One of the things that disturbs me the most about where medicine is going is the infiltration of quackery into academic medicine. So prevalent is this unfortunate phenomenon that Doctor RW even coined a truly apt term for it: Quackademic medicine. In essence, pseudoscientific and even prescientific ideas are rapidly being "integrated" with science-based medicine, or, as I tend to view it, quackery is being "integrated" with scientific medicine, to the gradual erosion of scientific standards in medicine. No quackery is too quacky, it seems. Even homeopathy and naturopathy can seemingly find their way into academic medical centers.
By far the most common form of pseudoscience to wend its way into what should be bastions of scientific medicine is acupuncture. Harvard, Stanford, Yale, M. D. Anderson, and many others, they've all fallen under the sway of the idea that somehow sticking thin little needles into points that bear no relationship to any known anatomic structure and that supposedly "unblock" the flow of some sort of "life energy" that can't be detected by any means that science has. Most recently, as I described, studies that seek to "prove that acupuncture works" have found their way into high quality, high impact journals whose editors should know better but apparently can't recognize that the evidence in the study doesn't actually show what the authors claim it shows. Even so, there are some journals that I didn't expect to see this sort of infiltration of quackademic medicine. Granted, I never expected it to show itself in one of the Nature journals, as it did in the study I just mentioned. I also never expected it to show up in that flagship of clinical journals, a journal that is one of the highest impact and most read medical journals that exists. I'm talking the New England Journal of Medicine, and, unfortunately, I'm also talking an unfortunately credulous article from Dr. Brian M. Berman, who is the founder of the Center for Integrative Medicine, University of Maryland School of Medicine and the holder of multiple NCCAM center grants, and other institutions, entitled Acupuncture for Chronic Low Back Pain.
The article appeared under the section of the NEJM known as Clinical Therapeutics. Articles published in this section begin:
...with a case vignette that includes a therapeutic recommendation. A discussion of the clinical problem and the mechanism of benefit of this form of therapy follows. Major clinical studies, the clinical use of this therapy, and potential adverse effects are reviewed. Relevant formal guidelines, if they exist, are presented. The article ends with the authors' clinical recommendations.
And this is the clinical vignette:
A 45-year-old construction worker with a 7-year history of intermittent low back pain is seen by his family physician. The pain has gradually increased over the past 4 months, despite pain medications, physical therapy, and two epidural corticosteroid injections. The pain is described as a dull ache in the lumbosacral area with episodic aching in the posterior aspect of both thighs; it worsens with prolonged standing and sitting. He is concerned about losing his job, while at the same time worried that continuing to work could cause further pain. The results of a neurologic examination and a straight-leg-raising test are normal. Magnetic resonance imaging (MRI) shows evidence of moderate degenerative disk disease at the L4-L5 and L5-S1 levels and a small midline disk herniation at L5-S1 without frank nerve impingement. The patient wonders whether acupuncture would be beneficial and asks for a referral to a licensed acupuncturist.
Berman et al begins with a simple discussion of low back pain, which, as he correctly notes, is an incredibly common clinical problem. He also points out that most of the really bad causes of low back pain (tumors, infection, or inflammatory disorders) are seen relatively infrequently in common practice. The most common cause of low back pain is the dreaded "I" or "N" word: idiopathic or nonspecific, both of which basically mean that we don't know what causes it. So far, fair enough. Berman et al even produce a fairly good discussion of the pathophysiology of low back pain, including the role of the central nervous system, behavioral elements, and musculoskeletal contributions, among others. Then, unfortunately, the authors go off the deep end:
Traditional Chinese medicine espouses an ancient physiological system (not based on Western scientific empiricism) in which health is seen as the result of harmony among bodily functions and between body and nature. Internal disharmony is believed to cause blockage of the body's vital energy, known as qi, which flows along 12 primary and 8 secondary meridians (Figure 1). Blockage of qi is thought to be manifested as tenderness on palpation. The insertion of acupuncture needles at specific points along the meridians is supposed to restore the proper flow of qi.
They even include a figure of acupuncture points
It was at this point that I wondered whether I was reading the NEJM or a quackademic medical journal such as the Journal of Alternative and Complementary Medicine. Here was an actual discussion of qi as though it might actually exist and as though meridians and qi were anything other than the result of prescientific concepts about how the body works and disease develops. One wonders if, for its next trick, the NEJM will publish Clinical Therapeutics articles touting the wonders of the humoral theory of disease and how the four humors must be balanced. Or maybe the miasma theory. That was a good one, and quite in accord with the modern day obsession with contamination and "detoxification."
My expectation to see greater woo appearing in the NEJM notwithstanding, as most CAM advocates do, Berman et al next try to justify acupuncture, starting with the belief that it works and then working backwards to cherry pick studies that they believe to support the hypothesis that acupuncture works for low back pain as anything other than a placebo effect. They begin with several inconclusive and conflicting animal studies, concluding by mentioning the study that I blogged so extensively about two months ago without noting that it didn't show what the authors thought it showed, nor did it demonstrate that adenosine mediates the effects of acupuncture. As I pointed out, what really irritated me about the adenosine study was that it was relatively interesting science but it was yoked into the service of trying to justify acupuncture with an animal model that had very little to do with acupuncture.
It gets even worse.
Next, Berman et al decide to delve into the clinical evidence for acupuncture:
A number of clinical trials have evaluated the efficacy of acupuncture for chronic low back pain. A meta-analysis in 2008, which involved a total of 6359 patients,44 showed that real acupuncture treatments were no more effective than sham acupuncture treatments. There was nevertheless evidence that both real acupuncture and sham acupuncture were more effective than no treatment and that acupuncture can be a useful supplement to other forms of conventional therapy for low back pain.
At least Berman's honest about this one in admitting that the meta-analysis showed that real acupuncture is no more effective than sham acupuncture, something that regular readers of this blog know. Then Berman tries to do what acupuncture apologists do every time they encounter studies that show that "true" acupuncture performs no better than the acupuncture control. Rather than simply admitting that acupuncture doesn't work and that acupuncture effects are placebo effects, they try to spin the results by pointing out that both sham and "real" acupuncture "work" and therefore are useful! In other words, they argue for placebo medicine without calling it placebo medicine. They then do it again for a German trial. Unfortunately for them, it's a study that I described in detail and explained why, as is the case with many acupuncture studies, the results didn't mean what the authors claimed they mean. Once again, Berman tries to represent the finding that sham acupuncture was just as effective as "real" acupuncture. In any randomized clinical trial of a conventional therapy, such a result would lead to the conclusion that the therapy doesn't work, but not in acupuncture. If both the placebo control and the treatment are indistinguishable from each other, then that means acupuncture must work.
The third study that Berman chooses is a so-called "pragmatic" trial. Basically, it's a mixed randomized trial with a non-randomized cohort. Let me quote one small passage from the trial that demonstrates why it is an utterly useless study:
In this study, neither providers nor patients were blinded to treatment. Therefore, a bias due to unblinding cannot be ruled out.
That's putting it mildly. Basically, the study is utterly worthless because it can't account for the rather large placebo effect that is common in intervention studies for back pain. In fact, it's fairly amazing that the peer reviewers at the NEJM let that pass, because the cherry picking of studies that went on in this particular review belongs at the National Cherry Festival, not in the pages of the NEJM, leading them to write:
Acupuncture is considered to be a form of alternative or complementary medicine, and as noted above, it has not been established to be superior to sham acupuncture for the relief of symptoms of low back pain.
In other words, acupuncture does not work. Even so:
However, since extensive clinical trials have suggested that acupuncture may be more effective than usual care, it is not unreasonable to consider acupuncture before or together with conventional treatments, such as physical therapy, pain medication, and exercise. Many pain specialists incorporate acupuncture into a multidisciplinary approach to the management of chronic low back pain.
In other words, even though acupuncture does not work, we should use it anyway because there are enough practitioners who believe it works and use it even though extensive clinical trials have shown that acupuncture is no better than sham acupuncture, and neither are better than placebo effects.
