Mar 25 2011

Another Acupuncture Fail

Here we go again. What if I said to you that drug X and a placebo of drug X (i.e. a placebo) worked better for the relief of a subjective symptom than no treatment, but no different from each other? I conclude from this that the “drug cohort” (whether the real drug or the placebo of the drug) did better than the no-treatment control group.  This may mean that the drug works through a general care effect and patient expectancy.

Let’s say, rather, that a pharmaceutical rep were saying this about their latest drug. It would be obvious that the drug company decided to use some Orwellian new speak in order to contrive a sentence in which they get to say that their drug works (by a general care effect). And in order to obscure the fact that their drug worked no better than placebo, they refer to both the drug and the placebo as the “drug cohort” and compare them both to a no-treatment group. The “drug cohort” had an effect.

In the real world of scientific medicine (not the bizarro world of CAM), when a treatment works no better than the placebo control we conclude that – the treatment does not work. Looking past all the obvious spin above, a more honest conclusion would simply be – drug X does not work for the tested symptom. Period. There is no “drug cohort”, and you don’t have to confuse the reader by calling placebo effects a “general care effect” or “expectation.” These are placebo effects.

Acupuncture still exists in the CAM bizarro world, and has not crossed over to the real world of scientific medicine. This is because, in my opinion, acupuncture does not actually work. It is not based upon any legitimate principles of physiology or anatomy, its underlying theories are mostly pre-scientific superstitions, and proponents cannot demonstrate in controlled clinical trials that it is more effective than placebo.

In the past, before the world collectively lost its mind, when a treatment consistently failed to demonstrate any plausible mechanism and any effect greater than placebo in clinical trials, eventually it would be discarded as a dead end – a failed hypothesis. But today treatments that fail to demonstrate any basis in reality or any clinical effect happily live on in the world of complementary and alternative medicine (CAM).

In the CAM world, rather than adjusting beliefs to the evidence, proponents adjust their approach to and interpretation of the evidence in order to accommodate their beliefs. For example, I wrote two days ago on Science-Based Medicine about Dr. Tonelli, who wrote:

Orthodox medicine should consider abandoning demands that CAM become evidence-based, at least as “evidence” is currently narrowly defined, but insist instead upon a more complete and coherent description and defense of the alternative epistemic methods and tools of these disciplines.

Let’s see how a recent acupuncture study avoids the “trap” of “narrowly” defining evidence (meaning defining evidence as actual evidence for efficacy).

Just published in PLoS One is an article comparing verum acupuncture (where needles are inserted into alleged acupuncture points) to sham acupuncture (in this case, fake needles were not inserted, just pressed against the skin, and not at the alleged proper acupuncture points) in the treatment of radiation-induced nausea in cancer patients. These two treatment groups were compared to a no-intervention group (usual care).

This is a reasonable trial design. It controls for the two main variables of acupuncture: needle insertion, and the existence of alleged acupuncture points. And it included a third usual care group.

The results of the study are fairly clear – there was absolutely no difference between the verum acupuncture and sham acupuncture groups (who had 7% and 6% nausea respectively). Both groups experience less nausea than the standard care group, at 15%.

In the real world, the interpretation of this result is very straightforward. Acupuncture does not work for radiation-induced nausea – it showed no difference between acupuncture and placebo acupuncture. Of course both groups showed improved over the standard care group  – and the study even documents why. Those subject receiving acupuncture, whether real or sham, had an expectation of benefit. Those receiving standard care did not have an expectation of benefit. Therefore any measured difference between these groups is due to the usual placebo effects that we control for in clinical trials.

Expectation can lead to not only a perceived decrease in symptoms, but decreased reporting of symptoms. Patients who think they are getting the treatment know they are expected to report an improvement, and so some will. Patients who are expecting a benefit may also have an elevated mood or decreased anxiety and this will affect their perception of their symptoms. The same is true of the attention from the caregiver – it may increase mood and reduce anxiety.

These effects are part of what we collectively call placebo effect, and is precisely why we do placebo-controlled trials when trying to figure out if an intervention has specific efficacy. We know there will be a measured benefit from the ritual of patient care – that is the baseline. And we know this effect is enhanced by various factors, such as expectation, the introduction of a novel element into patient care, and any type of physical contact.

