May 12 2009

More on Acupuncture

Published by under Uncategorized
Comments: 35

I am covering the in-patient service this month and so I am more busy than usual. I am trying to keep my blogging schedule without change, but it’s challenging, so forgive me if I occasionally miss a post or I am late.

For that reason, for my post today I am simply going to respond to a comment on my recent post on acupuncture and migraines. In response to this, frequent commenter, Sonic, wrote:

The conclusion that acupuncture does not work does not coincide with the evidence presented.

From the 2009 Cocrane review:

“Four trials compared acupuncture to proven prophylactic drug treatment. Overall in these trials acupuncture was associated with slightly better outcomes and fewer adverse effects than prophylactic drug treatment.”

This implies that acupuncture is a safe and effective treatment compared to the “proven” prophylactic drug treatment.

A drug company could go to the FDA with a study that showed their treatment to be better than the existing proven treatment and get approval based on that fact.

The conclusion that it is possible that the exact placement of the needles may not be as important as thought, does not invalidate the therapy.

Just because there might be a mistake in exactly how something works does not invalidate that it does work. The studies quoted would indicate that acupuncture works as well as or better than any other current therapy.

There is nothing in the evidence to indicate that the therapy did not work better than any existing therapy and better than doing nothing.

The fact that something works better than the proven therapy means that it works.

To conclude otherwise is to misread the evidence as presented.

Sonic has missed the central point of my position on this –  the trials comparing acupuncture to standard therapy or no therapy are not blinded, therefore we cannot draw any reliable conclusion from them. Placebo effects are well established, and they include the bias of the researcher as well as the desires of the subject. By Sonic’s criteria, placebos can be said to “work”.

The real question is – does placing acupuncture needles in the skin have any desirable physiological effect, and if so does it matter where we stick them? These are the two variables of interest as they form the basis of the very concept of acupuncture.

It is my contention, based upon the evidence and what we know from basic science, that there is no clinically relevant or significant effect from sticking needles in the skin. There is plausibility to the notion that needles may induce a physiological effect, such as counter-irritation, that could reduce the activity in various pain pathways in the nervous system. But this has not been adequately demonstrated. This also lends no plausibility to any medical claims for acupuncture, but we can stick to pain indications for now.

Whether or not sticking needles into patients had any significant biological effect, we would expect placebo effects from doing so – non-specific therapeutic effects and the artifact of bias and observation. So the fact that uncontrolled trials of acupuncture show these placebo effects tells us absolutely nothing.

Sonic brought up the point that in some studies acupuncture was compared to standard therapy. There are many problems with some of these studies as often the “standard” therapy given is not adequate and therefore is a poor control. But even if adequate standard therapy is given, the comparison is unblinded (patients know if they are getting acupuncture or not) and so the results are amenable to bias and not to be trusted. (As a side note, not all studies show a benefit over standard therapy, including some of the best studies, so even this dubious conclusion is not well established.)

And no, you could not get FDA approval for a drug only with unblinded data, especially for a subjective outcome like pain.

When we look at the evidence as it applies to the two real questions at hand – does it matter if and where we stick the needles – the evidence speaks quite clearly. There is no demonstrated difference between “true” and “sham” acupuncture – meaning sticking the needles where they are supposed to be stuck according to TCM principles, vs sticking them somewhere else.

It is completely absurd from a scientific point of view to conclude from such studies that sham acupuncture “works”. Such studies are absolutely not designed to form such conclusions. You can only conclude that there is no difference between true and sham acupuncture – and therefore there is no physiological effect dependent upon the location of the needles.

This, of course, completely contradicts any TCM principles of acupuncture points, meridians, and chi.

Further, these studies represent more reliable evidence than unblinded studies – and one of the principles of evidence- based medicine is that better data trumps weaker data. So the best data shows no difference between treatment and control – which in the real world of scientific medicine (as opposed to the altnerative world of CAM) this equals no effect.