Even though:
As noted above, the most recent well-powered 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.34,65 These studies also seem to indicate that needles do not need to stimulate the traditionally identified acupuncture points or actually penetrate the skin to produce the anticipated effect.
In other words, acupuncture does not work.
Let's put it this way. Berman concedes that "true acupuncture" doesn't work any better than sham acupuncture for low back pain. He concedes that it doesn't matter where you stick the needles. It makes no difference; the effect is the same. He concedes that any perceived benefit from acupuncture in low back pain is due to nonspecific factors, in particular psychosocial factors, patient's expectations, and the attention paid to the patient by the acupuncturist. What, I ask, do we call such a treatment, a treatment that is no better than placebo control and whose efficacy depends on beliefs and expectations, attention from the provider, and contextual factors.
We call it a placebo.
It turns out that using placebo medicine is just what Berman is recommending to this patient in the clinical vignette that started out this post. He recommends a trial of 12 courses of acupuncture, since the patient is interested in it. Oddly enough, his recommendation that the patient go to a licensed acupuncturist with "adequate training" doesn't even jibe with his findings in his review article. After all, if it doesn't mater where you place the needles, then it doesn't matter if the acupuncturist is trained.
As Steve Novella points out, what Berman is doing in this article in the NEJM is the same thing that CAM advocates in general and acupuncture apologists in particular have a maddening tendency to do. They either cherry pick studies that appear to indicate that their favored woo works. When, as Berman et al were, they are forced to admit that well-designed studies with lots of patients show that their woo is no better than a valid placebo control, they then shift to embracing the placebo, to owning it, so to speak, all without actually calling it placebo.
What I find so disturbing about this NEJM article is not so much that Berman et al pulled these usual CAM tricks. I expect that. I see it all the time in CAM journals and sometimes in unsuspecting legitimate medical or scientific journals. What I find so disturbing about this NEJM article is that the peer reviewers did not spot the obvious CAM abuses of language designed to obscure the fact that acupuncture is no better than placebo. The editors of the NEJM should be ashamed of themselves. The peer reviewers who reviewed this article should be ashamed of themselves. Those of us who rely on the NEJM for evidence-based findings and assessments of various treatments should be afraid.
After all, if quackademic medicine can infiltrate the NEJM, there's nowhere it can't go.
ADDENDUM: Right on schedule, a CAM advocate misrepresents this review article as being "evidence" that "acupuncture works."
- Log in to post comments
Perhaps some letters to the editor are in order? Maybe a parody article submitted for publication that substitutes a "novel drug" and "placebo" for every mention of "acupuncture" and "sham acupuncture", respectively.
Perhaps the reason that acupuncture appears to give some short term relief from pain, is due to the state of partial meditation that patients enter, to overcome the terror/horror of having sharp needles jabbed into their bare skin. The effect would be akin to a low dose muscle relaxant, albeit very short lived, then as normal posture and life stresses take over again, so does the pain.
I just can't imagine paying someone to jab needles into my skin unless she was leatherclad and wielding a whip. I mean really, have you seen these acupuncturists? They're all old and venerable. No turn on at all.
@Todd W.: This sounds like a really good idea. Downloading the article right now... :-)
I get the feeling, if the author hadn't given the unscientific introduction into qi, the whole article might have been useful as a study of treating treatment-resistant lower back pain with placebo effect acupuncture. Actually, a simple replacement of "ancient" with "pre-scientific" might have solved the conflict, taking it from embracing unscientific nonsense to using an otherwise unavailable treatment method.
"Perhaps some letters to the editor are in order?"
Or a "comment on . . ." piece, submitted to the NEJM? It looks like between you, your friend at Science-Based Medicine and Steve Novella, all the necessary research has been done; it just needs to be translated into the proper dialect of Academic.
@Mark in #2 - or the pain of being pierced by needles could release endorphins, inducing a sense of well-being. If so, then the application of self-flagellation to treat lower back pain might be worthy of study. I'll send Orac a cat o' nine tails and he can write something up for NEJM.
Having come along with my wife when she had acupuncture done, you can add to that the fact that (if this place was at all typical) they do a great job to create a relaxing, meditative atmosphere.
I believe this goes some way to explaining why sham acupuncture seems to be so effective. I would be interested in seeing a study that compared:
1) Acupuncture/sham acupuncture (doesn't matter because we know they are equivalent)
2) A standard spa treatment, but characterized as a medical procedure
3) A pill-based placebo
My intuition is that 1 and 2 would at least slightly outperform 3. It could be the combination of a placebo with a relaxation regime that just genuinely makes ya feel good.
Of course that's just my intuition; it could be totally wrong. But if it were true, that could potentially open up an avenue of treatment similar to the one suggested in the NEJM, but actually being ethical since no one would be required to actually lie to patients -- for patients with lower back pain, for example, doctors could prescribe "therapeutic spa treatments".
If it worked, of course. Like I say, it's just my intuition, based on a) the fact that sham acupuncture seems to perform better than expected in many trials, and b) when I've seem someone undergo acupuncture, it looked really relaxing.
I agree with Mu that the way scientists should probably approach acupuncture is as an effective placebo medicine, and study it on that basis. Acupuncture and sham acupuncture both seem to offer more relief for lower back pain than no treatment at all, and why is that? Does this work better than placebo pills or massage, and if so, why?
There's always been a problem with placebo delivery under the standard health care modality; maybe training PA's or nurses in a certain amount of needle-sticking would be a beneficial (and billable) way to get the placebo effect.
@DRK
Yeah, but then you get into the whole ethical quandary of lying to your patient, rather than treating them with the respect and right to be fully informed about the treatment they're receiving.
Also, I agree with Mu @#5. The problem with this article is one of tone and of insufficient caveatting, not of its basic methods and conclusions.
If it were more forceful in dismissing the premise of acupuncture as a load of hooey, and it were more circumspect in regards to the ethical issues entailed in "consider[ing] acupuncture before or together with conventional treatments", it could have been a pretty good article.
I do think the fact that many patients could benefit from a placebo is an issue worth wrestling with. It should not cause mainstream medicine to just jettison all ethical concerns... but it's worth reflecting on, and brainstorming on how to take advantage of this without the inherent erosion of informed consent which it seems to entail.
The quacks have two big leg-ups on mainstream medicine at this point: More time spent with patients, and the ability to freely prescribe harmless placebos (the latter being a result of either self-deception or a lack of ethics). The former is fairly uncontroversial... but do not underestimate the value of the latter. Many people with generalized chronic complaints seem to derive a lot of subjective benefit from quack treatments. At what cost, of course, and this is why I am as vocal in opposing quackery as anybody. But if there were a way to capture that benefit without the ethical and safety issues....
Exactly, informed consent -- one of the pillars of modern medical ethics -- is incompatible with the administration of a "true" placebo.
Which is why I think SBM would be well-served at doing some soul-searching on how this problem might be overcome. If there a way to properly inform the patient without undermining the placebo effect?
If there were, that could put a lot of quacks out of business, no joke...
To paraphrase Voltaire: The secret of acupuncture is to distract the patient while nature heals itself.
In other words, even though acupuncture does not work, we should use it anyway because there are enough practitioners who believe it works...
And that attitude is what makes alt med such a non-starter for me. Sure it doesn't work as far as we know, sure there's no proof that it does anything other than make people feel good, but let's use it because it gives people warm fuzzies and alties get to make some cash while they're at it!
Even when Big Pharma resorts to disease mongering to push their products into blockbuster territory, they have to at least show that their pills work better than a placebo. Although if we're talking about Prozac or Zoloft, that's not necessarily the case...
@truthspeaker:
I tried acupuncture once, years ago. The needles were so incredibly thin that I could barely feel them when they went in*. I bet if I was really engrossed in something someone could sneak up on me and do Stealth Acupuncture on me without it distracting me.