If all we see in clinical trials are these kinds of non-specific placebo effects, then we conclude that the treatment itself has no physiological effect – i.e. it does not work.

It is remarkable that PLoS One allowed this study past peer-review with the conclusions as stated.

Patients treated with verum or sham acupuncture experienced less nausea and vomiting compared to patients receiving standard care, possibly through a general care effect or due to the high level of patient expectancy.

This kind of spin is highly deceptive, and is a failure of the peer-review process in this and similar cases. But it seems to becoming standard fare for CAM modalities – the double standard is in full effect.

Worse than the report itself is the press release, which boldly claims, “Acupuncture is Equally Effective with Simulated Needles.” This is spin that would make the most cold-hearted industry marketers blush or make P.T. Barnum proud. But it gets worse. In the body of the press release they quote:

“It’s important to remember that the effects of the treatment are valuable to the patients, even if they can be said to have been caused by unspecific factors, such as the manner in which the patients were taken care of and their positive expectations,” says Dr Enblom. “So our next step is to study which part of the acupuncture procedure actually are of importance, to make possible the use of those components to further increase quality of care.”

Let me paraphrase – our study shows nothing else than the fact that acupuncture does not work for radiation-induced nausea. But we are desperate not to admit that and to hide that inconvenient fact with confusing statements about placebo effects. Even better – we are actually going to try to justify further research into acupuncture by stating that we need to study the very placebo effects that show that acupuncture does not work.

Dr. Enblom wonders which components of acupuncture are producing the perceived benefit. I will give her a clue – it’s not the acupuncture component. Yet again we see evidence that it does not matter where you stick the needles, or even if you stick the needles (in other words, do actual acupuncture). All that matter is that you give the patient positive attention. Acupuncture is therefore nothing than a worthless (and not entirely safe) superstitious ritual that is being wrapped around good bedside manner, and then trying to take credit for the nonspecific benefits of the good bedside manner.

In a way it’s brilliant (in an evil deceptive way, that is). I have to admire it, like the way you admire Hannibal Lecter. They actually got away with showing acupuncture does not work, and then stating in the conclusions of a peer-review article that the treatment worked. Then in their press release calling for more research to show how it worked – when the answer is already in front of them. It didn’t work, there were only non-specific placebo effects.

Acupuncture = fail

Orwellian newspeak CAM marketing = another (regrettable) win



47 responses so far

47 Responses to “Another Acupuncture Fail”

  1. Skepticoon 25 Mar 2011 at 11:26 am

    Dr. Tonelli writes:

    Orthodox medicine should consider abandoning demands that CAM become evidence-based, at least as “evidence” is currently narrowly defined, but insist instead upon a more complete and coherent description and defense of the alternative epistemic methods and tools of these disciplines.

    None of these people can write in simple English, can they? I’ve read that passage several times in an effort to determine what Tonelli is trying to say under all that dense impenetrable text. I’m going to have a stab at parsing it in understandable language:

    Medicine should abandon demands for evidence, and instead concentrate on describing how the CAM treatment works.

    If that is indeed what he means (and I’ll admit, I could be wrong), he’s putting the cart before the horse. Before you explain how CAM works, you need to determine if it works.

  2. gebradenkipon 25 Mar 2011 at 11:42 am

    Is it considered unethical to subscribe placebo treatment when it has been shown that it “works” better than no treatment at all?

    And would the conclusion of the article be more acceptable to you if it read:
    Patients treated with verum or sham acupuncture reported less nausea and vomiting compared to patients receiving standard care, possibly through a general care effect or due to the high level of patient expectancy.?

    Or is that not the large issue for you? I’m asking because for my untrained eye the statement seems perfectly true to me when you make that adjustment, and safe placebo-treatment seems beneficial to the patient.

  3. CrookedTimberon 25 Mar 2011 at 11:55 am

    True believers will never get it.

    Aside from all the reasons that acupuncture should be completely abandoned, I would like to ask them why they don’t change the standard to sham acupuncture. It has been shown to be equally effective and comes with less risk by not causing numerous skin punctures that increase risk of infection.
    I think they are afraid that would detract from their feelings of ancient mystique.

  4. Karl Withakayon 25 Mar 2011 at 12:23 pm

    It’s funny how studies like this are not death blows to the practice of acupuncture.

    A logical, reasonable person might use this study to conclude that acupuncture is bogus, and any effects are due to the placebo response.

    A less reasonable, logical person might decide that the control has a clinical effect too, but even if you speculate some clinical effect from the control, why recommend the treatment over the control? Shouldn’t this be an opportunity for CAM to learn, grow, and adapt based on new evidence? In scientific medicine, if you determine two treatments are equally effective, you generally go with the less invasive, less risky procedure and discard the one with more risks and complications.

    If a test drug and sugar pill were shown to be equally effective at alleviating chronic pain not otherwise treatable, why would you recommend the chemically active drug with its various side effects and complications over the sugar pill?

    Shouldn’t the recommendation at least be against the treatment, since the treatment is not without risks and adverse events?

    Why don’t they conclude, “Hey, we’ve found a safer, simpler, and less invasive way to do acupuncture, just randomly poke with toothpicks, no needling required.”

    Study conclusion: It is clear from this study that insertion of needles is not needed and it does not matter where you needle or poke. The fundamental aspects and understanding of acupuncture are therefore invalid. We see no reason to recommend verum acupuncture over the control.

  5. Steven Novellaon 25 Mar 2011 at 12:35 pm

    gebradenkip – My problem is not the word “experienced” vs “reported” – it’s that the comparison between acupuncture or fake acupuncture vs standard care is not blinded. Therefore you cannot make any statements about the relative effectiveness. Any such reports are overwhelmed with other variables.

    Because the verum and fake acupuncture groups were blinded to each other, you can make a comparison, and that comparison showed no difference.

    The other commenters are correct – even if you did try to deceptively use this and similar studies to claim that acupuncture (even fake acupuncture) works, you would have to also conclude that
    - acupuncture points do not exist
    - all of the notions about how acupuncture works having to do with chi etc. are wrong and should be abandoned
    - acupuncture training is worthless and should be abandoned
    - the only perceptible benefit to acupuncture is from the attention of the caring practitioner, and so the ritual of acupuncture should be simplified to it’s safest elements, which could just be gently palpating random points on the body while the person lays down and relaxes.
    - It is unethical to stick acupuncture needles through the skin when there is no benefit to this.

  6. jaranathon 25 Mar 2011 at 12:42 pm

    But do we know that isn’t their stance toward accupuncture? Is it safe to assume that I’m missing some context here that makes it clear the authors are pushing actual acupuncture? I would bet that they are, but the way they’ve worded that conclusion does make it seem like, as gebradenkip seems to be suggesting, that the authors are really just interested in studying the nature of the placebo effect and possible clinical uses. Even the citation from the press release could fit that interpretation. Probably extreme tap-dancing, I was just wondering if there were more clues.

  7. daedalus2uon 25 Mar 2011 at 1:02 pm

    It seems to me that sham acupuncture works slightly better than real acupuncture. I seem to remember a similar result on other studies that used both. This is something that I would expect, that the very slight tissue injury of real acupuncture does trigger a slight “fight or flight” response which counters slightly, the placebo effect.

    If this is correct, then eventually, with enough studies, the superiority of sham acupuncture over real acupuncture will be shown.

    What will they do then? Open up a “new” school of acupuncture? One that uses toothpicks instead of needles?

  8. daijiyobuon 25 Mar 2011 at 1:08 pm

    Re: “in the CAM world, rather than adjusting beliefs to the evidence, proponents adjust their approach to and interpretation of the evidence in order to accommodate their beliefs.”

    By belief, I believe [!] here you’re basically talking about ‘how we summarize phenomena / claims retroactively and proactively as informed by continually amassing quality evidence’ as opposed to ‘belief as dogmatic articles of faith’ since such is inherently not-evidence-supported and stagnant.

    This resonates with something rather tautological I’ve recently read in the Canadian naturopathy netherworld / bizzaroworld:

    “the naturopathic profession in Canada recognizes the value of research and seeks to make appropriate use of science to further the understanding and advancement of naturopathic medicine” (see ).

    One has to wonder what is “appropriate use”. The foremost context is obviously not science and evidence, but serving ‘the naturopathic’ beliefy [in the static way I've defined it above].

    That new Canadian Naturopathic Foundation site claims that naturopathy is a “branch of medical science” while therapies such as homeopathy are “powerful” and can be premised on such things as a “vital force”.

    Yet science, honestly appropriately used, has demolished homeopathy and such vitalistic contexts: science / evidence subset belief [in the dynamic sense I've defined above].

    But for naturopathy, it is: dogma / belief subset science.

    Quite bizzaro, in the the sense of the reversal of values that Bizzaro is all about.

    So, a very narrow view is taken of evidence in a ‘cherry-picked’ or even denialist manner — to protect the ‘essentially naturopathic’ like the nonsense above.

    I’ve sometimes called this, and admittedly it’s awkward, ‘science as the sword of sectarianism’.

    By sectarianism I mean belief-based [and now I mean belief in the 'dogmatic, article of faith sense'], science-ejected or -exterior stuff either simply ignorant of scientific preponderance or deliberately so: static belief / sectarian system subset science.


  9. Draalon 25 Mar 2011 at 2:13 pm

    Would it be unethical to design an experiment where the patients are placed in a drug induced sleep and repeat the above experiment?

  10. sonicon 25 Mar 2011 at 2:16 pm

    If something that doesn’t work gets better results than standard care, why is it standard care?

  11. Steven Novellaon 25 Mar 2011 at 2:32 pm

    sonic – you cannot conclude that it gets better results than standard care, because the comparison was not blinded. And keep in mind – it was not acupuncture vs standard care, it was acupuncture + standard care, vs standard care alone.

    There are many factors that influence patients to report that they feel better, even when there is no objective evidence that they are (even in studies where objective outcomes are being measured). There are psychological and memory factors at play.

    Also – no one doubts that taking time to lay down and relax a few times a week can improve subjective symptoms that are subject to stress. They should have had a fourth group where the patients went through the same process as the acupuncture groups, but no acupuncture (real or fake). When such studies are done, what we find is that the only variables that correlate with improved symptoms are having a supportive caretaker and the relaxation. The acupuncture (real or fake) is entirely superfluous. You could replace the acupuncture part with placing a ripe tomato on their heads and that would probably work as well.

    So how does this all add up to acupuncture “working” or having any basis in reality?

  12. sonicon 25 Mar 2011 at 2:50 pm

    Dr. N-
    Good point about the unblinded aspect.
    I think the researchers are going to do tests as you suggest- (that’s what I take from the statement about finding what part of the procedure works).
    It will be interesting to see what the results are.

  13. Steven Novellaon 25 Mar 2011 at 3:35 pm

    I would like to propose a tomato study where we have one group who gets relaxation, encouragement, and a ripe tomato on their head. A second group gets relaxation, encouragement, plus a simulated tomato. A third group will get relaxation and encouragement with no tomato. And a fourth group will get usual care with no extra intervention (and all the other groups also get usual care).

    My hypothesis is that relaxation and encouragement will make people feel better, as has been demonstrated in countless studies already and really isn’t controversial or novel.

    Come to think of it – let’s just skip the tomato bit.

  14. jaranathon 25 Mar 2011 at 3:46 pm

    I like Draal’s idea. Could you (maybe?) do it on surgical patients while they’re under anesthesia (say, for anesthetic-induced nausea)? That way you aren’t putting anyone under specifically for the research. Or has that been done already?

    Of course, the horse of acupuncture is so thoroughly dead it’s fossilized, so it’s hard for me to seriously suggest beating it again.

  15. peteron 25 Mar 2011 at 4:20 pm

    Hey Steve, I’m so glad that you addressed this acupuncture issue. There was an article in my local paper talking about how more kids are getting acupuncture treatments on March 15th. One of the things that really strikes me is that one of the guys that was practicing acupuncture said “I don’t know how to argue with something that’s been around for 3,000 years.” It seems to me that a lot of people have a real fascination for ancient Chinese practices. I remember you talking about mediation on the skeptics guide to the universe one time. About how people thought it was magical but was really no different from relaxation. Maybe there is something similar here too. Maybe because acupuncture is a 3,000 year old Chinese practice, people think that there is something magical to acupuncture and that somehow boosts people into thinking that it will actually work.

  16. sonicon 25 Mar 2011 at 4:31 pm

    I think the tomato wouldn’t work as well as the no tomato. It might be harder to relax with a tomato on your head.
    But this is the kind of question that can best be determined by experiment.

  17. norrisLon 25 Mar 2011 at 4:36 pm

    This is a very interesting read for skeptics and sbm type people. Of course, this book has been trashed by CAM type people. Petr Skrabanek was a highly regarded professor of medicine at Trinity College in Dublin.

    I draw your attention to Follies and Fallacies in Medicine, chapter one on placebo and page 126 on acupunture.

  18. norrisLon 25 Mar 2011 at 4:39 pm

    Oh, I forgot to say that the figures on nausea of 15% for untreated vs 7 and 6% for verum and sham groups indicates a > 50% improvement over the untreated group. At least, this s how CAM people would represent the “results”

  19. tmac57on 25 Mar 2011 at 5:52 pm

    Steve Novella-

    I would like to propose a tomato study where we have one group who gets relaxation, encouragement, and a ripe tomato on their head.

    You may not be too far off base on this.According to The Ketchup Advisory Board, ketchup (a popular tomato derivative) has “natural mellowing agents” that have all sorts of salutary effects.

  20. tmac57on 25 Mar 2011 at 6:14 pm


    What will they do then? Open up a “new” school of acupuncture? One that uses toothpicks instead of needles?

    Karl Withakay-

    Why don’t they conclude, “Hey, we’ve found a safer, simpler, and less invasive way to do acupuncture, just randomly poke with toothpicks, no needling required.”

    I originally posted this as a parody,but now you are forcing me to consider starting a real franchise:

  21. BKseaon 25 Mar 2011 at 6:52 pm

    To tmac57: Sign me up for one of those Pick n Poke franchises! Sounds like a great way to serve an untapped group of suckers, I mean patients.

  22. Mlemaon 25 Mar 2011 at 7:28 pm

    if I’m the patient, i really don’t care whether or not this sort of study is valid – I want the placebo effect to be delivered to me. Go ahead, increase my expectation and faith, make me feel like you care, cause me to relax. If you need to poke me with some toothpicks to accomplish that, then I, as a typical patient, trust that you, as the therapist, know what you’re doing.

    i think that is why people continue to utilize these therapies – the therapies are a way to deliver the placebo effect. (which DOES work!)

  23. nybgruson 25 Mar 2011 at 8:01 pm

    It is unethical to intentionally give patients a placebo. And you, as a patient, wouldn’t want that either.

    What I tend to do is give people the real stuff and explicitly build up their expectation of it. Never false promises, but I give timeframes of when it should become effective, what the effect should be, etc. And then people tend to have a heightened placebo effect. But giving a sugar pill and doing the same knowingly is simply unethical and dangerous.

    Besides, how would you manage to keep that intentional placebo going? The medical community would have to learn about, disseminate, and knowingly use, a placebo treatment whilst keeping it a secret from the population at large. How would that be done? And can you imagine the backlash once it is inevitably discovered?

    “Doctors charge money for sham treatment that only has placebo effect. You have been scammed!”

    This is exactly why we say that if it has the same effect as placebo it doesn’t work

  24. elmer mccurdyon 25 Mar 2011 at 8:47 pm

    Just thought I’d drop by and mention again that, according to Dr. Novella’s source, when you see your doctor, there’s a one in four chance that any treatment he prescribes will have absolutely no compelling evidence, experimental or otherwise, to back it up, and other sources would say that it’s even higher. That is the “real world of scientific medicine.”

    So that’s part of the reason I can’t be too outraged about those few ill-chosen words in this study (the other being that I have experienced so much truly deserving of outrage at the hands of my real-world doctors).

  25. reedonlyon 25 Mar 2011 at 9:27 pm

    Just as Steve has been advocating a language change to “science-based medicine,” I wonder if it is not time to try a change of language re “placebo.” We talk about a placebo effect, and in the mind of most lay people, anything with an effect is effective. We tend to objectify effects, assuming the agent is in the object rather than in the subject. Lucky shirts, interventional prayer, acupuncture, homeopathy – any “reported” change in the subject’s mental (and by extension, physical) state are generally assumed to be because of some quality in the object. If, rather, we talk about a “placebo illusion,” then it would be more apparent that the actions are not, in fact, effective.

  26. nybgruson 25 Mar 2011 at 9:37 pm

    elmer: are you arguing that we should then throw it all out and just go ahead and go willy nilly with whatever damn well pleases us? Essentially your argument is that since 25% of treatments prescribed don’t have compelling evidence we should instead make that 100%? Think before you write. Jeez.

