Stick It In Your Ear! How Not To Do Science
Have a sore throat? No worries! No need for lozenges, medicines, or home remedies. All you need to do is let someone stick needles in your ear! According to a recent study, ear acupuncture relieves sore throats. Do you believe that? I don’t. That’s one of those extraordinary claims that would require extraordinary evidence, but the researchers didn’t even provide ordinary evidence. The study is a great example of how not to do science.
Acupuncture theory is based on a prescientific vitalistic concept: an unmeasurable, undetectable energy called “qi” supposedly flows through meridians and can be accessed at acupoints, where needle stimulation is supposed to somehow unblock the flow of qi, which is somehow supposed to relieve pain and improve health. Qi, meridians, and acupoints are imaginary, but that doesn’t necessarily mean acupuncture can’t possibly work. It’s not implausible that sticking needles into the skin might have some physiologic effects, so it is reasonable to do scientific studies—and thousands of studies have been done, some better than others. The results have been disappointing. It is not reasonable to conclude from the existing published studies that acupuncture works.
Some people believe acupuncture is an effective way to relieve pain, but the evidence from scientific studies and systematic reviews is mixed. Any positive effects can be attributed to suggestion and the surrounding rituals, and the effects are too small in magnitude to have any clinical importance. The most rigorous studies have shown that sham acupuncture works just as well as “real” acupuncture. It doesn’t matter where you put the needles, and in fact it doesn’t matter whether you use needles at all. In one study, simply touching the skin surface with a toothpick worked just as well as penetrating the skin with a needle (Cherkin et al. 2009). There was even a study where subjects were given a “phantom limb” illusion that a rubber hand was their own hand, and they got similar results from acupuncturing the rubber hand! (Chae et al. 2015). The main thing that seems to matter is whether the patients believe acupuncture will work. If they get a sham procedure but think they got the real thing, they will think it worked. If they get the real thing but think they got a sham procedure, they will think it didn’t work. It has been pretty well established that acupuncture is nothing more than a theatrical placebo (Colquhoun and Novella 2013).
The Center for Inquiry issued a position paper on acupuncture that concludes:
It is becoming increasingly clear that the placebo effect accounts for much of the optimistic research on acupuncture published between the 1970s and 1990s. With the development of sham needles during the past decade—needles which retract like stage knives—researchers have since been able to more accurately assess the therapeutic effects of acupuncture. The result has been a complete unraveling of nearly all acupuncture claims... . The bulk of recent research strongly tends towards the hypothesis that acupuncture’s positive effects are mainly due to a built-in expectation... . (Slack 2010)
When The Medical Letter reviewed acupuncture in 2006, it concluded “Acupuncture alone has not been shown in rigorous, duplicated studies to benefit any defined medical condition.” Our own CSI fellow, Steven Novella, coauthored the paper that concluded acupuncture was nothing more than a theatrical placebo (Colquhoun and Novella 2013). Medical ethicists unanimously agree that doctors should not prescribe placebos. So why are some doctors still recommending acupuncture?
Unfortunately, there are many doctors who lack critical thinking skills and don’t really understand science or the psychology of cognitive errors. They don’t all read Skeptical Inquirer or the Science-Based Medicine blog. They should.
The history of acupuncture research is revealing. There were never any controlled studies to verify the alleged location of meridians and acupoints or their alleged connections to other parts of the anatomy. They transposed human diagrams onto animals without any testing for validity; the acupuncture diagram for horses, for example, shows gallbladder meridians. Horses don’t have gallbladders.
There are many different systems of acupuncture, some involving needles and some involving various kinds of stimulation of acupoints through intact skin (with electricity, light, heat, and other modalities). The number of meridians and acupoints varies among the different systems. In traditional acupuncture, there were originally 365 acupoints to symbolically correspond to the days of the year; now there are over 2,000. In Korean acupuncture, there are 300 acupoints, all on the hand. In other systems, they are found only on the ear, scalp, tongue, or foot. With all those different acupoints in all those different systems, is there any spot on the body that isn’t an acupoint in some system? Yes, there is: no system uses points on the male genitalia. Draw your own conclusions! A variant called “tong ren” doesn’t use acupoints at all; the patient rhythmically taps an acupuncture doll with a metal hammer. The online video of tong ren (https://www.youtube.com/watch?v=_I4r3FrzFBc) has to be seen to be believed: Is there no limit to human gullibility? As far as I know, no one has ever thought to try testing one system of acupuncture against another to find out which one is superior. For that matter, no one has ever thought to test homeopathy, chiropractic, acupuncture, and Reiki against each other. Alternative medicine simply doesn’t think that way. Anything goes.
The sore throat study didn’t use traditional acupuncture. It used ear acupuncture—and not even standard ear acupuncture but a simplified version. Why? Because the researchers were Air Force doctors, and the Air Force has been teaching ear acupuncture as “battlefield acupuncture” to its doctors under the guidance of Col. Richard Niemtzow. They clearly have drunk the Kool-Aid, because they allege that points on the ear correspond to areas on the body and stimulating them “balances the flow of the body’s energy or qi.” Their modified procedure involves five short needles that are left in place to eventually fall out on their own; the method is taught to non-acupuncturists in a four-hour workshop. The authors claim that there is research supporting ear acupuncture, but they admit that there is only “limited data” regarding the modified technique. You’d think that before they forged ahead to use it on wounded soldiers on the battlefield, they’d want convincing evidence that it worked. There isn’t any.