More recent acupuncture studies (although not yet with migraine that I have seen) compare placebo-acupuncture to true acupuncture. In placebo acupuncture special needles are used that do not penetrate the skin, and neither the acupuncturist nor the patient can tell the difference. In such studies there has been no difference between placebo acupuncture and true acupuncture.

So what the evidence tells us so far is that is does not matter if or where you stick acupuncture needles. These are the two variables specific to acupuncture. Therefore, acupuncture does not work. The ritual surrounding the use of acupuncture likely has some non-specific effects, and there are the usual biases and artifacts that constitute placebo effects. But acupuncture itself appears to be worthless.

There is no other rational interpretation of existing evidence, despite all the misdirections of proponents who are trying to turn scientific logic on its head.

In fact – here is yet another placebo-acupuncture study – a large study, just published, comparing true acupuncture, sham acupuncture, and “fake” or placebo acupuncture (using tooth picks that do not penetrate the skin) in chronic back pain. The results – no difference between the groups.

Scientific conclusion – it does not matter where or even if you stick the needles, i.e. acupuncture does not work.

Conclusion of proponents:

The leader of the trial, Dr Daniel Cherkin, from the Group Health Centre for Health Studies in Seattle, said: “We found that simulated acupuncture, without penetrating the skin, produced as much benefit as needle acupuncture – and that raises some new questions about how acupuncture works.”

Wrong – these results raise the question of if acupuncture works, not how it works.  His colleague, Dr. Karen Sherman, is much closer to the truth:

“Maybe the context in which people get the treatment has effects that are more important than the mechanically induced effects.”

Right – the context, meaning the placebo effects. And “more important” meaning that there are only placebo effects and no “mechanically induced effects.”

In other works – it doesn’t work. Give it up. Move on.

35 responses so far

35 Responses to “More on Acupuncture”

  1. MBoazon 12 May 2009 at 1:40 pm

    You’ve gotta love being told that you’re misinterpreting the data by someone who clearly isn’t aware of the need for rigid experimental controls in acupuncture studies. I heard a similar argument from personal ignorance from, of all places, a graduate biology student last week. It was a pretty sad display.

    Anyway, even a nonspecific effect from acupuncture is, by any reasonable standard, a negative. True, acupuncturist will insert needles at so called “ashi” points, points that are tender or painful on palpation, but as you alluded to, one of the central tenets of acupuncture is that different points do different things. If all points do the same thing, and in fact all non-points do the same things still, that’s a pretty resounding negative for “acupuncture.” Anyway, the non-specific effects are crap, so it’s a moot point.
    Thanks for your vigilance Dr. N.

  2. Cronanon 12 May 2009 at 1:49 pm

    I’m glad you picked up on the recent back-pain study, especially the (I thought) hilarious statement by Dr. Cherkin.

    Well done, keep it up.

  3. rc_mooreon 12 May 2009 at 2:19 pm

    It is interesting how the cited recent study changes from “Fake acupuncture same as real acupuncture” in news reports to “Acupuncture effective in relieving back pain” when one does a Google search.

    Also, from a Reuters report:

    “Acupuncture brought more relief to people with back pain than standard treatments, whether it was done with a toothpick or a real needle, U.S. researchers said on Monday in a study that raises new questions about how acupuncture works.”

    The reporter seems to miss the point that toothpicks are not acupuncture, so the premise of the first sentence is incorrect.

    I am also not sure the study actually reaches the conclusion that acupuncture (sham or not) brings more relief than “standard treatments”. I don’t think the studies design can support this conclusion.

    Science –> Popular Media –> Personal Biases –> Psuedo-Science

  4. RickKon 12 May 2009 at 2:35 pm

    Regarding the back pain study, it’s on the Time website now:,8599,1897636,00.html?cnn=yes

    HERE is a great reason to de-fund NCCAM:

    “There is a lot of data now that acupuncture can have an effect on the nervous system, and that the nervous system then has effects on other systems in the body,” says Dr. Richard Nahin, acting director of the division of extramural research at the National Institute of Health’s National Center for Complementary and Alternative Medicine.