  27. BillyJoe7on 26 Mar 2011 at 1:51 am


    Firstly, it’s “one in four medical treatments” not “one in four treatments that your doctor prescribes”. There is a big difference. The medical treatments with no evidence tend to be prescribed less often by your doctor than those with evidence.

    Secondly, just because you have a complaint against doctors, doesn’t mean you cannot complain about altmed practitioners. You can actually complain about both you know – as Dr. Novella has done when the circumstances have warranted it.

  28. BillyJoe7on 26 Mar 2011 at 1:55 am


    “I think the tomato wouldn’t work as well as the no tomato. It might be harder to relax with a tomato on your head.
    But this is the kind of question that can best be determined by experiment.”

    I do hope this is tongue in cheek.
    Otherwise you have missed the point of Dr. Novella’s little exercise.
    Hint: “Come to think of it – let’s just skip the tomato bit.”
    (ie, let’s just skip the acupuncture bit ;) )

  29. SimonWon 26 Mar 2011 at 5:01 pm


    there are good reasons placebos aren’t generally used in medicine. One is the issue of trust, trust is central to any relationship and the patient/doctor relationship is very important one in which to preserve trust.

    The more critical one from a skeptical perspective is the evidence for a “placebo effect”, actual improvement in any given condition due to expectiaton derived from taking a placebo is thin on the ground. It is likely most of the effect is people under reporting symptoms to fulfil their belief of their doctor’s expectations. Certainly the effect seems to disappear when objective measures of outcome are used.

    In the UK GP appointments are cheaper to the NHS than acupuncture appointments, so it was be cheaper (and safer) for doctors to do the toothpick thing than to send patient to someone who might actually pierce the patient’s skin (with the attendant risk of infection).

    If there is an “placebo effect” it is likely to attend any treatment the doctor provides (and possibly none if the doctor presents it in the right way), and likely to be stronger if he has the patients trust. But I’m not yet convinced there is anything going on with the placebo effect apart for patient optimism and attempts to convince the doctor he is doing a good job.

    In this case patients who expected nausea had more nausea, I suspect the relevant reflect is partially under consious control, so probably all the doctor has to do is tell the patient that whilst nausea is a possible side effect it doesn’t happen to most people and you can probably skip all the messing around poking patients with things.

  30. HHCon 27 Mar 2011 at 1:04 am

    I recall having several treatments with an acupuncturist at a chiropractors office after a road accident in this past decade. The acupuncturist and chiropractor insisted I was improving! There was actually more, not less pain and nausea because of the incense and further injuries to the epidermis where the needles were deeply inserted, creating a bloody tomato result.

  31. recourseon 27 Mar 2011 at 1:57 am

    Dr Novella,

    Another awesome post. Please never stop working.

  32. BillyJoe7on 27 Mar 2011 at 6:00 am


    I agree that the placebo effect is grossly over-sold.

    For a start, it seems to be observed only when they treat largely subjective symptoms like pain and nausea and….well, actually, not much else.

  33. banyanon 27 Mar 2011 at 10:38 am

    When I start soaking people for their riches with my new tomato therapy, I’m totally citing you for support.

    “Dr. Steven Novella of Science-Based Medicine says it deserves further study!”

    “Dr. Novella agrees that it works as well as acupuncture!”

    I’m gonna be a millionaire.

  34. eiskrystalon 28 Mar 2011 at 4:01 am

    -I think the tomato wouldn’t work as well as the no tomato. It might be harder to relax with a tomato on your head.-

    It’s not that hard to balance a tomato on the average head. Just tell them that it’s sucking out the “toxins”, and improving their balance and posture (which is probably true)… then try to sell them a holographic-tomato magnet.

    Next time I get asked, i’m going to demand the sham acupucture. It’s clearly far superior, has studies that prove it works better than real acupunture… and let’s face it, cheaper, since anyone can do it.

  35. pekka son 28 Mar 2011 at 5:53 am

    Steve, I am a long time fan of your work and writings on acupuncture. But I think that your critique might be a bit misguided this time.

