Acupuncture is a theatrical placebo: the end of a myth

It is easy to accept acupuncture as a treatment simply because it is clearly harmless. Massage therapy runs greater risks.  That it provides zero results is more surprising considering how it came to be somewhat accepted here over the past fifty years.

Yet not so surprising considering how it was adopted.   After the initial hoopla, clinical science remained scant and the protocol still was readily introduced.  I noted that at the time and wondered.

Here we are two generations later  and the mass of work shows essentially nothing.  No end of herbal concoctions survived their encounter with the scientific method and are willingly used today.  After all, if a safe herb will work for the bulk of victims why bother with a pharmaceutical with side effects until the herb is ruled out.

In the meantime i wonder about the nervous system mapping concept as well and if it is useful at all.  The same holds for reflexology which likely benefits at a modest level from reports to hand.  The  science there is at least real enough as nerve ends are been broadly stimulated.  Why should we not be doing this? .

.

.

Acupuncture is a theatrical placebo: the end of a myth

May 30th, 2013 · 60 Comments

https://www.blogger.com/blogger.g

Anesthesia & Analgesia is the official journal of the International Anesthesia Research Society. In 2012 its editor, Steven Shafer, proposed a head-to-head contest between those who believe that acupuncture works and those who don’t. I was asked to write the latter. It has now appeared in June 2013 edition of the journal [download pdf]. The pro-acupuncture article written by Wang, Harris, Lin and Gan appeared in the same issue [download pdf].

Acupuncture is an interesting case, because it seems to have achieved greater credibility than other forms of alternative medicine, despite its basis being just as bizarre as all the others. As a consequence, a lot more research has been done on acupuncture than on any other form of alternative medicine, and some of it has been of quite high quality. The outcome of all this research is that acupuncture has no effects that are big enough to be of noticeable benefit to patients, and it is, in all probablity, just a theatrical placebo.

After more than 3000 trials, there is no need for yet more. Acupuncture is dead.

ACUPUNCTURE IS A THEATRICAL PLACEBO

David Colquhoun (UCL) and Steven Novella (Yale)

Anesthesia & Analgesia, June 2013 116:1360-1363.

Pain is a big problem. If you read about pain management centres you might think it had been solved. It hasn’t. And when no effective treatment exists for a medical problem, it leads to a tendency to clutch at straws. Research has shown that acupuncture is little more than such a straw.

Although it is commonly claimed that acupuncture has been around for thousands of years, it hasn’t always been popular even in China. For almost 1000 years it was in decline and in 1822 Emperor Dao Guang issued an imperial edict stating that acupuncture and moxibustion should be banned forever from the Imperial Medical Academy.

Acupuncture continued as a minor fringe activity in the 1950s. After the Chinese Civil War, the Chinese Communist Party ridiculed traditional Chinese medicine, including acupuncture, as superstitious. Chairman Mao Zedong later revived traditional Chinese Medicine as part of the Great Proletarian Cultural Revolution of 1966 (Atwood, 2009). The revival was a convenient response to the dearth of medically-trained people in post-war China, and a useful way to increase Chinese nationalism. It is said that Chairman Mao himself preferred Western medicine. His personal physician quotes him as saying “Even though I believe we should promote Chinese medicine, I personally do not believe in it. I don’t take Chinese medicine” Li {Zhisui Li. Private Life Of Chairman Mao: Random House, 1996}.

The political, or perhaps commercial, bias seems to still exist. It has been reported by Vickers et al. (1998) (authors who are sympathetic to alternative medicine) that

"all trials [of acupuncture] originating in China, Japan, Hong Kong, and Taiwan were positive"(4).

Acupuncture was essentially defunct in the West until President Nixon visited China in 1972. Its revival in the West was largely a result of a single anecdote promulgated by journalist James Reston in the New York Times, after he’d had acupuncture in Beijing for post-operative pain in 1971. Despite his eminence as political journalist, Reston had no scientific background and evidently didn’t appreciate the post hoc ergo propter hoc fallacy, or the idea of regression to the mean.

After Reston’s article, acupuncture quickly became popular in the West. Stories circulated that patients in China had open heart surgery using only acupuncture (Atwood, 2009). The Medical Research Council (UK) sent a delegation, which included Alan Hodgkin, to China in 1972 to investigate these claims , about which they were skeptical. In 2006 the claims were repeated in 2006 in a BBC TV program, but Simon Singh (author of Fermat’s Last Theorem) discovered that the patienthad been given a combination of three very powerful sedatives (midazolam, droperidol, fentanyl) and large volumes of local anaesthetic injected into the chest. The acupuncture needles were purely cosmetic.

Curiously, given that its alleged principles are as bizarre as those on any other sort of pre-scientific medicine, acupuncture seemed to gain somewhat more plausibility than other forms of alternative medicine. The good thing about that is that more research has been done on acupuncture than on just about any other fringe practice.

