Is Evidence-based Medicine a Dead End?

A recent blog post promoted by the American Holistic Veterinary Medical Association (AHVMA) asks the question, “Is Evidence-based Medicine at a Dead End?” Since scientific evidence often fails to support the beliefs and claims of alternative vets, and the AHVMA has demonstrated many times that it accepts only evidence that supports the approaches it promotes, it is no surprise that the article contains a variety of tired misconceptions about EBVM and unproven assumptions about CAM that lead to the inevitable conclusion the author wanted to arrive at.

The first step is to suggest that EBM has somehow diverged from the presumably pure state in which it originated under the impetus of the “founding father” of EBM, Dr. David Sackett:

Evidence-Based Medicine (EBM) as practiced today (and not as originally conceived by Sackett), emphasizes fact-based medicine.

I suspect Dr. Sackett would be surprised to hear that emphasizing facts over beliefs and opinions is somehow a departure from his vision for EBM. The very fact that this can be used as a criticism of EBM highlights the disdain CAM practitioners and the AHVMA have for those pesky facts, and their preference for belief and pure faith-based medicine.

Large numbers of responses are analyzed, so this is valid for populations as a whole. In veterinary medicine this is called herd medicine, and decisions are made that will best protect the whole herd.

This is intended to build up to the claim CAM practitioners often make that because scientific research involves studying groups the results aren’t applicable to individuals, who are all unique. This is a common fallacy I’ve covered before, and which involves both the Vegas Delusion and the Snowflake Fallacy.

The Vegas Delusion is the idea that because statistics apply to groups and the outcome of any even for an individual is not perfectly predictable, we can ignore statistics in making decisions. The name for this fallacy comes from the fact that gamblers often use it to justify their hopes of winning despite the odds being overwhelming that they will lose. It is true that some individuals win at games of chance, and sometimes they win big. But casinos are lucrative businesses and gambling is a problem that ruins lives because the statistics do predict what will happen to most people most of the time.

Similarly, in medicine the prognosis or response to treatment can’t be perfectly predicted for any individual patient. But the averages that affects groups are often a useful and realistic guide to what will probably happen, and ignoring these to make up treatments haphazardly for each individual patient is a dangerous and unreliable way to practice medicine.

The Snowflake Fallacy refers to the belief that because every individual is unique, we cannot use information about a group of patients to guide the treatment of any particular patient. While it is true that we are all unique in many ways, we are also very much alike in many ways. And often the differences that we notice immediately, in appearance or behavior, aren’t relevant to how we respond to medical treatment. The significance of such differences has to be demonstrated with good research, not simply assumed or made up.

There is also a lie inherent in the claim that CAM individualizes treatment more than conventional medicine because it doesn’t rely on controlled scientific research. CAM practitioners do use information about groups of patients to guide their treatment of each individual, they simply use the haphazard and uncontrolled personal observations of their patients and of other CAM practitioners instead of scientific research. A homeopath who chooses an “individualized” remedy from a repertory is doing so based on patterns of symptoms and responses to treatment observed in other patients by other doctors. This is applying group results to individuals, it is just doing so without any effort to control for bias, placebo effects, and other important sources of error.

The author then goes on to point out that it is often difficult to control for placebo effects when testing non-pharmacologic therapies, such as acupuncture. Because of this, she believes EBM automatically finds fault with research in such therapies and so denies their obvious benefits unfairly.

While it is true that it is challenging to control for placebo effects in such approaches, it often can be done. In the case of acupuncture, needling at locations not considered to be “real” acupuncture points, or not needling at all but simply convincing the patient you have, often has just as much a clinical effect as verum acupuncture (1, 2). This most likely means that the “real” acupuncture is only an elaborate placebo, but acupuncturists are unwilling to accept this conclusion, choosing instead to claim that the method of testing doesn’t work. They have no alternative to propose, and simply expect such therapies to be accepted as effective on the basis of clinical experience alone. This is clearly a self-serving approach and not an example of a fatal flaw in evidence-based medicine.

This vet then goes on to argue that because scientific research studies try to minimize variables and simplify circumstances to make results easier to evaluate, that science ignores complexity and can only be useful in evaluating and treating very simple problems with single causes. Again, there is some truth to the notion that one will almost never find a real-world situation as clear and simple as even the most complex research study, so there are things such studies can’t tell us about phenomena in the real world. But the issue isn’t whether scientific evidence is perfect, it is whether it is more reliable than the alternatives.

The alternative proposed to scientific research is simply trial-and-error experience with individual patients. The CAM alternative to clinical research is simply to try and identify patterns in anecdotal experiences and our own clinical practice and rely on those. This is what human beings did for all of history until the scientific method was developed. In was a spectacular failure in comparison to what we have achieved using science.

Haphazard uncontrolled observation never doubled average life expectancy, dramatically reduced infant and maternal mortality, eliminated entire diseases, or accomplished anything like the amazing improvements in health and well-being in thousands of years that we have achieved in a mere couple of centuries. People like this author are not suggesting a new alternative or improvement in understanding disease diagnosis and treatment. They are proposing we return to the folk medicine methods that served us so poorly for so long.

Constructive criticism and improvement to scientific methods, including evidence-based medicine, is essential, and no one challenges concepts or practices in science more vigorously than scientists. But identifying, exaggerating, and fabricating weaknesses in EBM and then proposing we return to the even more limited and unreliable methods of history is not in the best interests of patients, human or veterinary.




This entry was posted in General. Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *