Acupuncture - Still Doesn't Work

Both Steven Novella and Orac posted today about the recent acupuncture study that supposedly shows acupuncture works better than real medicine. Both Orac and Novella’s posts examine the study and its weaknesses in some detail and so I won’t try to do the same. But I do have a couple of brief points I want to make about this study and acupuncture studies in general, and before I do that I’ll have to briefly summarize the main points of this study.

The Study

The researchers, amongst other things, wanted to compare the effectiveness of acupuncture on pain, compared with “groups receiving standard medical care.”  Remember that phrase. “Standard medical care” essentially means whatever their physicians had already prescribed for each patient – medical treatments or physical therapies. Remember that. The study divided the test patients into four groups, namely:”

1. “Standard medical care” plus individualized acupuncture.
2. “Standard medical care” plus standardized acupuncture
3. “Standard medical care” plus sham acupuncture (toothpicks that didn’t penetrate the skin)
4. “Standard medical care” only.

The reports of the study generally don’t describe the first three groups in quite that way, but the above is a correct description of what they actually did – all four groups continued with the standard treatment their doctors had prescribed before.

The result was that: there was no significant difference between groups 1, 2 and 3, but groups 1, 2 and 3 were each better than group 4. This is being reported as “both real and sham acupuncture better than conventional treatment”. But the study shows no such thing. There are (at least) three fundamental flaws with this conclusion, namely:

1. The study wasn’t blinded. This means there was no real control for placebo, and so any comparisons of acupuncture with the conventional treatment are worthless.
2. All three acupuncture groups had conventional treatment as well as acupuncture, so the study can not possibly show that acupuncture is better than conventional treatment. For that to be even possible, the acupuncture groups would have had to be acupuncture only – no conventional treatment. This is such an obvious “duh” point I don’t know whether to laugh or cry that apparently even the researchers didn’t think it was important.
3. Sham acupuncture was (again) just as good as the ‘real”stuff. But this does not mean that sham works as well as non-sham acupuncture, as has been reported. As Steven Novella wrote, when your real treatment is no better than your placebo (the “sham” acupuncture), you don’t get to conclude that, well, this means the placebo works too. No – you conclude the real treatment doesn’t work.

Worthless Acupuncture Studies

Yet again we have a worthless acupuncture study that is being falsely touted as showing acupuncture works. Again, this study is on supposed pain reduction. What about all the other things that acupuncture is supposed to fix? It releases blocked qi, yes? Shouldn’t it do more than reduce pain? Look at this list: Acupuncture: Conditions it Treats - Gastrointestinal Disorders, Urogenital Disorders, Gynecological Problems – the list goes on and on. So why are we always shown pain reduction only?

Second, why do acupuncture researchers always end saying something like this study “raises some new questions about how acupuncture works”? No it doesn’t. It shows it doesn’t work.

And third, why do they always call for more studies? Josephine P. Briggs, MD, director of NCCAM, is quoted saying:

Future research is needed to delve deeper into what is evoking these responses."

Why? There is enough research already for us to conclude that acupuncture is a placebo.

I’m going to make a prediction. Before this year is out, there will be at least one more study of acupuncture that shows “sham” acupuncture is as good as the “real” stuff, and / or that acupuncture is better than an (unblinded) non-acupuncture group. The researchers will say (1) this shows there is something going on with acupuncture, and (2) we need more studies to be done on acupuncture. More studies will be needed. I guarantee it.

Rinse and repeat next year, the year after…

Other links

Articles by Steven Novella:

Does Acupuncture Work or Not?

Acupuncture Does Not Work for Back Pain

Acupuncture and Back Pain - Part II

Orac’s Yawn...another overhyped acupuncture study

Several posts by me:

Placebo - pregnancy pain cure

Acupuncture does not cut blood pressure

Still no evidence acupuncture works

Acupuncture – it really really doesn't work

No point to acupuncture on animals

Equivocation on acupuncture

One More Time - Acupuncture Does NOT Work

Comments

[Swiftsure]

I think an acupuncture training manual should have a picture of an acupuncturist in his surgery, bearing the legend, "Prick with a needle."

[Dunc]

The last study of this type (almost identical in fact) that I recall only included subjects with chronic, non-responsive lower back pain - i.e. they only looked at people who had been receiving standard care for a prolonged period and for whom that care had not been effective. So it's not exactly surprising that standard care didn't perform well in the trial...

