Articles

Acupuncture Pseudoscience in the New England Journal of Medicine

Here is the conclusion quoted from a recent New England Journal of Medicine (NEJM) review article on acupuncture for back pain:

As noted above, the most recent wellpowered clinical trials of acupuncture for chronic low back pain showed that sham acupuncture was as effective as real acupuncture. The simplest explanation of such findings is that the specific therapeutic effects of acupuncture, if present, are small, whereas its clinically relevant benefits are mostly attributable to contextual and psychosocial factors, such as patients’ beliefs and expectations, attention from the acupuncturist, and highly focused, spatially directed attention on the part of the patient.

Translation – acupuncture does not work. Why, then, are the same authors in the same paper recommending that acupuncture be used for chronic low back pain? This is the insanity of the bizarro world of CAM (complementary and alternative medicine). Yesterday David covered the same article, which I had also covered on NeuroLogica, but we both thought this issue important enough to document our thoughts and objections on SBM.

Let’s break down their conclusions a bit. They have reviewed the clinical evidence, as I and others have done before, and found that when real acupuncture is compared to various forms of sham acupuncture (the acupuncture version of a placebo) there is no difference. As I have written many times before – it doesn’t matter where you stick the needles, or even if you stick the needles. Reviews have also concluded that there is no evidence for the mere existence of acupuncture points. Since acupuncture consists of sticking needles in acupuncture points, the only reasonable conclusion from this evidence is that there is no specific effect from acupuncture – acupuncture does not work.

The phrase, “contextual and psychosocial factors, such as patients’ beliefs and expectations, attention from the acupuncturist, and highly focused, spatially directed attention on the part of the patient.” is a fancy way of saying “placebo effects.” In other words, there are some non-specific subjective benefits to getting attention from a practitioner. There is this assumption, however, that these benefits are real and worthwhile. However, they are likely to be illusory – an artifact of observation and reporting, not a real improvement in the patient’s condition. In real science-based medicine, that is the underlying assumption – placebo effects are largely illusory – a variable to be controlled for.

But there has been recent controversy over the role of the placebo effect in ethical and evidence-based practice. This is, in my opinion, largely a back door attempt to justify CAM treatments that do not work. The claim is that placebo effects are real and useful. But a systematic review of the placebo effect in clinical trials concluded:

We did not find that placebo interventions have important clinical effects in general. However, in certain settings placebo interventions can influence patient-reported outcomes, especially pain and nausea, though it is difficult to distinguish patient-reported effects of placebo from biased reporting. The effect on pain varied, even among trials with low risk of bias, from negligible to clinically important. Variations in the effect of placebo were partly explained by variations in how trials were conducted and how patients were informed.

In other words – for any objective outcome, there is no important placebo effect. For outcomes that are subjectively reported by patients, there is a highly variable placebo effect. It is plausible that the expectation of benefit could result in the release of dopamine and endorphins and produce a physiological decrease in pain, for example, in a subset of people, and there is some evidence for this. But this is, at best, a transient symptomatic effect – not therapeutic.

Such effects are also non-specific – meaning they do not derive from the intervention itself, but from the therapeutic ritual surrounding the intervention. Even treatments that do not work may therefore provide these non-specific benefits. My opinion is that the non-specific benefits of the ritual of treatment should be combined with an actually effective treatment, not magic pretending to be medicine. There are many reasons for this. One is the ethics of patient autonomy and informed consent – giving a fake treatment to a patient violates the patient’s rights, in my opinion.

Further, there is potential downstream harm from convincing patients that fake magical treatments are effective, because of placebo effects. Then using obscure language to hide the fact that the treatment actually does not work. This distorts the public’s view of medicine, and of what works, and sets them up to be victims of fake treatments when their ailment is not subjective or self-limiting. In other words – refer them to an acupuncturists when they have back pain and they may rely upon acupuncture, or some other non-scientific intervention, when they have a more serious illness.

There is further harm caused by diverting research time, money, and other resources from more fruitful lines of investigation in order to pursue a theory that has no basis in biology. There are thousands of published studies on acupuncture – given the negative results of this research most of this has been a waste of time and resources.

The authors of this article recommend:

He has specifically requested a referral for acupuncture, and we would suggest a course of 10 to 12 treatments over a period of 8 weeks from a licensed acupuncturist or a physician trained in medical acupuncture.

