On alternative medicine, “labels,” and evidence

Alright, alright already! I get the message.

Over the course of the day yesterday I was bombarded by e-mails with a link to a New York Times article that shows a rather shocking lack of understanding of the science—more specifically, the lack of science—behind alternative medicine. Whenever something like this happens and I get so many requests to address a specific article, I’m always torn between my natural contrariness, which tempted me not to touch this article with the proverbial ten foot cattle prod (although something about this needs a cattle prod applied to it) and my desire to give the people what they want. In this case, the latter won because the the article is by someone whom I’ve actually cited on this blog before, someone who “gets it,” or so I thought. I’m referring to Aaron Carroll, a.k.a. The Incidental Economist, whose to whom I’ve referred in discussions of why cancer care is not worse in Europe than in the US and lead time bias.

It’s always a shock when someone who seemed to “get it” turns out not to. Actually, that’s not the way to put it. Clearly Carroll “gets it” when it comes to screening and cancer overtreatment. He also “gets it” about a lot of other things. However, when it comes to alternative medicine, he clearly doesn’t and demonstrates it in an article entitled Labels Like ‘Alternative Medicine’ Don’t Matter. The Science Does. (It’s reprinted here at Carroll’s blog, but unfortunately the comments are closed.) Now, reading the title, you might wonder why I’m even bothering. The title, after all, says nothing that I haven’t said myself many times, albeit not in quite that way. Indeed, one of my favorite sayings, cribbed from skeptics as varied as Richard Dawkins and Tim Minchin is that there is no such thing as alternative medicine. Alternative medicine that has been shown to work ceases to be “alternative” and becomes just “medicine.” Personally, I use a version that goes something like this: There is no such thing as alternative medicine. There is medicine that has been shown to work scientifically; there is medicine that hasn’t been shown to work; and there is medicine that has been shown not to work. So-called “alternative” medicine is made up almost exclusively of the latter two categories. Otherwise it wouldn’t be “alternative.”

You get the idea.

What apparently inspired Carroll’s article were the recent news stories about a course in alternative medicine being taught by an antivaccine homeopath, who thought that appropriate curricular material included writings by Andrew Wakefield and other luminaries of the antivaccine movement, you know, the one that I complained so bitterly about. This leads Carroll to write:

The dichotomy, however, between alternative and traditional medicine, or between Eastern and Western medicine, is a false one. We would be much better off if we could reframe the issue.

People often think of Eastern or alternative medicine as more “natural.” Many feel that Western medicine is built around technology and products produced in a lab. They’re not entirely wrong. Many of the gains that have been made in traditional medicine have been the result of innovation in laboratories.

But that doesn’t mean that everything doctors are taught in medical school involves a drug or device. I talk to patients all the time about diet and exercise. I don’t do this because there’s a company making money off it. I do it because both of these things have been proven to be important for health.

OK, so far, so good. There’s not much to disagree with here. Carroll just says what I’ve been saying in a different way, although I do take issue with his referring to “Eastern” or “alternative” medicine as somehow being more “natural.” That is, in actuality, nonsense. Is it any more “natural” to grind up animal parts, as, for example, traditional Chinese medicine does, than it is to isolate purified components from plants, as modern pharmacology not infrequently does? Is it more “natural” to rely on a vitalistic system divorced from science and reality in which “imbalances” between the five elements (very much like the four humors in “Western” medicine) than it is to rely on natural science, as science-based medicine does? Only if your definition of “natural” is a bit constrained and artificial in which the laboratory is somehow “unnatural or artificial” while everything else is “natural.”

That being said, Carroll is correct and makes a point that I’ve made many times about how diet and exercise are science-based medicine. Somehow, “alternative medicine,” then “complementary and alternative medicine,” and then “integrative medicine” have claimed these modalities as somehow “alternative” or outside of “mainstream medicine.” They do this because these are modalities that can have value for many conditions. Lumping them in with the hard core quackery, like homeopathy, “energy medicine” (healing touch and reiki, for example), and acupuncture gives credence to the quackery by association.