* Except for the ears. That hurt. After experiencing that, I don't know how anyone could stand to get their ears pierced.
I think the point DRK was trying to make is that if there is evidence that it improves subjective outcomes, then that entails a worthwhile benefit to the patient even if the mechanism is baloney and the objective outcomes are unaffected.
I agree that DRK is far too blase about the ethical concerns involved, but I think your characterization of his/her comment is a little unfair.
So if the evidence indicates that acupuncture works by the placebo effect, why talk about meridians at all, other than a brief mention that they don't seem relevant to outcome? There are serious ethical quandaries about if and how a physician should recommend a placebo ("I don't believe this really works, but it might make you feel better anyway"?). But if we skip over that, and assume for the sake of argument that a placebo treatment is an acceptable therapy, then the question becomes, "what are the benefits and risks of the various placebo therapies?"
So if we approach acupuncture as a placebo-based therapy, it seems like acupuncture would need to show compelling evidence of benefit over other placebos, because it has the potential for risk--it is, after all, an invasive therapy, with risk of infection or damage if done improperly, so it requires a properly trained, and therefore presumably costly, therapist. So is acupuncture actually superior to, say, a sugar pill? How about massage? How about a nice, friendly conversation with a sympathetic nurse?
@Matthew Cline
That's it! I'm going to open up a new school to teach Stealth Acupuncture. It'll be part vitalistic pokery, part ancient Japanese martial arts. We can rid the world of disease forever! Muahahahahahaha! [cue thunder]
@trrll
Good points. Comparing acupuncture (sham or otherwise) vs. other, non-invasive relaxation and/or placebos would be good, with endpoints being efficacy and adverse events.
It should also be borne in mind, though, that well-controlled studies have shown that it doesn't even matter if the skin is broken (i.e., using, basically, toothpicks instead of needles). That sort of sham acupuncture had effects equivalent to "real" acupuncture. With that in mind, then no ethical argument can be made in support of "real" acupuncture.
Pay no attention to the actual science on acupuncture behind the curtain, I am the great and powerful Woo.
So the article is saying "acupuncture is just a placebo, but the effect is so strong we should utilize it anyway" by using any wording short of Pig Latin to avoid language which would make it obvious what they are really advocating. Don't you need to establish if treatment with placebo is ethical and acceptable before you decide which placebos to use?
Why is it important to find a properly licensed acupuncturist? To avoid contamination with dirty needles or to enhance the placebo strength via proper dramatic effect?
The article is more grist for the woo-proselytizers' mills, eternally grinding reality into the fanciful : usually, they'll cherry pick an appealing point and free associate onwards to their desired end- here the author has already done the creative heavy lifting for them. It's bizarrely ironic to hear woo-meisters talk about "peer review", "RCT", "statistical significance","blinding","the placebo effect", dressing their fictions in the trappings of SBM, while habitually trashing SBM, research, and practicioners. I can just imagine how Adams, Null, Mercola, and lesser poseurs will spin this study : "It's in the *NEJM*!","It *proves* our medicine is better than yours!", "*Qi* is real !"..... ad nauseum. (( Speaking of woo: I haven't seen any disturbing sidebar ads here lately))
@ Todd W. (re your proposed school-) you'll probably make major money plus you'll get to wear cool clothes.
@Denice Walter
Not only can I charge a premium to train my needle-armed ninja, I'll also be able to walk up to people and say, "You've just been acupunctured. That'll be $1,000."
@Todd:
But can you get your friends free cable?
What I would personally like to find is more information on is needling directly into muscle tissue.
I spent over a year swallowing ibuprofen and Robaxacin like candy for the spasming knots behind my kidneys. I needed a cane to get into and out of a chair. Even rolling over in bed was agony. After a few months of not-so-helpful treatment, my physiotherapist tried needling the spasming muscles directly. I don't know what the actual effect was, but the apparent net result was they felt relaxed and I had pain-free full range o un-stiff movement. I flowed off the table and felt like I could do a cartwheel, not like I would fall over if I bent a little bit sideways. The effect lasted several hours, or until I spent too much time sitting still.
Exercise and stretching seem to have healed those muscles, but at the time, the needling was very interesting if confusing pain relief. I'd love to see if this has been studied outside of proprietary training clinics (the only info I'd found, and struck me as very similar to "Irlen" woo in it's lack of disclosure).
@Dave
One wonders if the same effect may have been achieved through some manner of massage of the offending muscle.
I'll be opening a chain of homeopathic acupuncture clinics. The needles go through a top secret dilution process and are invisible to the naked eye (and electron microscope). You won't even feel them go in!
I get the feeling that you decided immediately that acupuncture is invalid. I believed that in the 1960's, but around 1970 (I think) the Scientific American reported a study that found it to work. Since then many other researchers have reported that it works.
If someone who doesn't know how to do it fails to achieve results, that may be technique rather than the practice itself.
Please note, when I say "it is reported to work" this is reporting of observations, not of theory. I think that the standard theory behind it is silly, but if it works, it works.
N.B.: We don't *know* that acupuncture is equivalent to sham acupuncture. That would require either double-blind studies (and how to manage THAT), or long term studies with lots of experimental subjects and careful record keeping. It *might* be resolvable in 20 years. My personal suspicion is that it's a form of hypnosis, and that it does work. I have anecdotal evidence from acquaintances that it works. (But presumably the failures quickly gave up and so no reason to talk about it.)
P.S.: Many prescribed drugs have benefits that are only marginally superior to placebos, so there's no reason to disparage something relatively harmless merely because it produces it's results via placebo. (The drugs often have very worrying side effects on at least some patients.)
That's it! I'm going to open up a new school to teach Stealth Acupuncture. It'll be part vitalistic pokery, part ancient Japanese martial arts.
"Stealth" sounds too militaristic for your most likely customers. How about "homeopathic accupuncture?"
The placebo effect only works for SBMer if ORAC performs it.
CharlesHi--
The point is that there are good, blinded studies comparing "real" acupuncture to sham acupuncture. In particular, the patients didn't know what they were getting. If the blinding was imperfect on the practitioner side, that would mean that acupuncturists who knew that they were doing sham acupuncture were as convincing to the patients as those who thought they were doing something real.
The point of double-blinding is to avoid skewing things in favor of the "real" treatment, either because the doctor is more confident or because s/he (perhaps unconsciously) selects the healthier/better patients for the real treatment and the sicker ones for the placebo. If "real" acupuncture can't beat sham acupuncture even without double-blinding, that is not evidence in favor of acupuncture.
Consider a pill or injection as an analogy. It would not give me greater confidence in a medicine to know that, even when the doctors knew who was getting the placebo, the real pill was no better than a sugar pill, the real chemical no better than the same volume of simple saline.
@CharlesHi
Double-blind studies have been done. In order to accomplish this, the real and fake needles are enclosed in a sheath or are telescoping, so the practitioner cannot see whether there is penetration or not, and the sensation is still the same for the subject. Here's a study on using sham needles.
It is very sad how quickly alternative quack care is becoming accepted as credible. My healthcare policy (Great West Life, a Canadian company) will cover 80 % of chiropractic expenses. My friend has healthcare via the same (Great West Life) but a more comprehensive package. Hers covers $300/year chiropractic, $500/year naturpathic, $500/year accupuncture, $300/year massage, $250/year applied kiniseology and her employer has allowances for vitamins.
I also noted that our mall now has a company offering, for a large fee, "rejeuvenating oxygen chamber, great stress relief!" :(
At least both policies state that they will NOT cover homeopathic preparations
You should read more of what Orac's written on the subject. As he's repeatedly described, he used to think that there might be something to it - not meridians and qi, but some real effect of the needles. The DATA convinced him otherwise.
It's not that it fails to achieve results. It's that twirling toothpicks on the skin achieves the same results as the "best" acupuncturists.
"It is reported to work" is not a meaningful or useful observation. What's relevant is whether it works better than placebo. And the weight of the evidence firmly indicates that it does not.
You mean like the double-blind studies that have been repeatedly done and found that it indeed doesn't work any better than placebo?
Also consider the fact that being pampered in a nice, relaxing atmosphere genuinely does make people feel better. Or, in other words, even your anecdotes doesn't provide a jot of evidence that the needles are responsible.
This is more a reason to object to those drugs than to approve of acupuncture.
So happy I let my subscription lapse
Shoot, acupuncture can even be done at a person's home! In fact, a vet I know that is getting into acupuncture claims that it works better at home than it does in the clinic. The ramifications of this are pretty striking, I thought.
I think people should go back and read jay.sweet's post. I think there's real potential in the hypothesis he proposes, and people should at least discuss it.
According to studies cited by Ben Goldacre at "Bad Science", different placebos have different levels of effectiveness. A sham injection is more effective than a sham pill; sham surgery is more effective than a sham injection. My hypothesis is that acupuncture should fall somewhere between a sham injection and sham surgery, though that will depend on the belief of the patient.
So... if acupuncture is Only A Placebo, and if no treatment should be used for a condition unless it has been proven in clinical trials to be better than a placebo ... may I presume that you will no longer approve the use of analgesic drugs for conditions for which they have been shown to provide less relief than acupuncture?
MOB - note that it also depends heavily on the ailment being treated. The placebo effect is much more significant in more subjective ailments, like back pain, and not so big on more objective problems, like appendicitis.
Which conditions would that be?
Also: Cost of acupuncture per session $60 - $120
Cost of generic acetaminophen $2.99/100 (estimate)
jane:
Please name those drugs and the studies that show they do not work better than placebo.
Also, do they have the bonus of causing site infections like acupuncture done without proper hygiene protocols:
I once watched a video of Chinese veterinarians performing a small bowel resection on a horse anesthetized only with acupuncture. The horse was awake and showed no sign of distress. Discounting the possibility that the video was a sham - that the horse had surreptitiously been anesthetized pharmacologically - it was a pretty convincing demonstration of the analgesic effect of acupuncture. Do placebos work on horses?
@darwinsdog
The effect observed in animals is largely, I think, due to changes in the humans. The type of attention they give the animal, as well as the observations they make about the animal, may change.
As to Chinese "evidence", color me skeptical. There have been instances where western researchers have gone to China to observe acupuncture as an anesthetic. Only problem was that the subjects were either drugged surreptitiously or commanded to fake no pain. One subject that was observed was saying, in Chinese "Pain, pain, pain," before getting a stern warning from the person overseeing the observation. This was discussed on the Skeptic's Guide to the Universe. Can't recall the episode. So, my guess is that the horse was probably drugged. Can't say for certain, though.
Pablo,
That makes sense. Symptoms like drowsiness, pain, "jitters", nausea, and such should be more amenable to the placebo effect than, say, bleeding or some of the more advanced forms of death.
BTW - I've long believed that there really is no placebo effect and a double blind study comparing the effect of a placebo to a sugar pill would find no difference.
Acupuncture got a huge boost in China just before it was discovered by the west as part of their "a doctor for every village" campaign. Since it was impossible for Mao to actually train the millions of new doctors needed and equip the necessary medical facilities, they issues acupuncture kits to barely trained medics. I guess if my choices are "nothing" or "acupuncture", I try the latter 100% of the time.
What Todd W. said.
Yes, the placebo effect works in animals (although in that case it is the owner's/handler's expectations rather than the animal's that matters, of course), which by the way totally blew my mind when I heard it, but I have an anecdote(*) about that which really convinced me.
And no, I would not believe video evidence of acupuncture efficacy that came from China. Ever.
(*) Okay, anecdote time. My wife became concerned about one of our dogs, I can't remember exactly why, and when she approached to check the dog out, the dog put her tail between her legs, starting shaking, showing severe signs of distress. I was worried we were going to have to call the vet. Although my wife was making no overt signals that I could see, it suddenly dawned on me that this dog is pretty sensitive. So I told my wife, "Go hide behind the door so she can't see you" and then I called the dog. Perked right up, nothing the matter at all.
What's so weird about this story -- and you really had to be there to fully appreciate this -- is that it seemed to me like my wife was acting totally normal. Yeah she was expressing concern about the dog to me, but her tone seemed completely even and there was no body language that I could discern that was at all unusual. And yet apparently there were enough vocal/physical cues going on that the dog picked up that her owner thought there must be something wrong with her.
After seeing that, I never doubted the placebo effect in animals again. If anything, it might be more intense than in humans, because there is a positive feedback loop (the more my wife communicates concern, the more visibly worried the dog becomes, which means my wife becomes more concerned, etc.)
I have begun to doubt that the placebo effect exists at all in terms of objective outcomes. The idea that a good attitude improves cancer outcomes, for example, has recently taken a heavy blow.
If you believe you are experiencing less pain, you are experiencing less pain. If you believe you are happier, you are happier.
If you believe you didn't actually get hit by that bus... not so much.
jay.sweet
A second type of "placebo effect" in animals is in the human interpretation of their response. Since pain in animals generally requires human interpretation, that is also susceptible to a placebo effect. "Oh look, he had acupuncture and appears to me to be walking better now." With an emphasis "appears to me"
Vets are taught to assess pain in animals, and generally do a MUCH better job than owners do, but they are still susceptible to bias, and that is why even in animal studies, there needs to be proper blinding.
Thanks, folks, for your responses. I don't want to be interpreted as defending acupuncture in any way, but that video of surgery being performed on that horse reportedly anesthetized only with acupuncture seemed rather convincing. The film showed the entire procedure, from the insertion of the (numerous, large) needles in various parts of the horse's anatomy to the suturing of the body wall & skin. The horse appeared docile and calm throughout. I don't discount the possibility of fraud although I have no reason to suspect it, either. It seemed to me that some real clinical anesthetic/analgesic effect was being achieved by the application of those needles. Although the video spoke of "Qi," I presumed at the time that this effect was due to stimulation of endorphin secretion. My only point is that acupuncture may not quite belong to the same category of quackery as, say, homeopathy.
This has to do with pain and not with any disease. Sticking needles into someone precisely or randomly cannot provide pain relief. The study showed that either acupuncture or wildly-off-the-mark-puncture was better at relieving pain than "western medicine". Either those being pricked are "lying" in agony or they are no longer in pain. If they are not lying then their pain is to some degree psychosomatic. If this is so then the perceived efficacy of the treatment correlates with the confidence in it of the patient being treated.
Why is there apparently more confidence in a bogus treatment than in one which is evidence-based?
One theory I've heard. Painkillers such as pills are supposed to work. If you feel any pain after taking a pain killer, it's failed. On the other hand, being stuck with needles to relieve pain is counterintuitive. If you feel any relief at all, it's amazing! Thus you're more inclined to feel let down by the pill and more forgiving of what you do feel with acupuncture.
It's really quite astonishing how many commenters here feel it is somehow scientific or accurate to make bizarre statements while attacking "woo".
1) Acupuncture is not anything like "being stuck with needles", ie, it is nothing like an intramuscular injection. The needles do not even draw blood.
2) The idea that only psychosomatic pain can be resolved by placebo is nonsense. Read up on this. Post-surgical pain can be reduced with placebo. Asthma symptoms can be reduced with placebo. Etc.
What this means is that placebo is a PHYSIOLOGICAL phenomenon. And if you dispute that, then realize that YOU are the one introducing "woo" into the conversation. Are placebos made of magic?
Well, no, you might say, that's why I think the pain was psychosomatic! Ok, well, what the hell do you think "psychosomatic" pain is? Something the psychosomatic fairies bring to us? All the same physiological mechanisms are in action -- inflammation, stress hormones, and so on. These can happen if you're punched in the face, or are simply terrified you will be punched in the face. That doesn't the fear-induced reactions less physical.
3) If a placebo reduces human suffering, why in the world object to it?
Sonya:
Because modern medical ethics is against lying to patients.
Also, do you think that needles should be puncturing skin without using basic hygiene protocols like alcohol wipes, washing hands and gloves?
Brucy: "Also: Cost of acupuncture per session $60 - $120
Cost of generic acetaminophen $2.99/100 (estimate)"
Cost of original liver- PRICELESS!
@Sonya #53:
What's your take on the needles supposedly redirecting the Qi?
Sonya,
I used the phrase "being stuck with needles" but in no way did I say it was like an intramuscular injection (though I did suggest that in terms of placebo effectiveness it might be somewhere between a sham injection and sham surgery). But as I understand acupuncture, they insert needles into your skin - how that is not "being stuck with needles" is beyond me.
@Todd W.
Stretching, exercise, traction and massage were all tried in conjunction with the needling. (single-use sterile wrapped). No mumbo-jumbo was supplied just "Lets try thisâ¦"
I think psychosomatic pain is very real to the sufferer. It is also possible that the pain is real, but the relief is in the mind. When it comes to treatment, how does the doctor decide whether or not to charge for a sugar pill? You cannot guarantee the placebo effect. There is no physical cause and effect here. If the patient has not found relief and sues for malpractice, what is the doctor's defense?
Also there is a slippery slope. First psychosomatic pain, then cancer, autism, and many other diseases and conditions will be on the list of things treatable by a magic pill.
Academic medicine is completely woo-contaminated. Our medical school promotes it regularly (especially acupuncture) and rarely says anything close to "it doesn't work." Working at acupuncture clinics and going to China to learn about qi counts for credit for an MD. I probably got a lower grade because I questioned an attending when he suggested acupuncture for treatment of heartburn. It's especially upsetting because most med students take whatever their attendings say as fact.
On the horse video, another possibility is that the needles used for the acupuncture were laced with some sort of anesthetic or other drug. It is impossible to tell, just from a video, if everything was on the up and up.
This reminds me of something I was thinking in the original post. This part...
OK, either Orac cut out a citation or two, or I have to ask, how the bloody hell did this statement get past a reviewer? You can't just assert nonsense like this in a scientific paper without referring to some source for that claim. It is meaningless gobblygook.
The training includes how to look like an acupuncturist, how to talk like an acupuncturist, and how to act like an acupuncturist. The vocabulary and demeanor are essential to creating the placebo effect.
The bigger the mask, the better the witch doctor.
Well, you don't have to worry, unless you're a drunk in the habit of taking overdoses.
Just sayin'.
What kind of training do I need to set up shop as a sham accupuncturist?
First, buy a box of toothpicks.
To explain why toothpicks:
@Sonya- All pain is to some extent psychosomatic. Perception of pain from precisely the same surgery/injury/disease varies wildly from person to person.
I treat pain daily. The depressed middle aged back paineurs take way more pain medication than do the 82 year old crippling osteoarthritis patients. Why? Is their bulging disc more painful than disfiguring erosive arthritis? It shouldn't be.
I've been depressed. It was physically painful. Upon recovery, I realised that I was no longer in pain simply standing in line at the grocery store. I'd known all along there was "nothing" wrong with me physically. That sort of pain is really hard to treat without treating the depression.
Papercuts? Hurt like the dickens. But I'm sure everyone's played soccer/hockey/football and scraped all the skin off some body part and not noticed it for ages.
Pain really is all in one's head. To say "only psychosomatic pain" is to deny the experience of pain for he whole person. If I sound like a woomeister, I'm not.
I see nothing but placebo in the NEJM article, and it boggles my mind that they are allowing the report to be spun. I regularly tell my patients not to waste their money on non-evidence-based fakery.
due to the state of partial meditation that patients enter, to overcome the terror/horror of having sharp needles jabbed into their bare skin
Not quite. I got interested in acupucture woo a couple years ago and bought a box of needles on Ebay (they're cheap!) and experimented with inserting them into various parts of my anatomy. What I discovered was that it's very easy to tell when you're about to "drill" where there's a nerve - the natural tendency to flinch is a dead giveaway. When you drill where there isn't a nerve it's pretty easy to spin the needle quite deep (1/2" or more!) with no sensation at all, other than a weird twingeing feeling. Removing the needle results in a tiny red dot and no pain - the needles are fairly fine.
What appears to be going on is that the acupuncturist quack learns pretty quickly how to tell if they're drilling where there is a nerve, and to move the needle (after all it makes no difference where the needle is!) until it's someplace that won't cause pain. It's pretty simple. After I had put about 4 needles into my hand I found I was able to very quickly identify less enervated spots.
Need I mention that my amateur attempt at acupuncture didn't work? Other than enlarging my penis by 12", I mean.
Please know that I am not all all suggesting lying to patients! Mostly, I am thinking that the placebo phenomenon is 1) certainly real and 2) needs to be better understood; the billable hours comment was in jest, really. Shouldn't do jesting on blog comments -no one but me can see my cynical posting smirk. The traditional explanation for the efficacy of acupuncture is clearly total woo -- given that it was formulated probably before Galen's humors, this is not a surprise.
I guess really, the question I would have is, how could the placebo effect be ethically administered to patients, given that it might be a real, clinically valuable treatment when appropriate? And what would be the most effective placebo in that case? And is this a valuable question to pursue?
As to Chinese "evidence", color me skeptical.
Wouldn't lifespans in China be significantly longer than in the rest of the world, if acupuncture worked? And, of course, they aren't - lifespans didn't start to jump until Pasteur. "Traditional Chinese Medicine" has had centuries to produce solid, reliable, beneficial results, and it hasn't. Because it doesn't work.
This article will be the most frequently cited as ever. All the CAM specialist will mention in their papers and many who will argue against. NEJM lost 5 impact factors during the last 2 years. This paper will increase its IFs a lot.
I agree with ToddW's original thought - a letter to the editor from Orac and Dr Novella would be the most effective reply to this paper.
In order to be ethical, the patient would have to be explicitly informed that it has no specific effects. "The needles don't actually do anything, but the ritual surrounding the procedure is generally relaxing, which often makes people feel better."
Also, the risks of the placebo should be minimized. Which means that acupuncture is right out - can't justify breaking the skin (with corresponding risk of infection) when something like reiki could be selected instead.
Like jay.sweet said, it would be very appropriate to compare the classic woo treatments vs. getting a massage at a nice spa. If it's even close, then IMO doctors should feel free to recommend the massage to patients with a description like "it promotes relaxation and stress relief, which often makes patients with a wide variety of conditions feel better." Completely honest and above-board, and likely carries most/all of the benefits.
DRK--
One way to use the placebo effect is to combine it with treatments that are effective in other ways. For example, given medication M (say, a painkiller), the same dosage of the same chemical may be more effective if given in a small, bright red pill than in a small, pale green one. This has potential for confusion, given how many pills look alike already, but changing the color of a pill or the amount of inactive material added so the patient has something large enough to pick up is not an ethical problem.
A trickier one is that a patient who is told that their pill costs $10/dose is likely to have better results than one told that the exact same pill costs 10 cents/dose. However, there are obvious ethical problems in raising the price of a medicine in order to get this effect.
Some clinical trials in which acupuncture compared favorably to a pharma drug:
Hackett et al. The Practitioner 1988;232:163-4. For acute low back pain, improved mobility better than paracetamol.
Kittang et al. Tidsskr Nor Laegeforen 2001;121:1207-10. For acute low back pain, comparable to naproxen except that patients had fewer new episodes of pain and used fewer drugs, with side effect rates zero vs. about 50% (especially GI).
Sangdee et al. BMC Complement Altern Med 2002;2:3. For knee osteoarthritis, electroacupuncture superior to diclofenac.
Zhi. Zhongguo Zhen Jiu 2007;27:18-21. For dysmenorrhea, superior to indomethacin.