    “Everyone agrees that acupuncture is having some physiological effect. But we still don’t understand how it might be working,” says Nahin.

    No, it’s NOT having a physiological effect, it’s having a PSYCHOLOGICAL effect. Those that try something new and novel (“real” or “fake” accupuncture) feel better than those that just kept doing what they were doing. The question is: would you get EXACTLY the same response from “real” versus “fake” reiki, or “real” versus “fake” chiro, assuming it was a novel treatment for the patients? My guess is yes.

    It is no different than those studies that demonstrate sugar pills labeled “CAUTION! Do not exceed recommended dosage” are much more effective than sugar pills merely labeled “Safe and effective”.

    The power of the placebo!

    And of course, the article closes with the classic “what is needed is more studies”. BALONEY! Let’s waste not one more penny on this “research”. The case is proved, the work is done.

    But I do support health insurance covering toothpicks instead of chiropractic and accupuncture.

  5. rc_mooreon 12 May 2009 at 2:54 pm

    “No, it’s NOT having a physiological effect, it’s having a PSYCHOLOGICAL effect.”

    Well, the psychological effect could be causing a physiological effect, as opposed to something that only has a psychological effect.

    For instant, a magician’s indirection causes me to think the coin has disappeared, but this is not meant to result in any particular physiological effect.

    The placebo effect, however, can result in the release of endorphins, diminishing the intensity of pain in a real, physiologically measurable manner.

  6. TimKon 12 May 2009 at 2:59 pm

    Next time someone at med school suggests we should learn acupuncture, I’m going to recommend toothpick acupuncture instead. All the same benefits but less expensive and invasive!

  7. Scott D.on 12 May 2009 at 3:23 pm

    The article in Time magazine shows the level of corruption and dishonesty in the CAM community. Nowhere else would negitive results promote the use of a placebo?

  8. Kerry Maxwellon 13 May 2009 at 12:51 am

    One thing for certain, the way these stories are reported cause my wife and I great pain! The *toothpick acupuncture* story on our local news was accompanied by a great wailing and gnashing of teeth in our household. But we will salve our wounds with this post!

  9. wertyson 13 May 2009 at 8:15 am

    Acupuncturists should charge more for doing sham acupuncture, as it has all the benefits with none of the risks !

  10. jasonhenleon 13 May 2009 at 10:21 am

    There are so many acupuncture establishments in my town, I can’t help but hold my head in shame when I drive past them… but I can’t keep my head down too lone (I’m driving after all).

  11. sonicon 13 May 2009 at 4:59 pm

    Dr. N-
    Thank-you for the thoughtful reply.

    I agree that double blind is a good protocol for certain experiments.

    Data gathering and experimental design are interesting topics. The mathematical tools used to analyze the data (probability and statistics) place exact, mathematical constraints on the data that allow for the mathematics to be valid. Many experiments are run without enough attention to the constraints the math demands and the experimenters draw conclusions using incorrect mathematics.

    I agree that the ‘double-blind’ protocol is effective at diminishing certain difficulties that are inherent to data gathering in general, and specifically data gathering about the effectiveness of certain therapies.

    However, not all therapies can be subject to a double-blind analysis.
    But that is a different topic.

    From the article in “the Scotsman” of 12 May-

    “Researchers found that patients given the procedure, including a “fake version”, were much more likely to feel pain relief than those given normal remedies.”

    The evidence presented in the article goes on to indicate that the people who went to an acupuncturist were ‘much more likely to feel pain relief,’ and have ‘substantial improvement in function’ compared to those who did not.

    So when I have pain in my back, this evidence would indicate that I would be more likely to have less pain and substantial improvement in function, if I go to the acupuncturist. If I don’t see the acupuncturist, I’ll be ‘only slightly better’.

    I would agree with you that we cannot conclude that the acupuncture is what is causing this discrepancy in outcome. As you point-out, “the trials comparing acupuncture to standard therapy or no therapy are not blinded, therefore we cannot draw any reliable conclusion from them.”