Looking at the inclusion criteria for this study (at least as far as they are quoted on your NHS link), this one is better, but only slightly:

The researchers enrolled 641 adults aged 18 to 70 years who had experienced uncomplicated low back pain for between three and 12 months and who had never tried acupuncture before. To be eligible, the participants had to have rated their back pain as at least a three on a scale ranging from zero to 10...

[My emphasis]

Three to twelve months of at least moderate back pain makes me suspect that standard care isn't working too well.

"So why are we always shown pain reduction only?"

Oh, that's easy - pain is highly subjective, and therefore highly amenable to placebo effects. Hell, a high level of concentration will block out a surprising amount of pain, and I'm sure we've all had that experience of an unnoticed injury, which suddenly starts hurting as soon as you notice it.

[breakerslion]

I have a defective lower back. I've had it all my life. The best advice I have ever received for relieving pain without drugs is, "Get on all fours on the floor." It really helps not to be upright. What do you suppose is the benefit of lying on your stomach on a padded board for about half an hour?

[TechSkeptic]

I truly enjoyed one of the comments at oracs...

A better headline for this study would be:

"Randomly placed toothpicks work as well as real acupuncture!"

[Random]

You mean I don't need to dress in drag as a dominatrix to get paid for sticking needles in people?

Whew! That's a relief.

Shame to have to return those heels though...

[Pelger]

they show the pain thing only because it would be too obious even for them that an exposed bone fracture can't be cured adding insult to injury and sticking a needle on the poor guy's fractured arm :)

[Ian]

all of these mumbo jumbo 'cures' seem to involve a bit of good old human 'hands on' in one form or another, we always tend to feel 'better' regardless of the condition we have if someone concentrates their interest on us for an hour or two, in contrast to a gp visit where we are told to take our medicine etc in a quick 3 minute consultation. Also after some such 'therapy'we will almost always say we feel 'better' even if we do not, out of simple good manners!

[David King]

Oh Dear.

"NHS to give back pain acupuncture"

http://news.bbc.co.uk/1/hi/health/8068427.stm

[derek hudson]

I am not surprised that the NHS is to offer acupuncture, for 'back problems', as I understand it. If NICE looks at the evidence for effectiveness in this area it will find that acupuncture DOES work; that is to say, patients report an improvement in their condition after receiving acupuncture. Although the evidence points to a PLACEBO effect, the fact that IT does work AND that acupuncture is widely available puts NICE in a tricky position where they feel that they have to 'go with the evidence'. I feel quite sure that rubbing a patient's back with concentrated horse urine (or ANY treatment administered by a caring person, preferably in a white coat) would be equally effective, but such treatments are not well established, of course.

[Sue]

My mother has Trigenimenal Neuralgia for years and suffers greatly from periodical bouts with it. She takes medication for it. During one of the times she was suffering and the medication didn't help at all, a friend suggested she try accupuncture. She was very reluctant to try it but after weeks of severe pain she sceduled an appointment. She received a lot of relieve after the sessions, although not a cure and hasn't had a severe bout in 2 years.

Whether the accupuncture helped or it was just a placebo doesn't matter to me. What is important was that she was able to resume her life without being in immeasureable pain.

[Martin]

Hi Sue

Hypothetical situation: What if a doctor knew his treatment showed no effect beyond the placebo, but still continued to treat people with it -- would that bother you?

[TechSkeptic]

suffers greatly from periodical bouts with it

Why are you so sure that she simply isn't between bouts?

She received a lot of relieve after the sessions
How many sessions? over how long a period? how long does a 'bout' last?

Did she try anything else? There are many pharmacological avenues for this condition. There are even a number of surgeries. Is hers of dental origin? Did she get a root canal? Were all these things tried, or is simply settling for lying going to be good enough (it wont be forever).

I prefer my medicine to be administered without deceit being required to make it work.

[yakaru]

Sue, that is great for your mother, but if you are going to the trouble of making a public statement, why not take the extra step of trying to make a meaningful one instead of a declaration of willful ignorance? The evidence says clearly that any benefit your mother has had is a placebo effect, and that your acupuncturist is either ignorant or lying. Agreed?

Techskeptic, do you know of any studies showing what happens when a placebo is openly administered as a placebo? Deceit might not be necessary for it to work. I suspect the whole concept of a placebo is only necessary because of the scientific pretentions of quacks.

If we have an evolved mechanism that delays switching on the body's healing mechanisms till we relax (i.e. till danger has passed and its a good time to expend energy on healing instead of flight from a threat), then a situation that awakens trust and relaxation might switch on that mechanism without deception being necessary.