Ear acupuncture isn’t as silly as tong ren, but it’s still pretty silly. It was invented by a Frenchman, Dr. Paul Nogier, in 1957. He invented it all by himself—not through science but through intuition and epiphany. He looked at the external ear, the pinna, and imagined that it looked sort of like a fetus curled up in its mother’s womb. I can tell you he had a much better imagination than I do; I can’t see the resemblance. He arbitrarily assigned a spot on the ear to correspond to the part of the imagined fetus, tried inserting needles in ears, and confirmation bias kicked in to convince him that it worked on the corresponding part of the body. He originally used thirty points; now ear acupuncturists claim to have identified 120. The ear is pretty small to accommodate 120 acupoints, and anyway there are no connections between the ear and those other parts of the body.
The study, “Ear Acupuncture for Acute Sore Throat: A Randomized Controlled Trial,” (Moss and Crawford 2015) was published in the Journal of the American Board of Family Medicine. The fact that the editors did not reject the study is both surprising and disappointing.
Moss and Crawford studied fifty-four patients who presented to an Air Force family medicine clinic with pain from sore throat. They were randomized into two groups: one got standard treatment (ibuprofen, plus antibiotics when indicated); the other got standard treatment plus ear acupuncture. There was no blinding and no sham acupuncture control group. They didn’t even standardize the treatment: they placed “up to 10” needles. Unblinded research associates assessed the results by phone interview. Patients in the acupuncture group reported significantly less pain at fifteen minutes and twenty-four hours after the procedure, although at forty-eight hours there was no difference from the control group. Acupuncture patients reported taking fewer ibuprofen pills. There was no difference between the groups in time missed from work.
What’s wrong with this study? Practically everything. It is a typical example of the “pragmatic trial” favored by alternative medicine. Instead of asking whether acupuncture works better than placebo to relieve pain, they gave everyone standard treatment and added acupuncture to one group. When you add anything to the usual treatment, you are practically guaranteed to get positive results. Patients respond to the extra attention and to suggestion. They expect to feel better, so they do. In this case, the needles were left in the ear as a continuing reminder that they had received special treatment. Pain is subjective, and suggestion is powerful; that’s why we kiss our children’s minor injuries. Of course the acupuncture subjects reported less pain; that’s exactly what we could have predicted.
Pragmatic studies test the real-world practical performance of treatments. They have their place. A treatment that has been proven to work in a research setting with a select group of subjects and carefully controlled methods might not work as well in a real-world setting with patients who might have other illnesses or be on other medications, or where the providers might not be as rigorous in following treatment guidelines. Pragmatic studies can’t establish whether a treatment works better than placebo, and they were never intended to be done on treatments that had not already been proven to work.
That’s why studies must have a convincing control group. These researchers could have used a sham acupuncture control where they used retractable needles or put the needles in the “wrong” spots, in the ear or in other parts of the body. But they rejected the whole idea of a sham acupuncture control. Their amazing rationale: sham acupuncture can stimulate “C” fibers and produce analgesia too, so sham acupuncture wouldn’t be a placebo, it would actually be a treatment! If that were true, it would mean that acupuncture’s whole treatment rigmarole is unnecessary. There would be no reason to specify acupoints: you might as well just stick needles anywhere, and no training would be needed to administer the treatment. The authors didn’t even think of comparing a usual treatment group to an acupuncture group; both groups got the usual treatment. And why on Earth didn’t they blind the research assistants who did the telephone interviews? That would have been so simple to do, and it would have eliminated any suspicion that they might have subtly influenced the results in some way.
They got positive results, but they would have gotten positive results if they had added any other intervention to the usual treatment for sore throat. What if they’d offered massages, or training in meditation or relaxation exercises, or a warm scarf to wrap around the neck, or therapeutic touch to “balance their human energy fields,” or funny movies to take their mind off the pain? Or even a gift certificate for a shopping spree at the local mall to boost their spirits? We can predict that those would have “worked” too, maybe equally well or better. That’s why pragmatic studies such as this are useless.
It’s an ill wind that blows nobody any good. This study is astonishingly poor science. It should not have been done, and it should not have been published; it didn’t give us any useful information about acupuncture. But it can serve as a bad example to help us learn to distinguish between good and bad science.
- Chae, Y., I.S. Lee, W.M. Jung, et al. 2015. Psychophysical and neurophysiological responses to acupuncture stimulation to incorporated rubber hand. Neuroscience Letters 691 (March 30): 48–52.
- Cherkin, Daniel C., Karen J. Sherman, Andrew L. Avins, et al. 2009. A randomized trial comparing acupuncture, simulated acupuncture, and usual care for chronic low back pain. JAMA Internal Medicine 169(9): 858–66. Available online at http://archinte.jamanetwork.com/article.aspx?articleid=414934.
- Colquhoun, David, and Steven P. Novella. 2013. Acupuncture is theatrical placebo. Anesthesia & Analgesia 116(6): 1360–63. Available online at http://journals.lww.com/anesthesia- analgesia/pages/articleviewer.aspx?year=2013 &issue=06000&article=00025&type=Fulltext.
- Moss, David A., and Paul Crawford. 2015. Ear acupuncture for acute sore throat: A randomized controlled trial. Journal of the American Board of Family Medicine 28: 697–705. Available online at http://www.jabfm.org/content/28/6/697.full. Slack, Robert. 2010. Acupuncture: A science- based assessment. Center for Inquiry, Inc. (June). Available online at http://www.centerforinquiry.net/uploads/attachments/Acupuncture_Final_Paper.pdf.