    I would point out that the test comparing ‘normal’ acupuncture with other versions are not blinded either and therefore we cannot draw any reliable conclusions from them either.

    So we are left in the uncomfortable position of having to deal with the fact that those who went to the acupunturist got much better results than those who did not, even though we can not determine that it is the acupuncture that is the cause of this much better result.

    You say
    “The real question is – does placing acupuncture needles in the skin have any desirable physiological effect, and if so does it matter where we stick them? These are the two variables of interest as they form the basis of the very concept of acupuncture.”

    But there are two question of even more interest than those you ask-
    “How do I get over this back pain?” and “How do I get my function back?”

    The evidence presented in the article indicates that the people who went to the acupunturist were much more likely to do better on both counts than those who did not see the acupunturist.

    This indicates that going to the acupunturist is better than using standard therapies. This tells us that either a) there is something workable about acupuncture or b) that the standard therapies are actively harmful or c) both a) and b).

    I agree that it is not possible to conclude that acupunture is the cause of the better outcomes. But I would not draw a negative conclusion based on the evidence either.

  12. CKavaon 13 May 2009 at 6:47 pm

    “I would point out that the test comparing ‘normal’ acupuncture with other versions are not blinded either and therefore we cannot draw any reliable conclusions from them either.”

    Emm… but that’s not true? The recent study was blinded in that patients did not know whether they were receiving real acupuncture or sham acupuncture treatments.

    You can’t blind a patient who doesn’t even go to get acupuncture but you can make those who go believe they have received the treatment when you use a suitable sham treatment. That’s why we can draw reliable conclusions from comparisons of patients receiving real and sham treatments. They don’t know the difference.

    “The evidence presented in the article indicates that the people who went to the acupunturist were much more likely to do better on both counts than those who did not see the acupunturist.”

    It also showed that you can go to a fake acupuncturist who touches your back with toothpicks and you get the same benefit. In essence your advocating that we should promote a placebo effect as a valid treatment due to a single study in which the patients were well aware when they were in the group receiving extra attention from acupuncturists and when they were just getting regular treatments which were uncontrolled and variable (some included exercise, some painkillers etc. etc.).

  13. Karion 13 May 2009 at 7:24 pm

    Does anyone know where this study was published? I keep seeing “findings published” or “published in a medical journal” but the name of the journal is never identified.

  14. Karion 13 May 2009 at 7:30 pm

    Never mind. Found it.

    Cherkin DC, Sherman KJ, Avins AL, et al. A Randomized Trial Comparing Acupuncture, Simulated Acupuncture, and Usual Care for Chronic Low Back Pain. Archives of Internal Medicine. 2009;169[9]:858-866

  15. […] Steven Novella’s Neurologica blog which also addressed this study here. […]

  16. Steven Novellaon 13 May 2009 at 8:52 pm

    I discuss this study in more detail here:

    One further point – the study actually did not compare acupuncture (real or fake) with standard therapy. It compared acupuncture + standard therapy to standard therapy. Therefore you cannot even say that this study shows acupuncture has a better outcome (even if only due to placebo effects) than standard therapy, because the two were never compared.

    What this study clearly shows is that the addition of a novel intervention which involves the typical ritual of performing acupuncture had an added placebo effect to standard therapy for chronic back pain.

    This study, combined with the totality of acupuncture research, in my opinion makes it medically unethical to stick acupuncture needles into patients. It clearly adds nothing but is invasive and entails some risk.

    It also shows that there is no utility to studying where to place acupuncture needles. You can basically randomly poke patients with toothpicks and get the same benefit.