    I am aware of the way acupuncture- and CAM-supporters often try to twist negative results and interpret them as positive. But in this case I don’t think that the authors were trying to justify the use of acupuncture as you might think. I know for a fact that one of the authors, Martin Ingvar, is one of the good guys here in Sweden when it comes to critique of alternative medicines. He is very interested in studying the mechanisms of the placebo effect, since he acknowledges that most CAM therapies are useless beyond the placebo.

    Note that the authors of this study doesn’t say that acupuncture works in it self. Rather it’s other factors such as positive expectations, extra care, tactile stimulation and were given extra time for rest and relaxation. Factors that I suspect that Ingvar (and others) are eager to study further.

    I understand why you interpreted the authors as you did, but I’m pretty sure that they had other, more neutral, intentions.

    In the swedish press this study has mainly been reported as a fail for acupuncture. The method is being called placebo.

    The main author had the following to say to Dagens Nyheter (my translation):

    “Our study shows that it wasn’t the needles and the stimulation of the traditional acupuncture points that gave the effect. We believe that it was the general care taking of the patient that relieved the nausia.”

    “But in our study we could se that stimulation of the traditional acupuncture points and stimulation with the needle – qi as it’s called in traditional chinese medicine – wasn’t effective in our case with nausia caused by radiation.

    Pekka S

  36. BillyJoe7on 28 Mar 2011 at 7:21 am


    He may have been the token sceptic.

    Here is Dr Enblom as quoted in Steve’s article:

    “So our next step is to study which part of the acupuncture procedure actually are of importance, to make possible the use of those components to further increase quality of care.”

    This is their NEXT step.

    They’ve just shown that the acupuncture procedure is not important, and their NEXT step is to extract the parts that are important?

    This is spin.

  37. locutusbrgon 28 Mar 2011 at 9:33 am

    I think we need to throughly research this tomato treatment you speak of.
    I mean it is natural and we all know makes it a good treatment. Organic tomatoes would of course be better. You are a doctor therefore the treatment must be helpful. Several people have tried the treatment, and have told me that it was effective. At least half of people got better at some point after getting hit with the tomato. Maybe we could formulate a tomato homeopathy treatment to spray in their face so less trauma. I am pretty sure Oprah would be willing to help us. I will start the NIH application to study this. We will use rocks as a placebo. Of course the subjects will get to choose the rock, the tomato or the homeopathic preparation to make it fair. I figure a study of about 5-10 people will be enough. I think this holds a lot of promise at least a book. What do you think??
    PS:I know the medical pharmaceutical complex will try to stop but we must not let them. Tom-Knocko treatment, 3 out of 4 doctors just can’t be wrong.

  38. Steven Novellaon 28 Mar 2011 at 9:57 am


    What you say is interesting. It is possible the spin in Sweden has been different than here in the US.

    The conclusions in the actual paper, and the press release are still problematic, as I discussed. It is also possible, as has been pointed out, that different authors may have had different agendas.

    Ingvar sounds like he may be another Edzard Ernst – which I hope is the case. But he may have also been used a bit in this study, lending credibility to the study, which is now being spun to support acupuncture (at least in the US).

    Or it may be that the authors are trying to undermine acupuncture through stealth, in which case they may have been too stealthy for their own good. It should have been made crystal clear in the actual conclusion of this paper that the results show acupuncture does not work, and that the only thing that is going on here is placebo effects.

  39. pekka son 28 Mar 2011 at 11:26 am

    I doubt that the conclusions were spun to lend credibility to acupuncture. Studying where in the acupuncture procedure the effects lie might end up excluding the needles all together?

    I also doubt that Martin Ingvar would lend his name to research that he couldn’t stand for himself.

    One might think that he’s some sort of since he is a professor of integrative medicine and the chief of the Osher Center in Sweden. I know that other Osher-institutions have bad reputations for being a bit woo-woo. But the one in Sweden is actually very science-based and “skeptic friendly”.

    Anna Enblom might be more pro-acupuncture since she has practiced it in her work as a physiotherapist (as I understand it).

    Perhaps I’ll be able to provide you with her own point of view in the short future since we have plans of interviewing her for our podcast (I’m one of eight panelists in a skeptical, swedish podcast – Skeptikerpodden).

    I’ll write an update as soon as I know more…

  40. Karl Withakayon 28 Mar 2011 at 12:53 pm

    “Patients treated with verum or sham acupuncture experienced less nausea and vomiting compared to patients receiving standard care, possibly through a general care effect or due to the high level of patient expectancy.”

