The Truth About Evidence-Based Medicine

When I was working on the “healing systems” section of this website, I stumbled across Wikipedia’s article on naturopathy, which described it as using “an array of pseudoscientific practices,” and being “based on vitalism and folk medicine, rather than evidence-based medicine.”  (And conveniently is “semi-locked” for editing, so only experienced users can edit it and it can’t be labeled as biased.)  There are some problems with that claim, and I want to talk about why — but I’m going to do so within the context of a larger discussion of evidence and health practice.

Evidence Level Varies

The main issue with the statements on Wikipedia is that they treat naturopathy (and, by implication, conventional allopathic medicine) as homogenous entities.  No modality has a monopoly on evidence-based practice, and no modality is universally evidence-based.

By way of a brief example, consider my college texts.  There is one particular book we referenced at times, which was written by a couple of naturopathic doctors and, in my opinion, pretty well characterizes the naturopathic approach.  It’s packed full of references to studies, etc.  I would consider it to be heavily evidence-based.  My more mainstream texts (like pathophysiology & psychology), on the other hand, treated a lot of things as information that “everybody knows,” and did not bother to support the claims with evidence.

This is disappointingly common in allopathic medicine — there’s a sort of “proof by consensus” at play.  A majority of the field accepts an idea or adopts a practice, so it becomes something that “everybody knows” and considered to be above the need for proof.  The problem is, in many cases there never was any proof (and in some other cases, the evidence is so old it really ought to be re-examined).

A recent example involves the hubbub about saturated fat.  The data that “proved” saturated fat clogs arteries was twisted and/or fabricated from the start.  The short version is that investigational studies only reported on the evidence that fit the hypothesis (see The Cholesterol Myths for a very enlightening pair of graphs) and that much of what Keys used in his own experience was implied to be butter, but was actually hydrogenated oil — a radically different substance. 1

Unfortunately, this thinking quickly took over and persisted for decades, during which the highly-processed oils with which America replaced butter actually did cause health issues.  The USDA still includes the obsolete recommendation to avoid saturated fats, even on the very same page with their own data that show how far off-base this is.

Perhaps a less controversial example is the routine use of episiotomy during childbirth.  When episiotomy was introduced, the idea was that by cutting into the vaginal opening to enlarge the orifice, tearing during childbirth would be reduced.  This practice spread throughout the obstetric field until it was common routine practice.  When the practice was actually submitted to scientific scrutiny, it was determined that it does not accomplish its intended goals.  In fact, it increases severe tears and makes postpartum recovery more difficult. 2.  (It does have a place in select instances, but not as a routine procedure.)

Lest you should think I’m trying to demonstrate that alternative medicine is evidence-based and mainstream medicine is not evidence-based, let me assure you that is not the case!  I’m simply pointing out that the reverse is not true.  There is a tendency to assume anything in mainstream medicine is evidence-based, but a surprising amount of medical practice is not particularly evidence-based at all.  Conversely, there is a tendency to assume anything alternative is “quackery,” but you might be surprised by how much of it has solid scientific backing.

All of it calls for due diligence.

Types of Evidence

When I talk about “evidence-based” here, I’m thinking primarily of structured, intentional studies, but that isn’t the only form of evidence that exists.  Even anecdotal reports are a form of evidence, albeit a considerably less reliable one.  (The scientific community recognizes this; that’s why case reports are published in journals.)

The randomly-controlled, double-blinded trial is the “gold standard” for good reason.  It enables researchers to carefully isolate a single variable and reduce or eliminate confounding factors.  But it isn’t well-suited for every intervention that calls for study, and it shouldn’t be allowed to wrongly push all other evidence into the category of “non-evidence.”  (It also is subject to the quality of an individual study’s design.)

All evidence should be carefully considered for what it can — and cannot — tell us.  It should also be given an appropriate degree of weight based on the potential consequences of being wrong.

How Strong Does the Evidence Need to Be?

This last point is a critical one.  Not everything is studied by means of a randomly-controlled trial.  Some things simply cannot be studied in this manner.  Others just aren’t, for a variety of reasons.  For one thing, lab studies and clinical trials cost money, and they’re least likely to be funded where there’s least potential for financial gain.  (Those investing want to recover their investment, after all.)  Should we always avoid the use of an intervention that isn’t well-studied?  Not necessarily.

The wise person considers the likelihood of risk and the potential consequences of an intervention’s being ineffective, and requires an appropriately-corresponding degree of evidence.  Let’s consider a couple of examples by way of illustration.

First, evaluating risk.  Let’s take the herb chamomile, and the pertussis vaccine.  Chamomile has been used for centuries.  We have numerous reports of its use, we have folk histories of what it’s useful for, and (the key factor here) we have an absence of any notable mention of toxicity.  It seems unlikely that the herb could have been used for so long without any notes of toxicity if it’s dangerous, so we can likely consider it fairly safe when used in reasonable amounts and with common sense application.

Vaccines as we know them are only around 100 years old, and the pertussis vaccine in use today is only decades old.  It does not have a centuries-long history of use to prove its safety, and it has been the subject of controversy throughout its entire existence.  Thus, we probably want some scientific trials to prove its safety before assuming it to be safe.

In addition to the risk, though, we’re also concerned with the efficacy of a given intervention.  How important this is will depend on what we’re trying to accomplish.  If you have a minor illness like a cold, and you’re hoping for something to reduce your nasal congestion, an unproven herb might be an acceptable risk to take.  The cold isn’t life-threatening, so if the herb doesn’t work, you simply still have a stuffy nose.  Annoying, but no big deal.

Even in a more chronic situation, like chronic pain, or even a life-threatening one, like cancer, an unproven remedy might be worth the risk if none of the more tested options have worked.  If every pain medicine has been tried, and your pain is still present, does it hurt anything to try an energy medicine treatment because it “might not work”?  Well, if it doesn’t work, you’re no worse off than you already are, so no (at least as far as efficacy/inefficacy is concerned).  If you have cancer, an untested option is probably not the best primary approach.  But what if doctors have exhausted all their options and told you there’s nothing more they can do?  In this case, you might as well try the untested option.  (Likewise, if it won’t interfere with a conventional treatment, it can be a “safe risk” to use both.)

But what if your child has a systemic infection that’s threatening to shut down her organ function?  This is a serious, life-threatening situation, and one which requires fast action.  This would be a foolish time to try something unproven.  In this case, the antibiotic that has a demonstrated ability to wipe out the infection in question is a wise choice.  (This is, of course, assuming access to standard medical care.  If you’re somehow stuck in the bush with access to nothing but some plants, try the most likely option you do have access to!)

The bottom line is, there is not a one-size-fits-all solution.  All (well, most) modalities have strong points and weak points in their evidence bases, and every situation should be weighed on its own merits.


One thought on “The Truth About Evidence-Based Medicine

Comments are closed.