And now for something completely different.
Yes, it’s about time for that, isn’t it? I’ve probably beat the Tribeca Film Festival story into the ground, even for me, having spent the last week blogging about it. Scratch that. There’s no “probably” about it.
I frequently write at length about the quackery that is homeopathy. One reason I do this is because it is one of the most perfect forms fo quackery there is. There is about as close to no chance that it could work as there can be, and the only reason I don’t call homeopathy completely impossible is because I have a hard time calling anything completely impossible. However, it’s about as close to impossible as one can imagine, as, for homeopathy to work, multiple well-established laws of physics and chemistry would have to be not just wrong, but spectacularly wrong. Moreover, homeopathy is based on two principles that have zero basis in science. The first is the so-called law of similars, which states that to relieve a symptom you use a substance that causes the symptom in healthy volunteers. No basis in science and medicine. The second is the law of infinitesimals, which states that diluting a remedy (with vigorous shaking between serial dilutions, a process that homeopaths call “succussion”, to “potentize” it) makes it stronger. Of course, it’s more common than not that homeopaths dilute their remedies far more than Avogadro’s number, such that it is incredibly unlikely that there is even a single molecule of remedy left, barring carry over from serial dilutions. Basically, most homeopathic remedies are water or ethanol, depending on what was used to dilute them. Actually, many of them are not even that, because the final dilution is often pressed into sugar pills. Homeopathy is, as I said, the perfect quackery. It can’t work, and it doesn’t work.
Over the many years that I’ve been writing about homeopathy and other alternative medicine, one question that’s come to mind on many occasions is simple: How on earth can a physician come to believe that something that is this obviously quackery and this obviously doesn’t work actually does work? Naturopaths, I can understand. They are steeped in a world view that is part prescientific and part pseudoscientific. Homeopathy is part of their training—a major part. They have to know homeopathy to pass the NPLEX, the naturopathic licensing examination. Physicians, on the other hand, are trained in evidence- and science-based medicine, or at least they should be.
So it was with interest that I read Edzard Ernst’s post on The Making of a Homeopath. Part autobiographical, given Ernst’s experience before he became a skeptic of alternative medicine, part fictional, the story helps to explain why someone trained in medicine could fall for the pure pseudoscience that is homeopathy.
The story begins:
After he had finished medical school, our young and enthusiastic doctor wanted nothing more than to help and assist needy patients. A chain of coincidences made him take a post in a homeopathic hospital where he worked as a junior clinician alongside 10 experienced homeopaths. What he saw impressed him: despite of what he had learnt at med school, homeopathy seemed to work quite well: patients with all sorts of symptoms improved. This was not his or anybody else’s imagination, it was an undeniable fact.
This is why placebo controls are so important for any treatment designed to relieve symptoms. Basically, any treatment or intervention, effective or not, will produce an apparent improvement in symptoms due to placebo effects. That’s why personal clinical experience can be so misleading, something many physicians have a hard time accepting. Ernst then explains how this can happen, using the example of an otherwise healthy young woman suffering from infertility:
Our young physician was not convinced that he could help his patient but, in the end, he was persuaded to give it a try. As he had been taught by his fellow homeopaths, he conducted a full homeopathic history to find the optimal remedy for his patient, gave her an individualised prescription and explained that any effect might take a while. The patient was delighted that someone had given her so much time, felt well-cared for by her homeopaths, and seemed full of optimism.
Months passed and she returned for several further consultations. But sadly she failed to become pregnant. About a year later, when everyone involved had all but given up hope, her periods stopped and the test confirmed: she was expecting!
Everyone was surprised, not least our doctor. This outcome, he reasoned, could not possibly be due to placebo, or the good therapeutic relationship he had been able to establish with his patient. Perhaps it was just a coincidence?
There’s no “perhaps” about it. It was very likely coincidence, particularly given that it was a year later that the woman finally became pregnant. However, news of this anecdote provided by this seeming success of homeopathy spread, leading the young doctor to see several other young women with fertility problems, ten to be precise. The results were as follows:
The doctor thus treated several infertile women, about 10, during the next months. Amazingly most of them got pregnant within a year or so. The doctor was baffled, such a series of pregnancies could not be a coincidence, he reasoned.