    That’s a very conservatively worded conclusion that seems to fit more with someone who is seeking to “harness the power of the placebo” rather than someone who is attempting to determine if the effects acupuncture are due to a placebo response.

    A more strongly worded conclusion would be:

    Study conclusion: It is clear from this study that the effects of verum acupuncture are indistinguishable from those of sham acupuncture. The insertion of needles is not needed, nor does it matter where you needle or poke. The fundamental aspects and understanding of acupuncture are therefore invalid. We see no reason to recommend verum acupuncture.

  41. Enzoon 28 Mar 2011 at 4:18 pm

    Studies like this getting published and positive press is worse than a double standard for evidence, I think.

    Drug companies routinely get slammed by the media for producing expensive medication with “minimal” benefit. Imagine what the negative press would look like if this were a pill produced by “big pharma”. It would be public outrage. Amazing.

  42. Wholly Fatheron 29 Mar 2011 at 1:22 pm

    I’ve noticed a trend in study design for CAM trials, in general, and acupuncture trials, in particular: the use of an unmasked usual care group. There is really no need for such a group in this study. It would be unusual to have such a group in an analgesic drug vs placebo trial. It adds to the size, cost, and complexity of the study design. It is often very ambiguous in the Method section of these papers exactly how the various groups will be compared, and whether or not appropriate adjustments are made for multiple comparisons.

    Use of the usual care group, however, gives the authors a “fail safe” when the acupuncture procedure does no better than sham acupuncture group. Just as the authors of this study have done, they can still spin the results by showing a benefit over the (unblinded) usual care cohort.

  43. Steven Novellaon 29 Mar 2011 at 2:11 pm

    WF – A usual care group can serve the purpose of demonstrating that the study is capable (power, outcome measure, etc.) of showing a statistical difference, if there were one. Otherwise, if you have only two groups with no difference you could argue that the study was not capable of seeing a difference.

    So, it’s a way of calibrating the study.

    If anything, then, the inclusion of such a group in this study strengthens the negative results of the study – absence of difference between verum and simulated acupuncture.

    But you are correct – it is being abused as a fail safe for negative studies.

  44. BillyJoe7on 29 Mar 2011 at 4:28 pm

    Steven Novella

    I’m with Wholly Father on this one but, then, I don’t understand this point:

    “if you have only two groups with no difference you could argue that the study was not capable of seeing a difference. ”

    How would it not be “capable” of showing a difference.

  45. Steven Novellaon 29 Mar 2011 at 5:07 pm

    BJ – as I said, for example you could argue that the outcome measure was not sensitive enough to pick up a difference. But if there is a difference between the placebo and usual care group, that proves the study had the power and sensitivity to measure that much of a difference.

    This is more common when comparing two drugs, you always include a third arm with a placebo to calibrate the power and sensitivity of the study.

  46. Wholly Fatheron 29 Mar 2011 at 7:48 pm

    Its worse than we thought!

    In “Methods” it is explicitly stated that patients were randomized between real and sham acupuncture. It is a little ambiguous as to where the standard of care group came from.

    Now, look at reference 13, a previous report from the same group. They discuss 4 studies, and it is clear that the standard of care group in the new paper was culled from a separate cross-sectional study! It was not randomized or even contemporaneous with the acupuncture groups.

    They also mention one analysis which favored the sham group (62% nausea) over the acupuncture group (70% nausea).

    The idea that these authors are presenting conclusions based on a statistical analysis comparing the combined data from 2 randomized groups to a historical control is about as irresponsible as can be.

  47. Wholly Fatheron 29 Mar 2011 at 9:08 pm

    mea culpa:

    In the PLoS paper the do explain the origin of the usual care group, but I missed this in the first reading. In the Discussion, they do acknowledge that their findings could be due to non-randomized nature of the usual care group. In any case, I suspect the inclusion of this group in the analysis was an afterthought when the comparison of the real and sham acupuncture groups failed to find a difference.

    Bottom line is that the authors ignored the primary analysis of their paper (comparison of the randomized groups), and based their conclusions on a comparison with an unblinded, non-randomized, non-contemporaneous group. How did this get through peer review?

Trackback URI | Comments RSS

Leave a Reply

You must be logged in to post a comment.