misusing it. Otherwise we're going to allow them to call the shots, causing us to throw the baby out with the bathwater.

I fear that his comments, if left to stand alone, will lead some to think that EBM is a less than worthy endeavor. I doubt that Saul believes such a thing. Just because some quack can produce some statistics after treating a few patients doesn't justify them in calling their results "EBM," nor their quack treatment-for-payment "research." One of the areas where I believe that legitimate scientists are in danger of misusing the term is in connection with meta-analyses. Poorly done, they can compare apples and oranges. Just because someone counted and measured them in an RCT doesn't mean the conclusions of the meta-analysis are valid. The Cochrane Collaboration attempts to avoid this danger.


Paul Lee, PT