This contradicts their own conclusions. Why is training in acupuncture necessary? That training largely consists of identifying acupuncture points, knowing which points to use on an individual patient, and knowing the technique of needle insertion – but none of these things matter. The sham ritual is all that matters – you can literally fake it and get the same response. I bet a 10 minute video is all that is necessary. In fact I bet even that is not necessary – you could probably fake it well enough to get a maximum placebo effect without any prior demonstration.

What the authors of this article have done is something that is increasingly common in CAM (when it is trying to infiltrate academia and peer-reviewed journals like the NEJM) – reviewing the evidence, admitting that the CAM treatment does not work, then making an elaborate and misleading appeal to placebo effects, and ending with a recommendation to use the treatment that does not work. Specifically, they not only recommend using the treatment, but in its fullest magical form, complete with all the disproven claims (that is what “medical acupuncture” is). It’s a bait and switch con game, nothing more. Come for the placebo effect, then be treated with magical nonsense.

Posted in: Acupuncture, Science and Medicine

Leave a Comment (29) ↓

29 thoughts on “Acupuncture Pseudoscience in the New England Journal of Medicine

  1. rork says:

    Nice. I very much liked the longer treatment of the dark side of utilizing placebos, since it seems common to hear apologists. I’m still a bit openminded about administering what may be de facto placebos for certain mental states (anxiety, depression), which may be what’s happening with some drugs, and would like to obtain more opinions than what I heard on NPR a while back from Arthur Barsky (Professor of Psychiatry, Harvard Medical School):
    http://www.npr.org/templates/story/story.php?storyId=126802142
    It made me uneasy.

    On training: I wouldn’t be shocked if shamans of any stripe, trained more thoroughly, might get better subjective responses. The show may be better, and the mystical language employed may be more fluent and convincing. The healer’s belief might help too, since most of us are crappy actors.
    Doh, forgot my crystal – bad carpal tunnel day for sure.

  2. windriven says:

    I would like to make a modest suggestion: Rubber Goose awards for quacks of distinction. There is nothing like a little good-natured mockery to foster introspection. Along the lines of the Darwin awards, a Rubber Goose would include a recitation of the malpractice that earned the award. One supposes there could be special Rubber Geese for lifetime achievements in quackery. I wonder how someone like Berman would react to enshrinement next to Christian Hahnemann? I wonder if his colleagues would look at him in quite the same way? I wonder if those same colleagues might shy away from publishing woo?

  3. I was a massage therapist in a past life, and I routinely saw significant harm done by acupuncture and other ineffective therapies. Far from enjoying a robust mind-over-matter placebo effect, most patients seemed to believe all the more in their back pain as an unassailable affliction that “even acupuncture” couldn’t help. More tragic than simply wasting time and money on a treatment that doesn’t work is that so many patients conclude not that the treatment was ineffective but that acupuncture was defeated by an unusually serious case of back pain. Patients are strongly predisposed to anxious assumptions that their problem is “really bad,” and the failure of acupuncture confirms it. The acupuncturist is given the benefit of the doubt, while the back pain is seen as a fiercer enemy. A nice trap.

  4. qetzal says:

    It is plausible that the expectation of benefit could result in the release of dopamine and endorphins and produce a physiological decrease in pain, for example, in a subset of people, and there is some evidence for this. But this is, at best, a transient symptomatic effect – not therapeutic.

    I was recent referred to an interesting review on this topic. Based on the studies summarized therein, it does seem that patient expectations can stimulate release of endogenous opioids for pain, and even dopamine in Parkinson’s patients.

    Of course, the review authors are clearly convinced that placebos have ‘real effects. (They cite many of their own studies.) It’s possible they’ve played up the ‘positive’ studies, and played down any negative ones.

    Even accepting the review as accurate, I didn’t see anything to contradict Dr. Novella’s statement that these are transient, essentially non-therapeutic effects.

  5. urodovic says:

    I e-mailed the NEJM editors and got a quick reply. Due to confidentiality issues I can not post their exact response.

    Briefly, they state that they are aware of the controversial aspect of the published article but that the ultimate judge once “the evidence” is presented, will be the reader.

    Which reminds me of the Discovery Institute’s campaign in regards to evolution and creationism (ID)…”teach the controversy”

    http://en.wikipedia.org/wiki/Teach_the_Controversy

    I am very disappointed :-(

  6. @urodovic, that is disappointing, since your gripe was not with publication of this evidence, but with the publication of recommendations that are strikingly at odds with the evidence. This is “letting the reader judge the evidence”? Seems like it’s telling readers exactly what to do with the evidence.

    Evidence-denied medicine!