Resim et al. 2005;33:285-90. For relieving pain during lithotripsy, electroacupuncture comparable in most parameters to tramadol and midazolam, with average pain on a visual analog scale of 5.0 vs. 8.0 and without the side effects of the drugs. (I recall that there are similar studies in which the EA group has modestly more pain than the drug group but also avoids more severe side effects).
Waler et al. J Clin Oncol 28:634-40. For vasomotor symptoms in breast cancer patients, equivalent to venlafaxine with zero side effects vs. high rates of side effects.
Gan et al. Anesth Analg 2004;99:1070-5. For prevention of postoperative nausea and vomiting, electroacupuncture comparable to ondansetron (superior in reducing incidence and severity of nausea and postsurgery pain).
Allais et al. Headache 2002;42:855-861. For prophylaxis of migraine without aura, superior to flunarizine in reducing frequency of attacks at two and four months, equal at six months; acupuncture was the only treatment to significantly reduce pain intensity and had significantly fewer side effects.
Lee et al. J Urol 1992;147:16-8. For renal colic, relieved pain comparably to Avafortan injection but with far faster analgesic onset and a side effect rate of zero vs. 44%; patients receiving Avafortan were more likely to have paralytic ileus.
Long-term consumption of various OTC analgesic drugs, by the way, has been reported to increase the risk not only of GI problems (and liver damage for Tylenol) but of heart problems, kidney damage, and premature deafness. You can prate about the hypothetical deadly risks of a skin infection all you like; it's clear that the burden of death and disability from casual use of potent drugs is much greater. Certainly, I have never heard of a single head-to-head comparison of a drug and acupuncture in which the acupuncture group had more side effects.
The upshot for me is that if I had lower back pain or knee arthritis, rather than embark on long-term drug treatment that would be very likely to cause me serious side effects (since my stomach is sensitive to analgesics), I would most certainly try acupuncture first. If its Only Placebo effect is equal to the placebo plus mechanist effects of the drug, then as far as my interests are concerned, "it works." Asserting that it "doesn't work," even in the context of a study in which it was statistically superior to a drug that "does work," is not a scientific judgement; it's a philosophical opinion, and as such, "science" cannot demand that we all embrace it. Cheers.
An effect not limited to medication. Food and wine generally taste better if the person tasting them believes them to be more expensive, for instance.
Acupuncture is possibly the least harmful woo out there.
-It's relaxing, very much so: a good thing, surely.
-It "treats" non-specific pain of all sorts.
-It is so much related to general well-being that an acupuncunist might be more likely than other wooists to refer a real problem to a doctor (bearing in mind, of course, that most/all patients have already seen a doctor).
-The placebo effect is indeed strong with this one. That's a good thing for the patient.
Those are a few reasons why I didn't argue with the two acupuncturists I met at a barcue recently - that and I didn't want to get thrown out like the time when I argued bitterly with the homeopathists at a friends' house. But then homeopathy is malign nonsense whereas acupunture is....begnignish. Or at least, begnignishy.
But no, it's utter rubbish, and this study has no place in any proper scientific journal.
And yet, and yet - if CAM has any use whatsoever, debatable I know, perhaps a doctor suggesting that even though he's sceptical, it helps some patients (at a price of course...placebos don't come cheap!) and a patient thinking he's getting relief, that may be something for the poor thing.
Sigh, yes I know: acupuncture's pure shite, serves no purpose in any healthcare system, and promotes stupid magical thinking.
But be honest - for chronic, long-term, back pain, say, wouldn't you be tempted to give it a shot?
It'll cost you, mind.
Oh god, I just supported acupunture.
I hang my head in shame.
My regrets to the two people who asked for details on the conditions for which acupuncture methods have tested favorably against pharma drugs in head-to-head clinical trials. I wrote up a message yesterday morning citing nine such trials (mostly for various painful conditions, such as lower back and knee pain and migraine prophylaxis), which consistently reported acupuncture to be virtually free of side effects, in notable contrast to the drug arms. The comment was automatically directed to moderation; I assumed this was simply because it was long, since it included no links, but alas, Orac did not choose to release it. So if you want to know what actual scientific research has been done (assuming you see this message), you will have to get on PubMed yourself and look up whatever conditions interest you. Cheers.
jane,
Don't assume Orac is oppressing your voice if you don't see your comment. The spam trap here is weird. I'm probably in it right now in fact.
It's not enough to quote studies that show acupuncture works. That's just cherry-picking and can be a means to make anything look good.
You have to look at *all* the evidence out there regarding acupuncture. When you do that, you see that a few poorly controlled studies are positive while larger, better controlled studies are negative. That's the pattern we get for stuff that simply doesn't work.
Considering the rather lax moderation on this blog, I would suggest that you try again, jane.
Though, I just remembered there are certain keywords. So to avoid that, jane, just post the PubMed ID number for the studies.
It was probably the length of the message (several paragraphs' worth) that triggered automatic moderation, but only Orac knows why he chose not to release it. From the abstract printouts I have on file, I get the numbers 1402745, 11914160, 15971087, 17378197, 12390610, 1729516, 20038728, and 15385352. I've also seen a reference to a second study on low back pain (Hackett et al. The Practitioner 1988;232:163-4) that is not listed in PubMed.
Now, none of these are large enough to be conclusive (ca. 200 patients maximum), but if the only relevant research on a particular condition has found acupuncture to be as good as or better than a drug - while enabling the patient to dodge sometimes a >50% drug side effect rate - I'm not inclined to simply presume that "the drug must be better, never mind what the research says" simply because we have a better idea of HOW the drug relieves pain.
Those are a few reasons why I didn't argue with the two acupuncturists I met at a barcue recently - that and I didn't want to get thrown out like the time when I argued bitterly with the homeopathists at a friends' house.
You go to someone's house as a guest and feel you need to engage in debates in the name of your science? I wonder how many SBMer's have gotten into arguments over their "view of the world" and tried to set someone "straight" at a social function.
I wonder because you seem to be somewhat wishy washy open to the placebo effect where others have dismissed it as "not counting" because it is scientifically unexplainable.
AnthonyK : "...and a patient thinking he's getting relief, that may be something for the poor thing."
Don't you know that pain is in the brain? It's a subjective experience. If someone "thinks he's getting (pain) relief" he ACTUALLY IS getting pain relief. Good grief. Get off the fence man.
And I think it's a good strategy for the all knowing empiricists to go to parties and set stupid people straight in scientific factual ways. And you wonder why SBMers are still a cult and it's the rest of the world that's stupid?
Two words: Emotional Intelligence. You don't have it.
More accurately, we know both how the drug relives pain, and that acupuncture CANNOT relieve pain by anything even vaguely related to the rationale for it.
I couldn't care less about "the" rationale for acupuncture, or more accurately, anyone's rationale. We can stipulate that when the originators of TCM thought acupuncture worked by "adjusting qi," they were wrong. Likewise, when ancient Greek doctors used opium to relieve pain, they were just as wrong to think it worked by affecting "humors." Yet it did work, and none of us would refuse to use codeine for pain because Galen believed in humors. Most of us are pretty selective in deciding which cultures' knowledge should be rejected en masse because it historically included errors.
jane, I looked at them and suggest you get a copy of this book. It has a good section on how to look at these studies and evaluate them.
I will assume that the first PMID contains a typo, because it was a comment titled "Determining total method level of detection and level of quantitation for breath alcohol analysis programs."
I thought you meant analgesics like ibuprofen. I have never heard of the specialty drugs like diclofenac and venlafaxine (the latter not really being an analgesic).
These are all on electroacupuncture, which is not really needling:
11914160:
BMC Complement Altern Med. 2002 Mar 21;2:3.
Electroacupuncture versus diclofenac in symptomatic treatment of osteoarthritis of the knee: a randomized controlled trial.
Sangdee C, Teekachunhatean S, Sananpanich K, Sugandhavesa N, Chiewchantanakit S, Pojchamarnwiputh S, Jayasvasti S.