  17. sonicon 14 May 2009 at 4:16 am

    The person doing the treatment knows if they are doing a sham treatment or the real thing- therefore the trial is not (and probably) can not be double-blind.
    This is the point- not all therapies can be studied in that manner.
    (We can’t do a double-blind study of the effects of jumping off the Golden Gate Bridge, I guess we will have to go with the anecdotes on that one)
    All experiments suffer flaws from the mathematicians stand point. But we know from experience that some flaws are worse than others. (One big flaw is that conclusions are often made about populations that were not randomly drawn from for the experiment- for example)

    Dr. N-
    Perhaps the main difference in evaluation is in the priors— just a thought. It seems to me that you are willing to draw conclusions based on an unblinded trial and then use those conclusions to draw further conclusions that do not match the major findings.

    Given that I think your analysis is based in logic, I think our priors must be way different in this case.

    I’ll bet I’m right about that…

  18. Steven Novellaon 14 May 2009 at 7:36 am


    The acupuncture arms of this trial were single blind, which is better than unblinded. Yes, double blind is better. There have been double-blind studies in acupuncture – with opaque sheaths containing needles that either penetrate the skin or don’t and even the acupuncturist doesn’t know. These are all negative.

    Further – unblinding will tend to produce a false positive, not false negative. So we can conclude that the blinding in this trial was probably adequate.

    Bottom line – whenever the two variables of location and skin penetration are controlled for there is no difference from these variables. The better the trial, the more likely it is to be negative. If we take the totality of acupuncture research what we find is a consistent placebo effect from the ritual of acupuncture, and no documented effect related to the location of acupuncture points (which therefore probably don’t exist), or the penetration of the skin by needles.

    Acupuncture = placebo medicine

  19. tmac57on 14 May 2009 at 9:22 am

    Title: ‘A Randomized Trial Comparing Acupuncture, Simulated Acupuncture, and Usual Care for Chronic Low Back Pain.”

    Subtitle: ‘Who moved my Chi’s’

  20. CKavaon 14 May 2009 at 7:01 pm

    Sonic> I know it wasn’t double blind. However, as Dr. N pointed out the acupuncture, personalised acupuncture and sham puncture treatments were blinded for the patients. Obviously double blind would be better but single blinding is better than no blinding.

    Is it your contention that we can’t use clinical trials to establish the efficacy of acupuncture? If so why? This trial is a good example of how to compare acupuncture treatments against a placebo treatment. Even if double blind trials weren’t possible (although Dr. N seems to have examples) that wouldn’t make single blind studies useless or mean that we couldn’t use them to show that there is a high chance of a treatment being a placebo.

  21. sonicon 14 May 2009 at 9:21 pm

    I do think it is possible to use clinical trails to study acupuncture. I’m not sure those trails can be double-blind, which means that they are not as certain as to meaning as trials that are.

    In the case presented (and this case is fairly typical of the evidence) what we know is that the people seeing the acupuncturist had a better result than those who did not. It is not clear why that is true.

    My point is that if a person gains function and losses pain, you can call the treatment whatever you want, but I would say that it is worth a try given that the treatment has few or no adverse reactions. (Which is a fairly common finding as well)

    Am I advocating placebo? I’m not sure.
    I am advocating the notion that improved function and pain relief is better than lack of function and pain. That’s all.

  22. CKavaon 15 May 2009 at 5:29 am

    Sonic> OK, I see your point but you also surely must see that going to a fake acupuncturist who touches your back with toothpicks is also demonstrated in this study to provide just as much relief as regular acupuncture.

    That strongly suggests the benefit is from the extra attention and the placebo effect. If the sham acupuncture had not shown the same effect

    Would you recommend that people take a drug that was shown in a trial to be no better than a placebo? It’s very likely that if you included a group who knew they were just getting ‘standard’ treatment (as in this acupuncture study) the ones getting the new drug and the ones thinking they are getting it would feel better due to the placebo effect. It doesn’t mean the new drug is any more effective at relieving pain. In fact, if it produced an effect that was indistinguishable from the effect of the placebo treatment it is almost certain to be the placebo effect.

  23. CKavaon 15 May 2009 at 5:31 am

    That sentence in the middle should have read ‘If the sham acupuncture had not shown the same effect then things would be different but it did’.