  7. Steve says:

    I would like to point out that the likely reason for indicating a qualified acupuncturist is to avoid the possibility of more harm. That statement is a non-sequitur for benefit, but makes sense if you want to avoid liability of encouraging “garage” acupuncture. I do not think it invalidates Steve’s point, just illuminates that this is not without risk.

  8. Steve – acupuncturists routinely insert needles without proper antiseptic technique. Otherwise the needles are small enough that you have to try (or be very unlucky) to cause direct harm. Your average physician would probably do a better job in terms of safety just following standard procedure without any specific acupuncture training.

  9. There is a lung-puncturing risk as well. I have heard of at least two cases of holy-crap-that-was-so-not-worth-it pneumothorax here in Vancouver, both caused by physical therapists using acupuncture needles as tools for a non-acupuncture dry-needling treatment for muscle pain (“intramuscular stimulation”).

  10. urodovic says:

    @Paul Ingraham

    Exactly my point, “the evidence” in the article amounts to “sham acupuncture was as effective as real acupuncture”

    So why allow the publication of an article in such high impact journal like the NEJM by an author than in a very subjective final recommendation states: “we would suggest a course of 10 to 12 treatments over a period of 8 weeks from a licensed acupuncturist or a physician trained in medical acupuncture

    Anyone knows how much a licensed acupuncturist will charge for one session?…My oh my…12 “treatments”….that surely looks like a very expensive placebo therapy..

  11. pmoran says:

    “The simplest explanation of such findings is that the specific therapeutic effects of acupuncture, if present, are small, whereas its clinically relevant benefits are mostly attributable to contextual and psychosocial factors, such as patients’ beliefs and expectations, attention from the acupuncturist, and highly focused, spatially directed attention on the part of the patient.”

    I haven’t access to the whole paper, but see no real quarrel with what is stated here, apart from the reluctance to call a spade a spade and allow that acupuncture is mostly placebo. It does not read as an endorsement of ancient Chinese medical superstitions. Note the cautious “if present” .

    The thing is, I suppose, that the word “placebo” has acquired strong negative connotations in too many minds.

    Berman might point out that acupuncture not working better than a sham treatments is not the same thing as saying that “acupuncture doesn’t work”, if BOTH have irrefutable clinical benefits (and who can say they definitely don’t?).

    “Acupuncture” has become a loose descriptor for a variety of interventions that involve taking the patient out of their normal environment for a spell once or twice a week, keeping them still while a mildly invasive and often painful treatment ritual is performed and certain social and medical interactions may occur. There may be counterirritant and distractant effects, perhaps an opportunity for resetting of pain tolerance levels, even perhaps some switching out of illness mode altogether in what can sometimes be a complex psychosomatic illness.

    Or, you continue with usual care, and the patient is much the same six months later!

    I take all the points that are being made. They all have SOME validity, without necessarily carrying the day. Just be careful not to completely overlook patient needs in what is normally a very difficult clinical problem.

  12. Dr Benway says:

    NEJM, your pants are on fire.

  13. weing says:

    “Anyone knows how much a licensed acupuncturist will charge for one session?…My oh my…12 “treatments”….that surely looks like a very expensive placebo therapy..”

    I think the more expensive the therapy, the more powerful the placebo effect. The $5 aspirin is more effective for a headache than the 50 cent one. The brand name drug works better than the identical cheaper generic. My take on this is that we canuse sham acupuncture, just charge the patient a lot of money for it so they get a more powerful placebo effect. You don’t need to waste your time and money getting trained in acupuncture.

  14. Jann Bellamy says:

    In this “case vignette” the authors are free to make up whatever facts they choose, to wit, a 45-year-old construction worker with a 7-year history of intermittent low back pain, and a gradual increase in back pain over the past 4 months despite certain vaguely described interventions. The pain is described as “a dull ache in the lumbrosacral area with episodic aching in the posterior aspect of both thighs.”

    Question: does such a patient exist? In other words, have they described a plausible clinical presentation? If so, shouldn’t they state the basis for creating this patient? All the authors say is that low back pain is common. What are they suggesting? That all low back pain patients be “referred” for acupuncture? Or only ones who present with the symptoms they manufactured? If the former, why bother with the patient description? If the latter, what is the connection between the referral and this patient’s specific symptoms.

    Second, it is worth noting that, even with the freedom to create a patient and his clinical presentation, they fail to construct a plausible basis for the “referral” to an acupuncturist.