Cuddling up with a teddy bear, for example, might be just as successful as paying some pretentious wanker to stick pins in you.

[Sue]

Hi Martin and TechSkeptic,

Thank you for your input.

I would think it very unethical if a Doctor knew his treatment was only a placebo and administered it anyway.

My mother is 85 years old and has suffered with Trigenimenal Neuralgia for over 30 years. She has had false teeth since her early 30's, so it wasn't caused by a root canal.

She has been to lots of doctors and has tried different medicines. Surgery is her last resort for treatment as it carries a certain amount of risks.

I think she had 3 or 4 accupuncture treatments. She had weeks of excruciating pain before having accupuncture and the medicine she was taking wasn't helping.

She has not had a severe bout since she had the accupuncture, however, she continues to take the medicine that is prescribed for her condition.

I am not supporting accupuncture as I am no expert on the subject, I am only saying that it seemed to give my mother relieve and after seeing the way she suffers when her Trigenimenal Neuralgia ia acting up, I don't blame her for trying it.

[Sue]

Hmmm, the Mayo Clinic says acupunture may help with pain in some cases:

http://www.mayoclinic.com/health/acupuncture/sa00086

[Sue]

Oops, I stand corrected. After googling acupuncture and reading all I could find on the subject including the long and well written article on acupuncture in The Skeptic's Dictionary, I have to agree that any benefit received from acupuncture is probably due to the placebo effect.

[yakaru]

Well done Sue. I hope the placebo continues to help your mother. Sorry if I came across unsympathetic above, but it wasn't meant personally. The success treating your mother is exactly the kind of case which acupuncturists use to overstate their effectiveness.

For every case like your mother's there are, statistically speaking, two or three other people who are also at their wits' end with pain who try acupuncture without success.

[Martin]

With the new information Sue now possess, should she tell her 85 year old mom that it is probably a placebo, and risk interfering with her current state of healing and well-being, or should she keep quiet and (to put it bluntly) lie to her mother?

I guess the truth really hurts sometimes..

[TechSkeptic]

Martin,

The question is not whether she should tell her that acupuncture is quackery or not. The reason is that her mother will likely not care. I do this with my sister all the time with her newest cure for something (no milk, aloe treatments, some other woo, etc). I explain about why we examine evidence, about how our minds have evolved to play certain tricks on us, paredolia, placebo etc etc.

It doesnt matter. The response is always "well it works for me". Which is likely what her mom will say. And honestly, that is fine. The harm from placebo, usually doesn't come from the placebo itself, it usually comes when the patient forgos an effective treatment. Like treating cancer with acupuncture and not medicine. This is what I watch out for with my sister.

I doubt very much the her mom will be affected one way or another by disclosing the information the acupuncture is nonsense. However, sue should probably watch out for stubborn loyalty to this treatment when the placebo wears off. That is the conversation to have.

Sue, good for you. Its a rare instance we see anyone have one set of idea, look it up and discuss it, and actually change their mind. We admire true fence sitters who get off the fence based on evidence.

She has not had a severe bout since she had the acupuncture, however, she continues to take the medicine that is prescribed for her condition.

This is common for all CAM treatments, and what happens, for some reason, is that any relief is assigned to the CAM and not to the medicine, or the natural tendency for many symptoms to come and go in waves (you call them bouts).

Yakaru,
I can certainly look for some specific studies on placebo, but Singh's book Trick or Treatment covers it pretty well. Placebo may not be an act of deliberate deceit, but ignorance also. I have no doubt many CAM practitioners beleive they do what they can do, but that is becuase they don't check with strict testing protocols.

Medical procedures and pills have placebo also. Its very powerful to say "this treatment is expensive and has been shown to work well" and administer a sugar pill. I have no idea why "this treatment is over 2000 years old" imparts any placebo at all.

:)

[Jimmy_Blue]

Sue:

She has not had a severe bout since she had the accupuncture, however, she continues to take the medicine that is prescribed for her condition.

As TechSkeptic said, it is common that people who recieve acupuncture are also continuing to recieve their conventional medical treatment - what a lot of people don't also know is that often when studies are carried out into acupuncture the patients who recieve acupuncture are also continuing to recieve conventional medical treatment.

Yet the studies are always touted by the media and woos as showing acupuncture alone works better than x treatment and you have to look very hard at the studies themselves to see exactly what other treatments the acupuncture group was recieving at the same time. Combine this with the large number of very poorly designed studies and you get a seemingly large volume of studies showing acupuncture has some sort of effect.