Speaking of acupuncture, this is where Carroll goes off the rails. If only he stopped with what he wrote above. But he doesn’t. First, he mentions that “not all medications get cooked up in a lab,” mentioning folic acid for pregnant women to reduce the risk of major birth defects, vitamin C for the prevention of scurvy, and vitamin D for the prevention of rickets. Of course, the wag in me can’t resist getting snarky and pointing out that most of these things actually are “cooked up in a lab,” at least when administered as a supplement. There’s nothing more “natural” about them than there is about Lipitor.

Unfortunately, acupuncture begins Carroll’s downfall in this article:

There are many other forms of nontechnological medicine that have the weight of scrutiny behind them. In a meta-analysis published just a few years ago, researchers looked at all the accumulated randomized controlled trials examining how acupuncture fared in treating people with chronic pain. They found that not only did acupuncture work better than no-acupuncture control groups, but there were also significant differences between acupuncture and sham acupuncture. This suggests that not all of the benefits are placebo effects.

No, no, no, no, no, no! How can someone who is usually so spot-on most of the time get something like this so spectacularly wrong. Before I explain why, let me ask you a question. (Longtime regular readers have a good chance of knowing the answer to this one.) What is the study to which Carroll’s link goes? Here’s a hint: It’s in the JAMA Archives of Internal Medicine. Give up? It’s the meta-analysis by Andrew Vickers, a meta-analysis that was widely touted in the press as slam-dunk evidence that acupuncture “works,” when in fact it showed nothing of the sort. I myself discussed why that is, particularly the part about how the purported effect size is well below the minimum clinically important detectable difference (MCID). In other words, Vickers’ meta-analysis, despite the play it got in the press, did not show that acupuncture works better than placebo for chronic pain. Quite the contrary as Steve Novella, Mark Crislip, and I argued. To go into detail is beyond the scope of this post; so I leave the links for interested readers to peruse, along with the one article where, in the context of critiquing “integrative oncology,” I also discussed how existing evidence does not support the efficacy of acupuncture for pretty much anything in the peer-reviewed medical literature in a high impact journal. The bottom line is that when you look at the totality of evidence for acupuncture, there is no convincing evidence that it is more than an elaborate placebo. How Carroll missed that, I don’t know. One wonders if he even read the paper.

On the other hand, a lot of doctors were fooled by Vickers’ meta-analysis. You really have to dive into the weeds of the exact methods to see its problems, and even then it’s not easy.

On some level, Carroll and I seem to be saying the same thing. Elsewhere in his article, he points out how many drugs used by “conventional” medicine derive from natural products. Digitalis, for instance, comes from foxglove; quinine from cinchona bark; penicillin from bread mold; and aspirin from willow tree bark. This is nothing more than pharmacognosy, the branch of pharmacology that studies natural products for potential pharmacologic activity. Indeed, in the case of aspirin, it’s not the substance isolated from the raw willow tree bark that is used, but rather a chemically modified version of it, acetylsalicylic, derived from the salicylic acid in the bark. He does this as a prelude to discussing all the herbs and plants used as “natural” alternative medicine.

There’s a big problem, though. He’s comparing apples and oranges. Carroll concedes that the reason that all the drugs mentioned above are so effective is because “conventional medicine may have improved our ability to purify these substances” (there’s no “may have” about it), but then plunges into a false equivalence between modern natural product-derived pharmaceuticals, purified, standardized, packaged, and reliable, compared to a variety of natural medications

Butterbur, a plant extract, has been found in medical studies to be as effective as antihistamines in treating allergic rhinitis, without the sedating side effects conventional drugs often have. Horse chestnut seed extract appears to be safe and effective in the short-term treatment of chronic venous insufficiency. Peppermint oil can be used to relieve the symptoms of irritable bowel syndrome. I know of few physicians who promote these therapies as often as they do prescription or over-the-counter drugs. Granted, that could be because it’s potentially hard to be sure of supplements you’re buying, but there are ways to overcome those problems.