  15. tmac57 says:

    pmoran said “…a variety of interventions that involve taking the patient out of their normal environment for a spell once or twice a week, keeping them still while a mildly invasive and often painful treatment ritual is performed and certain social and medical interactions may occur.” Sounds kind of kinky when you put it that way. Maybe S&M is really a form of acupuncture ;)

  16. Ben Kavoussi says:

    There conclusions complement a post on this site called “James Reston’s Tooth of Gold:”

    “Systematic reviews of literature, notably one by Howard H. Moffet of Kaiser Permanente Division of Research, indicated that although acupuncture can affect outcomes and is distinguishable from a placebo, trials that compare distinct needling regimens often do not indicate statistically significant differences in outcomes. Indeed, the dominant scientific rationale for acupuncture involves the release of neurochemicals (such as endorphins) by the irritation and injury it causes, but there is little evidence that this release depends on any specific points or means of stimulation.”

    Read the rest at http://www.sciencebasedmedicine.org/?p=928

    Ben Kavoussi

  17. JMB says:

    The old style saline injections and sugar pills used for placebo effect in the 50′ and 60′s fell into disrepute because of the argument of respect for the patient. The idea that we had to actively engage the patient in the healthcare process led to the ideas of informed consent and patient education. Informed consent and patient education really curtailed the use of blatant placebos.

    The new wave of integrative medicine seems to be abandoning the idea that we try to share the best of our scientific knowledge with the patient. I guess the paternalistic attitudes of the 50′s and 60′s has been reincarnated with an elitist attitude that the patient will be fooled by “attention from the acupuncturist, and highly focused, spatially directed attention on the part of the patient”, or homeopathy, or naturopathy, or Reiki, etc.

    Where is the patient education in the clinical scenario presented in the paper? I have not read the entire paper. Perhaps it was mentioned in the full article? Since the patient asks for acupuncture, is the clinician absolved of the responsibility of education? Can you refer the patient for $1200 in treatments by a specialist, while explaining to them that the treatments could be done by a high school student with 1 hour of training just as effectively? At least the high school student would also mow your lawn 10 times in addition to the acupuncture treatments for $1200.

  18. Karellen says:

    However, [placebo effects] are likely to be illusory – an artifact of observation and reporting, not a real improvement in the patient’s condition. In real science-based medicine, that is the underlying assumption – placebo effects are largely illusory – a variable to be controlled for.

    Are you sure? Could you comment on the studies mentioned by Ben “Bad Science” Goldacre in the videos linked in his Placebo! Nocebo! article? (Links to the actual studies in the comments section.)

    I bet a 10 minute video is all that is necessary. In fact I bet even that is not necessary – you could probably fake it well enough to get a maximum placebo effect without any prior demonstration.

    I actually doubt that, for the same reasons that Rork mentioned. If you take someone to a room that’s bare except for a couch, lie them down, stick needles in them for 20 minutes, and then set them on their way, you might get some effect. But if you take them to a room with full-colour anatomy diagrams (including Qi lines) and large Hànzì on the wall, plants around (bamboo?), Chinese music playing softly, wear some elements of traditional Chinese clothing, have a stethoscope round your neck (need not be used and can be taken off before actual treatment begins), spend 20 minutes talking to the client about the history of their problem, throwing in Chinese words they’ll be unfamiliar with, lie them down, spend another 10 minutes waving your hands around “diagnosing” where the “blocks” in their Qi are (or even “finding” their Qi lines, or something equally vague), then stick needles in them for 20 minutes, and then ask them to lie there for another 10 for some other reason (maybe turn the music up a bit at this point) – then from what I’ve read, you might get a much better response. And all those things I’ve just come up with off the top of my head. The point is, AIUI, the ritual surrounding the treatment is the only thing that matters; that’s the point of the placebo effect. So you could probably do a lot of useful training on how to use the placebo effect to its best extent – and so could real doctors for that matter.

  19. urodovic says:

    Ahh the placebo effect… we just cant keep up with it…this is from today…

    http://www.sciencedaily.com/releases/2010/08/100802165412.htm

  20. Dr Benway says:

    urodovic, I do not like that study.

    The experimenters created four placebo groups. They got a statistically significant increase in dopamine release in one of the four groups. Meh.

    Meanwhile, they spent an assload of money and exposed 35 people to radiation for no good reason.

    THE PLACEBO IS NOT THAT INTERESTING, PEOPLE!

    It’s a conditioned response at best. Think Pavlov’s dogs. You can extinguish the drooling-after-the-bell response just by ringing the bell occasionally without delivering any food. The dogs will gradually learn that bell = nothing.

    Likewise, placebo responses fade with time. The reality of nothingness eventually trumps the illusion of usefulness.