Naturally, the cases that were talked about were the women who had become pregnant. And naturally, these were the patients our doctor liked to remember. Slowly he became convinced that he was indeed able to treat infertility homeopathically – so much so that he published a case series in a homeopathic journal about his successes.
In a way, he had hoped that, perhaps, someone would challenge him and explain where he had gone wrong. But the article was greeted nationally with much applause by his fellow homeopaths, and he was even invited to speak at several conferences. In short, within a few years, he made himself a name for his ability to help infertile women.
One more time: Personal clinical experience can be very misleading. It’s not just how anecdotes can make it seem as though ineffective treatments “work.” It’s also how good such anecdotes make physicians feel. Think about it. Why do people go into medicine? One reason that it is profoundly rewarding to help patients. Being able to help a fellow human being overcome an illness, to relieve suffering, to cure a disease, all of these things make a doctor feel very, very good. Helping our fellow human beings is a powerful reward. I’ve felt it myself after a particularly difficult operation to remove a breast cancer. Saving lives and removing suffering, it’s what we as physicians are about. Homeopathy can make it seem as though that’s what we are doing is the same. When there are no controls, we can easily confuse regression to the mean or the normal clinical course of, say, infertility with an actual therapeutic effect.
Ernst schools us on why:
But there are other, even simpler and much more plausible explanations for our doctor’s apparent success rate: otherwise healthy women who don’t get pregnant within months of trying do very often succeed eventually, even without any treatment whatsoever. Our doctor struck lucky when this happened a few times after the first patient had consulted him. Had he prescribed non-homeopathic placebos, his success rate would have been exactly the same.
As a clinician, it is all too easy and extremely tempting not to adequately rationalise such ‘success’. If the ‘success’ then happens repeatedly, one can be in danger of becoming deluded, and then one almost automatically ‘forgets’ one’s failures. Over time, this confirmation bias will create an entirely false impression and often even a deeply felt conviction.
I am sure that this sort of thing happens often, very often. And it happens not just to homeopaths. It happens to all types of quacks. And, I am afraid, it also happens to many conventional doctors.
This is how ineffective treatments survive for often very long periods. This is how blood-letting survived for centuries. This is how millions of patients get harmed following the advice of their trusted physicians to employ a useless or even dangerous therapy.
I would go even further than this. It happens to all doctors at one time or another. In particular, it happens to surgeons. We tend to remember the good outcomes and forget the bad. It’s not that we do this intentionally. It’s just that confirmation bias is part of human nature. It’s how our brains are wired, every bit as much as how we tend to confuse correlation with causation. While it’s true that correlation can often indicate causation, more often it does not. In fact, if there’s one thing that Ernst doesn’t emphasize enough, it’s just how much this happens not just to quacks, but to doctors who like to think of themselves as science- and evidence-based.
It’s a two-way street. Both the doctor and the patient get something out of alternative medicine, particularly in the case of chronic disease or chronic symptoms for which conventional medicine doesn’t have a treatment that does what the patient wants and completely eliminates his symptoms. The patient derives hope and seeming temporary relief due to placebo effects. In addition, because so many alternative medicine modalities practically require “personalization” or, more specifically, for the practitioner providing treatment to pay more attention to the patient, the patient also gets a human ear to listen to him.
But what does the doctor get out of this? First of all, he gets something he can offer patients other than the unsatisfying options that conventional medicine provides him. For patients with, for instance, chronic pain this is a fantastic thing, as patients will tell him they feel better even in the case of using homeopathic remedies due to placebo effects and regression to the mean. Also, as Ernst points out, even in the case of “hard” endpoints like a successful pregnancy after treatment for infertility, there are a lot of conditions that improve on their own a lot more frequently than people tend to believe. Another perceived reward doctors receive for using alternative medicine comes from how medicine is practiced today. Today, everything is about productivity, about RVUs. In primary care, doctors seldom have the time to be complete doctors of the old-fashioned variety who can spend time with their patients because it’s not financially viable to do so. Alternative medicine allows a way out, and, because it’s not reimbursed by insurance, physicians can charge whatever the market will bear to prescribe it.
Ernst is quite correct to conclude, as he does, that “health care professionals need to systematically learn critical thinking early on in their education.” However, in the US at least, medicine itself needs to change the features in it that reinforce the use of alternative medicine, such as the hamster wheel schedule that so many primary care doctors must adhere to in order to support themselves.