None of these studies are particularly impressive. The butterbur study, for instance, was a non-inferiority study comparing butterbur to Zyrtec (cetirizine) that concluded that butterbur was equivalent to Zyrtec. However, it’s not as clear as that. For instance, a more recent randomized study, which, unlike the study referenced by Carroll, had a valid placebo control to compare to butterbur, found no difference between placebo and butterbur on allergic rhinitis symptoms. Another found a difference. Summing it all up, a systemic review from 2007 concluded that butterbur had some promise, but “independent replication is required before a firm conclusion can be drawn because of the financial support from the manufacturer of P hybridus extract to the 3 large trials.” Carroll also forgot to mention that butterbur can cause severe liver toxicity. Indeed, in 2012 in the UK, the Medicines and Healthcare products Regulatory Agency (MHRA) urged British herbal shops to remove products containing butterbur from their shelves. Oops! There’ goes the “non-toxic” claim. Repeat after me: If an herbal medicine has actual, measurable effects on a symptom or disease, it is a drug, and if it is a drug it can cause toxicity along with its therapeutic effects.

Of course, the reason that butterbur has an effect on allergic rhinitis is because it contains substances that block the production of leukotrienes, key molecules modulating inflammation. Given that, one wonders why an extract of the plant would be better than isolating these molecules. The same is true of horse chestnut extract, whose evidence supporting its use in chronic venous insufficiency is stronger. Ditto peppermint oil. Ironically enough, the very problem with supplements recognized by Carroll are the same problems each of these “natural remedies” have. The way to overcome that problem? Apparently Carroll thinks that state attorney generals going after sellers of adulterated or mislabeled supplements is a model to overcome the problems with using herbal medicines like the ones described above.

Carroll quite correctly recognizes that the evidence for homeopathy is resoundingly negative. He also correctly recognizes that evidence doesn’t matter to its proponents. Disappointingly, this leads Carroll to a painful-to-read false equivalence:

Those who favor conventional medicine, though, can be just as blinded. Too often, when confronted with evidence that advanced technology might not be providing benefits, the medical community refuses to change its behavior. My Upshot articles are littered with examples of this, including potentially too-widespread mammography screening, advanced life support and many surgical procedures. Supporters of Western medicine are often blind to their own prejudices.

As I’ve conceded more times than I can remember, yes, physicians can be frustratingly slow to change in the face of new evidence. Indeed, from time to time I’ve repeated a joke I first heard from faculty in medical school that outmoded treatments never truly go away until the physicians who learned them during their training either retire or die off. That’s a bit of an exaggeration, of course. As a counterpoint, I can point to my own specialty of breast cancer surgery and note that the way I approach breast cancer surgery now is strikingly different from the way I learned to approach it during my residency and fellowship in the 1990s. Practice has changed enormously since then based on science and clinical trials. I’m sure other physicians in other specialties can report the same thing. In other words, in medicine, change is the rule, not the exception. It is true that there are holdouts that resist evidence (such as vertebroplasty for osteoporotic compression fractures of the spine). Some even lash out at those who produce evidence that goes against their orthodoxy.

Here’s the difference, though. In science-based medicine, science and evidence eventually win out. Practice changes in response. The process might be messier than we like, meet more resistance than might be reasonable, and take longer than one would like, but in time practice does change. It really does. Contrast that to homeopaths and other alternative medicine practitioners, who are completely impervious to evidence. Yes, we practitioners of science-based medicine do all too often resist new evidence, but eventually we are forced to capitulate because science and reality compel us. That’s because our practice is rooted in science. Therefore, we speak and think in scientific terms. We justify our treatments with science. If science turns against a treatment we cannot resist indefinitely. Contrast this to alternative medicine practitioners, whose practices are rooted in magic, fantasy, and prescientific vitalism. Let’s just put it this way. When your entire practice is rooted in pseudoscience and unreality, science and reality have a hard time breaking through.

It’s a shame that Carroll doesn’t seem to understand that.