  24. Watcheron 16 May 2009 at 2:13 pm

    On a side note, this made “Wait, wait … Don’t Tell Me” this week in the form of a limerick 🙂

    I hate needles flung with an uncouth flick,
    But I want chi to flow for my youth kick,
    Let us use wooden pins
    that do not pierce my skin,
    I think this might work with a _____________.

    For those that don’t listen to the show, a caller must get three out of four of these limericks completed to get the prize. Needless to say, Steve got me ready for this one 😀

    Though they didn’t do a very good job explaining that this shows that a placebo effect is occurring. Just that “Toothpicks or needles doesn’t matter, it still works the same!”

    This is the link to that specific portion of the show, though I’d suggest wasting an hour of your time listening to all of it.

  25. sonicon 17 May 2009 at 7:07 am


    “The researchers found that at eight weeks the individualized, standardized, and simulated acupuncture groups all improved their dysfunction scores significantly more than the group receiving usual care.”

    This tells us that the one fact that we can be certain of is that the people seeing the acupuncturist got better results than the people who did not see the acupuncturist and got the usual care.

    It is possible that the ‘usual care’ is actively harmful and that the people seeing the acupuncturist were not subjected to this harmful activity.
    It is possible that the acupuncturist does something helpful for the person.
    It is possible that both those statements are true.

    I would prefer to have less pain and better function regardless of what the therapy is called—

    Is it usual for placebos to get better results than the usual therapies? (Don’t forget the major finding here- those seeing the acupuncturist got better results than those getting the usual therapy). If so, I would suggest taking them, yes.
    If placebos don’t work better than the usual therapy, then I would say, don’t take them.

    I care more about what works than what it’s called.

  26. Steven Novellaon 17 May 2009 at 7:37 am

    sonic – this is an absurd position, and you have not addressed points already made.

    1 – we cannot conclude that the acupuncture groups had better outcomes, only that they reported better outcomes, and since this is an unblinded comparison the role of bias, expectation, and other psychological factors cannot be ruled out.

    2 – acupuncture was not even compared to standard therapy. Acupuncture + standard therapy was compared to standard therapy. So there was an added procedure known to produce a placebo effect added to existing standard therapy – and it showed what was perfectly consistent with a placebo effect, because the single-blinded variables made no difference.

    3 – standard therapy was not standardized in any way. It was – whatever people were doing anyway. Since subjects were selected because they were having symptoms it is reasonable to conclude that subjects who entered the study were not getting good results from whatever therapy they were currently using. They were selected for having already failed, and then some of them had an unblinded added procedure.

    Therefore patients were not randomized to a specific standard therapy, which was then initiated. Which is why this study was not even designed to compare standard therapy to acupuncture and it is scientifically absurd and completely deceptive to try to draw that conclusion from this data.

    This study was only capable of comparing cookbook, individualized, and placebo acupuncture – which showed no difference. No difference between acupuncture and placebo means acupuncture does not work.

  27. tmac57on 17 May 2009 at 10:30 am

    It seems like many of the “yes but” comments concerning all forms of CAM keep coming back to “placebo effects make people feel better,so it works!” I wouldn’t be surprised if practically any elaborate ritual that someone could dream up would produce the same kind of relief as the study above. Would we then conclude that it too is a legitimate treatment. Some will say yes, but to me it’s like seeing a magic trick that you can’t figure out, and then saying ” I guess that coin really did disappear because I saw it”. Subjective experiences are tricky to make sense of in science.

  28. sonicon 18 May 2009 at 3:30 am

    Dr. N-
    I agree with your statements, I’m just not sure they apply to the study at hand.

    To wit:

    I think you are mistaken about the groups studied-

    “The participants were randomly assigned to one of four groups: individualized acupuncture, … standardized acupuncture, … simulated acupuncture, … or usual care, which is standard medical care.”

    So there are four groups, three seeing an acupuncturist and one getting usual care. (The word in the above sentence is ‘or’.)

    “The researchers found that at eight weeks the individualized, standardized, and simulated acupuncture groups all improved their dysfunction scores significantly more than the group receiving usual care.”