[Julian09]

Acupuncture has been a hot issue of discussion in the medical world since the late 20th century. The WHO has recognized and concluded that acupuncture is effective for the treatment of 28 diseases and still there is evidence to suggest that it would be effective for several other diseases.

Recently, acupuncture is acknowledged as the most efficient way of treating various diseases all over the world. There are lots of benefits in acupuncture. It is not only an effective way of treating diseases but it is also inexpensive when compared to other forms of treatment.
Acupuncture Books

[Ramel]

Yeah, but Rodger Daltry is also a homeopathy booster so their opinion is pretty meaningless...

[yakaru]

Julian,
First, a rhetorical question. Why is it that people who claim that "studies have shown" alternative medicine works never take the time to link the actual study?

I went looking for it and the first thing that came up was this site
Acupuncture Schools.us which mentions the study but doesn't link to it, and doesn't give any reference details, even though they quote extensively from it. Why such a lapse in basic referencing skills? Didn't they attend high school?

Anyway, I dug a bit more and found a pdf of the report.

Yes, it is published under the auspices of the WHO, and yes it does finish up claiming acupuncture is effective for a long list of illnesses.

However, the author, Dr Xiaorui Zhang, from the "tradition medicine" office of WHO plays the usual tricks in data gathering - or should I say data-mining.

There is a constant shifting of the goal posts going on from start (where he notes that the studies were at times poorly reported, but he would use them anyway and try to be honest with the data) to finish, where he presents it as if the results are unequivocally in favor of acupuncture.

He says he has only used studies with a control group (good) but doesn't seem to mention whether or not he excluded studies which showed no effect above the placebo. What makes me especially suspicious is that although the reported results show acupuncture always spectacularly outperforming the control, he nevertheless takes a couple of paragraphs to argue that comparing acupuncture to a placebo control will show acupuncture in an unfair light.

The he skips a few steps further --
...However, if rapid improvement can be achieved in the treatment of a long-standing chronic disease...the effect of acupunture should be viewed in a more favorable light, even when a well-designed, controlled study has not been carried out. (p.4)

Then he lowers the bar a bit more, claiming that
the therapeutic effect depends greatly on the proficiency of the acupuncturists...This may partly explain the disparities or inconsistencies by different authors, even when their studies were carried out on equally sound methodological bases.

So his assumption that acupuncture works is now acknowledged as a possible tool in evaluating whether or not acupuncture works. Sound methodology indeed! So what happened to these studies with "discrepancies? Do they make it into the final cut up or not?

A bit further down he proclaims acupuncture works "by mobilization of the organism's own potential...". How the hell does he know that? He doesn't say. But he does go further to explain that "it may not work if the mobilization of the individual's potential is not adequate for recovery"

That squeaking sound you hear is the bar being lowered again.

So then we get to the table of the results of the many dozens of controlled studies. In the vast majority, acupuncture has a success rate higher than 90%, while the control (often treated with "western" medicine) is often as low as 30%.

Certainly looks impressive. Every bit as impressive as when African dictators get re-elected with 97% of the vote.

Why then all the hand-wringing about how difficult it is to evalauate, and why discrepancies might appear? Why no mention of whether he excluded studies which showed no significant return? Why use studies which were poorly recorded?

The condition of "convulsions in infants due to high fever" was treated with phenobarbital with a 51% success rate, where acupuncture returned 98%.

No wonder acupuncture fans don't provide the link to this piece of embarrassing lunacy.

[yakaru]

Aaaagghhhhh!!! I screwed up the italics again! Now it's going to make everyones comments in italics. Jeeeezuz. Mea Culpa.

[Big Al]

Don't worry, yakaru. God will magic away your sinful italics...

Well, come on then! Get busy with the omnipotent schtick, Jehovah!

[Big Al]

There you are. Told you the Big Man would intervene in His mysterious way.

[yakaru]

It's the power of prayer.

[Big Al]

Hmmm... but suppose I got it wrong?

Suppose it wasn't a non-existent Jehovah, but an existent Mithra, Quetzalcoatl, Odin, Ahura Mazda, Horus, Astaroth, John Frum or some other god in whose name I didn't happen to be brought up and of whom I am therefore unfortunately unaware?

They'll be nasty to me forever after, in their mercy! I'm so confused? How do I get the right god?

I am now wearing a tinfoil skullcap and getting nervous of thunderstorms...