Department of Pharmacology, Faculty of Medicine, Chiang Mai University, Thailand. csangdee@mail.med.cmu.ac.th
15971087:
Urol Res. 2005 Aug;33(4):285-90. Epub 2005 Jun 22.
Effectiveness of electro-acupuncture compared to sedo-analgesics in relieving pain during shockwave lithotripsy.
Resim S, Gumusalan Y, Ekerbicer HC, Sahin MA, Sahinkanat T.
Department of Urology, KSU Medical School, Kahramanmaras, Turkey.
Zhongguo Zhen Jiu. 2007 Jan;27(1):18-21.
[Randomized controlled study on superficial needling for treatment of primary dysmenorrhea]
[Article in Chinese]
Zhi LX.
Eye-Ear-Nose-Throat Hospital of Jiaozuo Coal Corporation, Henan 454000, China.
15385352:
Anesth Analg. 2004 Oct;99(4):1070-5, table of contents.
A randomized controlled comparison of electro-acupoint stimulation or ondansetron versus placebo for the prevention of postoperative nausea and vomiting.
Gan TJ, Jiao KR, Zenn M, Georgiade G.
Department of Anesthesiology, Duke University Medical Center, Durham, NC 27710, USA
This was interesting (and why you need to read the book listed above), because of the title of the comment:
20038728:
J Clin Oncol. 2010 Feb 1;28(4):634-40. Epub 2009 Dec 28.
Acupuncture versus venlafaxine for the management of vasomotor symptoms in patients with hormone receptor-positive breast cancer: a randomized controlled trial.
Walker EM, Rodriguez AI, Kohn B, Ball RM, Pegg J, Pocock JR, Nunez R, Peterson E, Jakary S, Levine RA.
Department of Radiation Oncology, Henry Ford Hospital, Detroit, MI 48202, USA
The comment on this is:
J Clin Oncol. 2010 Apr 20;28(12):1979-81. Epub 2010 Mar 22.
Conventional and complementary therapies: a tale of two research standards?
Levine MN.
I also looked through the abstracts of the articles you mentioned. VERY poor methodology, in all cases.
1402745: This article is "Determining total method level of detection and level of quantitation for breath alcohol analysis programs." Typo?
11914160: Only single-blind, with four groups of placebo, diclofenac, electroacupuncture, and diclofenac+EA. Accordingly it must be considered to not have a meaningful placebo control - no single placebo can be appropriate for both treatments, and there is no mention that the single-treatment groups received a placebo of the other treatment. Also, any success of EA does not validate acupuncture since electrical stimulation has known physiological effects.
15971087: Unblinded, and EA.
17378197: Unblinded with no placebo.
12390610: Unblinded with no placebo.
1729516: Unblinded with no placebo.
20038728: Unblinded with no placebo.
15385352: Unblinded, but at least there was a proper control (the ONLY one on the list which did). EA.
Not a single double-blind placebo-controlled study; when the weight of the evidence points firmly to the procedure in question being pure placebo, these are crucial. Indeed, the only one with a meaningful placebo control was completely unblinded and the only one to even be single-blind had no meaningful placebo.
Even if they were huge, they wouldn't provide any substantial support for the efficacy of acupuncture. If they were large and there was no prior implausibility they would be good as hypothesis generators. That's pretty much it.
I realize both Chris and Scott touched on this, but it bears repeating. Electro-acupuncture is not acupuncture. It is electro-stimulation with a veneer of Eastern mysticism. The effects they are seeing are, most likely, due to the electro-stimulation.
The proper way to test acupuncture is to use sham needles that neither the patient nor the practitioner can tell from real needles. This has been done (someone linked to a study earlier, I believe), and the results showed that real needles do not differ in their effects from placebo needles (i.e., ones that do not break the skin). Other studies have also shown that the placement of the needles does not matter, either. The effects are the same whether the needles are placed on "meridians" or not, whether they are placed on the "proper" meridians for the condition being treated or placed elsewhere.
So, proper placement, improper placement, breaking the skin, not breaking the skin...they all have the same results, meaning that it is more than likely that any effect seen is placebo, and nothing more.
Chris - Sorry for the typo; the first one is 11407745 (a Norwegian paper comparing acupuncture to naproxen for back pain). Diclofenac is an NSAID; it's not available OTC in the U.S. but is prescribed here. As for the venlafaxine, that was used in a trial for reduction of hot flashes in women with breast cancer; of course, an analgesic would not be a suitable comparator in that case.
We will have to disagree about whether electroacupuncture studies should be mentioned. The standard dogma here is that all types of skin stimulation are equal placebos, including Poking Yourself With Toothpicks - despite the fact that many or most studies of real vs. sham acupuncture find otherwise - and I see no reason to assume that electroacupuncture would less stimulate the skin or have less placebo activity than the legendary Toothpick.
I've looked at Snake Oil Science; the author seems to me to have a definite bias that prevents me from uncritically accepting his judgement of published research or potential future research.
As for the comment you link to, the writer's criticisms of the hot flash study imply the philosophical position that placebo effects should be avoided rather than benefited from (demanding sham acupuncture be included with the medication for the sake of "double blinding") and simultaneously privilege his own opinion regarding mechanism (because if both acupuncture and sham acupuncture actually caused useful nerve stimulation, "sham" acupuncture would be a totally inappropriate placebo). To others, a head-to-head comparison between methods as they are practiced in real life is entirely reasonable; if acupuncture gave relief equal to a drug without drug side effects solely because of its enormously superior placebo activity, that would be just fine with many patients. You can never answer that question if you demand that acupuncture be always and only compared to groups receiving (possibly bioactive) "sham acupuncture." Some of the other questions approach nitpicking (no study is written up perfectly, and many studies end up enrolling fewer patients than they would have liked; he opines without evidence that the study was "likely" underpowered to detect differences between groups even though the one significant difference reported favored acupuncture!). Of course one small study is never the last word - but so long as these are the best data we've got, if I were a patient looking at that 18[incidences]/25 side effect rate for venlafaxine, I'd sure start by trying acupuncture.
Toothpicks also work!
The point is that, even if needles do anything, acupuncture cannot be meaningful. Culture has nothing at all to do with it.
Actually, jane, you need to read that book more closely. He actually did not have much of a bias, at first. But he is critical of the studies that come from certain countries, like where several of the studies you cited were from.
As a person who is one of the 10% who cannot tolerate narcotic pain killers, I would love it if magic needles would work. But they don't. Or they work as well as toothpicks.
Ibuprofen does work. And fortunately I was also given anti-nausea medication along with the analgesic when I had a colonoscopy, which also worked.
To reiterate Todd's point, I find it anywhere from amusing to interesting to annoying that this thread has gone into discussing the efficacy of acupuncture. The actual article that is discussed in the post in fact stipulates that acupuncture doesn't really work, and is nothing more than placebo. So for Todd's last paragraph,
as Kevin Bacon would say, "These are the facts, and they are not disputed." At least, not by the pro-acupuncture authors of the article in the NEJM. Therefore, arguments that acupuncture IS actually effective, particularly compared to placebo, is changing the topic and getting away from the claims of the paper (topic drift always occurs, of course) and turning into another same old acupuncture discussion. I am saddened to see it moving away from the real topic of the paper, which is not about whether acupuncture works but how we should just use it anyway because it is an effective placebo.
Personally, I find the ethical aspects of the question far, far, far more interesting than just another thread about whether acupuncture works.
11407745 is "Successful removal of a left main coronary artery thrombus induced by vasospasm to the aorta [correction of vasospasm of the aorta] after the injection of contrast medium." Still wrong.
That's not what we're saying. What we're saying is that electrical stimulation has its own effects and hence no study of EA can hope to disentangle the effects of the electricity vs. those of the acupuncture. It's like inserting needles coated in an analgesic, then attributing the result to the needles.
Additionally, I should point out that if acupuncture is just an elaborate placebo, it is:
(A) Unethical to present it as if it is more.
(B) Inappropriate to charge premium prices for it.