    I assume that the ‘functionality’ is not completely subjective and that the people determining functionality were blinded- this might be an error on my part.

    I have not read the original study (I couldn’t download it from where I am), so I am relying on the published reports in news articles and the procedures that I know from other studies. This could be a mistake on my part.

  29. jo5efon 18 May 2009 at 5:39 am

    It would be nice to see a similar study with an extra category: Usual care + made up but convincing sounding alternative therapy: aromamagnetism or something, throw in a bit of ritual, ancient ethnic history and ego massaging of the patients. If the patients in this group also improved dsycfunction scores cf usual care would the researchers conclude”there must be something to this aromamagnetism after all?”.

  30. daedalus2uon 18 May 2009 at 9:16 am

    sonic, you are mistaken. The authors describe their protocol in great detail, and all patients had access to “standard care” for the duration of the trial.

    The protocol they planned to use is available open access.

    All patients had access to what was termed “standard care”. Acupuncture and simulated acupuncture was added to what was termed “standard care”. No patient was denied any treatment that was considered to be “standard care”.

    Their 3 primary aims were (paraphrased):

    1. Is acupuncture and simulated acupuncture more effective than “standard care”?

    2. Is acupuncture needling more effective than simulated acupuncture?

    3. Is individualized acupuncture more effective then generic acupuncture?

    Hypotheses 2 and 3 were falsified. Acupuncture needling was not more effective than simulated acupuncture and individualized acupuncture was not more effective than generic acupuncture.

    The authors expected acupuncture needling to be more effective than sham acupuncture. It wasn’t. They expected individualized acupuncture to be more effective than generic acupuncture. It wasn’t. They concluded (from the abstract):

    “…tailoring needling sites to each patient and penetration of the skin appear to be unimportant in eliciting therapeutic benefits. These findings raise questions about acupuncture’s purported mechanisms of action. It remains unclear whether acupuncture or our simulated method of acupuncture provide physiologically important stimulation or represent placebo or nonspecific effects.”

    The statistical analysis of the study protocol projected a 99% chance of achieving a significance of 0.05 in measuring differences between the different modes of acupuncture. There was no significant difference, the different modes are indistinguishable.

    If the different modes are indistinguishable, then those acupuncture practitioners who believe they are distinguishable have been mistaken. Those acupuncture practitioners who persist in the belief that they are distinguishable in the absence of data are deluding themselves.

    There might be a very slight difference that the trial didn’t have the power to measure. It can only be a slight difference, and not a difference that is of therapeutic significance. EBM has a difficult time trying to distinguish between the efficacy of different placebo treatments. I think it is a waste of resources to try and do so.

  31. sonicon 19 May 2009 at 4:37 am


    Thank-you for the link.
    It is clear from the protocol that this particular trial is close to worthless.
    They did not examine the populations they intended to draw conclusions about (getting acupunture vs. getting usual treatment.)
    They failed to test a null hypothesis.
    The tests of efficacy did not include examinations that were objective. (In the study of knee pain ‘range of motion’ can be determined objectively, for example)
    The statistical analysis is pure hokum.

    I would suggest a different study to draw any conclusion about this or any other treatment as these flaws make the whole thing questionable at best.

    The original post was about a different study, I guess the idea that this is better than the one we originally discussed was a mistake.

  32. daedalus2uon 19 May 2009 at 8:47 am

    sonic, did you read the paper describing the protocol that I linked to? The study linked to in the original post derived from that protocol. They did look at 3 types of acupuncture, with real needles in individualized acupuncture points, real needles in generic points, and with fake needles (aka toothpicks) and compared it with no acupuncture.

    The lead author of the protocol is also the lead article of the paper describing the study.

    The raters were blinded to the treatment given the patients so the outcomes couldn’t have been subjective. Back pain is something that acupuncturists believe and state (and advertise and charge for) that acupuncture is effective at treating.

    What protocol would you suggest to test the efficacy of acupuncture?