    In medicine the rule has been: must be better than mere placebo. Only in the case of a strangely particular set of therapies (supplements, chiropractic, acupuncture, yoga, homeopathy, energy medicine) that people try to bullshit their way around that rule.

    The world of folk medicine is a lot more than herbs, yoga, and needles. Why no voodoo, CAMers?

    Our local hospital just opened a department of “integrative medicine.” Guess where the head MD trained….

    If you said, “University of Arizona in Tucson,” you would be correct!

  21. MKirschMD says:

    This is not about science. We’re swirling in a culture of feel good ‘wellness’ and magical healing. There is a CAM Industrial Complex that is exploiting the limitations of conventional medicine and those who seek its elixirs. These treatments leapfrog over any evidence testing and ask for faith of efficacy. Even traditional hospitals have jumped onto the wellness train. http://www.MDWhistleblower.blogspot.com

  22. DevoutCatalyst says:

    from the Online Etymology Dictionary:
    placebo
    early 13c., name given to the rite of Vespers of the Office of the Dead, so called from the opening of the first antiphon, “I will please the Lord in the land of the living” (Psalm cxiv:9), from L. placebo “I shall please,” future indic. of placere “to please” (see please). Medical sense is first recorded 1785, “a medicine given more to please than to benefit the patient.”

    For what it’s worth…

  23. baldape says:

    As I like to say, accupuncture is good for back pain in exactly the same way that a mother’s gentle kiss is good for a child’s scraped knee. The difference is that the mother doesn’t charge $100/hour for the service.

  24. BillyJoe says:

    Dr. Benway,

    “THE PLACEBO IS NOT THAT INTERESTING, PEOPLE!

    It’s a conditioned response at best. Think Pavlov’s dogs. You can extinguish the drooling-after-the-bell response just by ringing the bell occasionally without delivering any food. The dogs will gradually learn that bell = nothing.

    Likewise, placebo responses fade with time. The reality of nothingness eventually trumps the illusion of usefulness.

    In medicine the rule has been: must be better than mere placebo. Only in the case of a strangely particular set of therapies (supplements, chiropractic, acupuncture, yoga, homeopathy, energy medicine) that people try to bullshit their way around that rule.”

    Nicely put. :)

    Reminds me of a relative who is convinced her collection of symptoms are due to allergies – but she’s always allergic to something else. As the treatment for one allergy loses its effectiveness, there’s always another allergen around the corner to be discovered and treated.

  25. BillyJoe says:

    Someone please sack the moderator.
    He is less than useless. :(

  26. pmoran says:

    I again point out that the Hrobjartsson study cannot provide an accurate appraisal of placebo potential.

    All conventional clinical trials are performed in such a way as to minimise placebo influences. The subjects in the placebo arms of the trials included in that study would not even know if they are supposed to feel better or not.

    How is that a fair test of placebo?

  27. JMB says:

    I have no expertise in this field of medicine, and I haven’t bothered to do any research, so feel free to shoot me down on this. I think that IV saline or RL for a hangover cure is mostly a placebo effect, because you have to run the IV drip at a fairly rapid pace to provide hydration faster than drinking fluids.

    I know there are plausible explanations, but has there ever been a randomized controlled trial to indicate that the IV is better than placebo for a hangover? Certainly, there are many medical personnel that swear by that cure. So if I am right, then many on this thread may have personally experienced placebo effect.

    Hangover — a good example of a self limited disease that will probably be recurrent. I’ll drink to that!

    If I’m wrong about that, then consider medical student’s disease, a nocebo effect. Surely everybody has witnessed their classmates suffering from this? Of course, not ourselves.

  28. E says:

    Here’s yet another annoying item to come across:

    http://www.youtube.com/watch?v=VBDMA6gwGIQ

    ‘No, no, not veddy good. Veddy bad!’

  29. Doc says:

    If it’s contextual then that means the sham and traditional acupuncture are the same but both are probably different from other treatments. How different one can’t tell from this study.

    So the recommendations would make perfect sense. Acupuncture or sham acupuncture is almost like the movie Matrix and Neo’s choice to take blue or red pill from Morpheus.

    I haven’t read the study but how matched are the treatment groups for beliefs and cultural background of the patients.

    Then of course there is the issue of isolation between the traditional & sham arms of the experiment. They maybe entangled in something analogous to quantum superposition and pyschokinetic / telepathic effects between the experimental arms are make the results the same.

    All much more interesting than boring neurology ;-)

Comments are closed.