(C) Unconscionable to penetrate the skin, posing risk of infection, when other placebos are safer
I'm not going to get into the niggling over what counts as acupuncture - you can't simply demand that acupuncture must work according to ancient Chinese philosophy, therefore if a specific practice is shown to work by some other method than chi flow, it magically becomes "not acupuncture." If we are agreed that electroacupuncture is bioactive, then, I take it that you do not object to people using that, or doctors recommending it, as a first-line alternative to drugs in conditions where it has clinical trial backing?
Scott, you say Gan et al. 2004 (prevention of postop nausea and vomiting) was the only paper I mentioned with a "proper control" because it included both real and sham electroacupuncture as well as ondansetron. You announced that this was "unblinded" apparently by assumption, as the free full text that is available through PubMed notes twice that "The screen on the unit (measuring 4 Ã 2 cm) was covered with an opaque tape in all groups so that the clinicians and research personnel were unaware if the unit was on or off" and that "The anesthesiologists and care providers were blinded to the study group as the screen on the HANâs unit was concealed."
Now, it so happens that in this study, on all measures the real EA looked a lot better than sham EA, with ondansetron somewhere in between (nausea at 2 hours: 19% for EA, 40% for ondansetron, 79% for sham EA; p<0.0001). So if you were the patient, you would darn well want the real thing. However, what if the sham EA had had an effect similar to the real EA, and we both pretended we were sure that the sham EA did not cause meaningful nerve stimulation? You would believe that EA was "only a placebo" and therefore should be rejected as worthless. I would believe that EA was such a bang-up placebo that it was better (and cheaper) than ondansetron, and therefore should be preferred. The demand for the inclusion of sham acupuncture in every study is essentially a demand that every researcher in every country in the world adopt your view of the matter - because if they hold my view, than the sham group only makes the study more costly and/or more likely to be underpowered to answer the questions they really want to ask.
Drat! 11402745. Need more coffee. Or possibly less coffee.
If acupuncture gives more relief than a sugar pill, or indeed a real pharma drug, it's not unethical to present it as doing such. And it's not "unconscionable" to continue to engage in a traditional practice with minute safety risks when the alternative is a drug with enormous side effect rates. Please. Nothing we do in life is 100% risk free nor is that a desirable goal, but it seems pretty darn obvious that (as in the trial referenced by that number, for real this time), you will certainly reduce your risks by electing acupuncture rather than naproxen.
I certainly do so object, for the reasons described previously.
That blinds the comparison between needling and EA; it does not blind the comparison between drug and EA. I should have been more clear; my apologies. More explicitly, there is a reasonable control to measure the effect of the electricity, but there is still no reasonable control to measure the effect of the needles.
All untrue. If naproxen doesn't outperform a placebo for a condition, that's a reason NOT to use naproxen for that condition. It does NOT in any way, shape, or form justify use of the placebo!
Re comment 95 - apparently I triggered one of those HTML things that ate the end of my message by saying that the p value was less than 0.0001. My following argument ran like this:
Suppose that the sham had actually been nearly as good as the EA, and that we knew or imagined we knew that the sham did not cause meaningful dermal stimulation? You would then believe that EA was "only a placebo" and therefore worthless, so they should not use it. I would then believe that EA was such a bang-up placebo that it was much better (and cheaper) than ondansetron, so they should use it as a first choice. The demand for similar "placebo" controls in all studies essentially demands that all researchers worldwide adopt your philosophy. If they share my view of it, the inclusion of a "sham" arm merely increases the study's cost and/or the likelihood that it will be underpowered to answer the questions that they really want to ask.
11402745 is another case of no blinding + no placebo = no meaning.
Sorry for the redundant comment (and thanks to Orac or the webmaster for restoring the eaten portion of comment 95).
Scott, this is not a scientific argument but a cultural one. Science deals only in facts; it has no means of telling us what behaviors are "unethical" or better yet "unconscionable." You want to use only treatments that make you feel better for a specific subset of the possible reasons; I want to use the treatment that will make me feel most better or that has the lowest risk of side effects. I think my belief is as valid as yours; if you want to assert that everyone should adopt your view, you need to make an overt philosophical case for it.
@jane
Proper placebo for EA: electrical stimulation without breaking the skin (preferably blinding the subject and the researchers)
Proper placebo for acupuncture: sham needles that do not break the skin (again, preferably blinding the subject and the researchers). Also, comparison should be made between being "on" the meridians vs. "off" or "wrong" meridians
Proper placebo for EA+drug: as above + sugar pill identical to study drug
The whole reason for using a placebo is to make sure you can tell what is due to the treatment vs. what is due to chance and other non-treatment effects (e.g., expectation, ritual, relaxation, etc.). The reason to blind the study is to prevent biases from creeping in as much as possible, either consciously or unconsciously.
Until you know what a treatments effects are when compared against an appropriate placebo, you cannot compare it against other things (e.g., a drug or a sugar pill), since you don't know what its actual effects are.
No. First, EA needs to be shown to have an effect above and beyond simple electrical stimulation. If it does no better than ES, then there is no justification for it, as it adds risk of infection with no equivalent increase in benefit. Only if EA does better than non-skin-breaking ES can it be studied in comparison to a drug that is also indicated for the condition under investigation. For any indication, the gold standard drug should be used for comparison. If the results of that study show that EA is no worse than or better than the drug (in terms of efficacy + side effects) should it be considered as an alternative treatment. It should not become routine until numerous other studies have replicated the results. That's a lot of "ifs" to overcome, and we aren't there yet.
I'm talking about generally accepted medical ethics.
re: 98 - There, see? You are not the sole arbiter of what constitutes meaning; apparently the journal editors and reviewers thought the paper contained some level of meaning. "No meaning" really means "They asked a question that I personally do not think deserves to be asked as such."
Science is nothing more or less than a method for answering questions about the world; it's not a set of beliefs that a person must hold before he can set out to answer the questions that fit within the framework of that belief system. Ideally, it lets people of different worldviews work together to identify a certain number of things that they can agree on as true or false. This does not work if one demands that everyone adopt one's worldview going into the process. You might consider being just a little bit more humble.
Actually, minor correction. Before EA and ES are compared, you would need to show that ES is actually effective for the condition being examined. Then you could look at EA vs. ES. It might be possible to shorten the approach by using multiple arms to include placebos for both EA and ES.
Ladies and gentlemen, I present to you a textbook example of blowing up meaningless semantics instead of actually addressing the point.
@jane
What Scott means by "no meaning" is that due to methodological flaws, the results gained from the study are too uncertain to draw any meaningful conclusion from. There is too great a possibility that the results are due to chance or bias and do not represent reality. If there are no controls and no blinding, the chances of biases skewing the data jumps way up.
Scott: "All untrue. If naproxen doesn't outperform a placebo for a condition, that's a reason NOT to use naproxen for that condition. It does NOT in any way, shape, or form justify use of the placebo!"
Scott, why would you dismiss the placebo or the naproxen when you never asked the question "how good is the placebo?"
That's very rigid of you. But along the lines of SBM so it's expected.
http://www.nature.com/neuro/journal/v13/n7/full/nn0710-783.html
I dont really care about acupuncture myself, but I thought that this pre-digested article talking about a possible mechanism of action for acupuncture was interesting food for thought. It will be very interesting to see if people can replicate this work and expand further on it.
@ Sust:
asked and answered.
Oops, I tried to post this as a hyperlink and it got stuck in moderation. @Sust:
if you type 'nature neuroscience' into the search box up on the left, you'll find a comprehensive discussion of this 'food for thought'.
http://scienceblogs.com/insolence/2010/06/when_what_an_acupuncture_stud…
@jenbphillips: Thanks for the link :)
My jaw dropped when I saw that article in Nat. Neuro. and I too was wowed by the scientific methodology they used plus the cool implications of adenosine in pain control. Even if they don't follow-up the work in the proper direction, I think their work will get the ball rolling.