  33. sonicon 19 May 2009 at 5:21 pm


    Good question- here are some thoughts-

    The protocol would have to be set-up so that it would be possible to compare the treatments one is studying.
    So if you wanted to know if acupunture worked as well or better than the current standard therapy, you would give one group one thing and the other group the other thing. This is not how this study was done.

    The protocol would have to study the actual treatment. This study put the time, number and exact placement of needles in place before the treatment began. This is not how acupucturists actually work (as far as I know).

    I would like a study that includes an actual physical, objective examination of results. In the case of artheritis of the knee, for example, range of motion tests (a number of degrees) can be done. This test gives results that do not depend entirely on ‘how does it feel now?’ type of results. This study did all the follow up over the phone.

    The hypothesis that should be tested statistically is ‘it makes no difference’.

  34. Steven Novellaon 20 May 2009 at 8:08 am

    FYI –

  35. ghulseon 30 Aug 2009 at 2:24 pm

    Novella: It is my contention, based upon the evidence and what we know from basic science, that there is no clinically relevant or significant effect from sticking needles in the skin. There is plausibility to the notion that needles may induce a physiological effect, such as counter-irritation, that could reduce the activity in various pain pathways in the nervous system. But this has not been adequately demonstrated. This also lends no plausibility to any medical claims for acupuncture, but we can stick to pain indications for now.”

    Hello Steve,

    I am a big fan of both your neurologica blog and the skeptic’s guide podcast. I have here a question related to acupuncture and a condition my wife has, called brachioradial pruritus. Just so you know, I’m definitely not a proponent of acupuncture in any way, but many of these medical articles I have come up with seem to suggest it as a viable treatment for this condition. Although it could be these articles are outdated and that more recent evidence is more definitive that acupuncture is not effective? Or maybe acupuncture could show some efficacy for these specific neuropathic conditions?

    Here’s the background. This summer my wife began having a real problem with itchiness in her arms and legs (especially the arms) which apparently is neuropathic in nature. My wife describes it as something that feels almost but not quite like itching. She gets the most relief from ice, but also with Sarna lotion. She thinks the Sarna lotion’s menthol and camphor somehow “distracts” the signals from her skin cell receptors (something to do with the gate control theory of pain). My wife had this problem once before and it went away on its own after a couple of months.

    I call it “itchy arm syndrome.” But in our research we have learned that this condition is called brachioradial pruritus.

    The primary cause is one or more of the following:

    – excess exposure to ultraviolet radiation (sunlight, primarily)
    – cervical spondylosis (arthritis of the neck around cervical joint C6)
    – other compromises of the C6 joint (narrowing of exit foramen where the nerve root branches off from the spinal cord and heads off toward the arm)

    My wife’s brachioradial pruritus is almost certainly related to being out in the sun too much. This latest incident occurred soon after a week-long trip to the beach in South Carolina.

    Anyway, several of the medical articles we’ve come up with point to a small study done in England. Here’s the link to an article in the the British Journal of Dermatology (Vol. 155, Issue 5, Pages 970-976; published online: 31 Jul 2006)

    Excerpt: In a recent retrospective study, symptomatic relief of neuropathic pruritus was reported in 12 of 16 patients treated with acupuncture.15 Twelve patients had symptoms of brachioradial pruritus, seven of notalgia paraesthetica and four of meralgia paraesthetica. Relapse occurred in six of the patients within 1–12 months following treatment. Localized neuropathic itch such as notalgia paraesthetica, brachioradial pruritus and other peripheral neuropathies displays aberrant sensory innervation.16,17 The question is whether this sensory innervation may be a target for acupuncture. In other words, does acupuncture exert any effects on sensory nerve fibres in the skin?

    Anyway, my wife is seriously considering having acupuncture and I’d love to hear your opinion (while also completely understanding that I’m essentially the annoying guy at the cocktail party trying to get free advice from a doctor.) 🙂

    Cheers, and keep up the great work.

Trackback URI | Comments RSS

Leave a Reply

You must be logged in to post a comment.