The outcome of this research, we propose, is that the benefits of acupuncture, if any, are too small and too transient to be of any clinical significance. It seems that acupuncture is little or no more than a theatrical placebo. The evidence for this conclusion will now be discussed.

Three things that are not relevant to the argument

There is no point in discussing surrogate outcomes such as fMRI studies or endorphine release studies until such time as it has been shown that patients get a useful degree of relief. It is now clear that they don’t. 

There is also little point in invoking individual studies. Inconsistency is a prominent characteristic of acupuncture research: the heterogeneity of results poses a problem for meta-analysis. Consequently it is very easy to pick trials that show any outcome whatsoever. Therefore we shall consider only meta-analyses.

The argument that acupuncture is somehow more holistic, or more patient-centred, than medicine seems us to be a red herring. All good doctors are empathetic and patient-centred. The idea that empathy is restricted to those who practice unscientific medicine seems both condescending to doctors, and it verges on an admission that empathy is all that alternative treatments have to offer.

There is now unanimity that the benefits, if any, of acupuncture for analgesia, are too small to be helpful to patients.

Large multicenter clinical trails conducted in Germany {Linde et al., 2005; Melchart et, 2005; Haake et al, 2007, Witt et al, 2005), and in the United States {Cherkin et al, 2009) consistently revealed that verum (or true) acupuncture and sham acupuncture treatments are no different in decreasing pain levels across multiple chronic pain disorders: migraine, tension headache, low back pain, and osteoarthritis of the knee.

If, indeed, sham acupuncture is no different from real acupuncture the apparent improvement that may be seen after acupuncture is merely a placebo effect. Furthermore it shows meridians don’t exist, so the "theory" memorized by qualified acupuncturists is just myth. All that remains to be discussed is whether or not the placebo effect is big enough to be useful, and whether it is ethical to prescribe placebos.

Some recent meta-analyses have found that there may be a small difference between sham and real acupuncture. Madsen Gøtzsche & Hróbjartsson {2009) looked at thirteen trials with 3025 patients, in which acupuncture was used to treat a variety of painful conditions. There was a small difference between ‘real’ and sham acupuncture (it didn’t matter which sort of sham was used), and a somewhat bigger difference between the acupuncture group and the no-acupuncture group. The crucial result was that even this bigger difference corresponded to only a 10 point improvement on a 100 point pain scale. A consensus report (Dworkin, 2009) that a change of this sort should be described as a “minimal” change or “little change”. It isn’t big enough for the patient to notice much effect.

The acupuncture and no-acupuncture groups were, of course, not blind to the patients and neither were they blind to the practitioner giving the treatment. It isn’t possible to say whether the observed difference is a real physiological action or whether it’s a placebo effect of a rather dramatic intervention. Interesting though it would be to know this, it matters not a jot, because the effect just isn’t big enough to produce any tangible benefit.

Publication bias is likely to be an even greater problem for alternative medicine than it is for real medicine, so it is particularly interesting that the result just described has been confirmed by authors who practise, or sympathise with, acupuncture. Vickers et al. (2012) did a meta-analysis for 29 RCTs, with 17,922 patients. The patients were being treated for a variety of chronic pain conditions. The results were very similar to those of Madsen et al.{2009). Real acupuncture was better than sham, but by a tiny amount that lacked any clinical significance. Again there was a somewhat larger difference in the non-blind comparison of acupuncture and no-acupuncture, but again it was so small that patients would barely notice it.

Comparison of these two meta-analyses shows how important it is to read the results, not just the summaries. Although the outcomes were similar for both, the spin on the results in the abstracts (and consequently the tone of media reports) was very different. 

An even more extreme example of spin occurred in the CACTUS trial of acupuncture for " ‘frequent attenders’ with medically unexplained symptoms” (Paterson et al., 2011). In this case, the results showed very little difference even between acupuncture and no-acupuncture groups, despite the lack of blinding and lack of proper controls. But by ignoring the problems of multiple comparisons the authors were able to pick out a few results that were statistically significant, though trivial in size. But despite this unusually negative outcome, the result was trumpeted as a success for acupuncture. Not only the authors, but also their university’s PR department and even the Journal editor issued highly misleading statements. This gave rise to a flood of letters to the British Journal of General Practice and much criticism on the internet.

From the intellectual point of view it would be interesting to know if the small difference between real and sham acupuncture found in some, but not all, recent studies is a genuine effect of acupuncture or whether it is a result of the fact that the practitioners are never blinded, or of publication bias. But that knowledge is irrelevant for patients. All that matters for them is whether or not they get a useful degree of relief.

There is now unanimity between acupuncturists and non-acupuncturists that any benefits that may exist are too small to provide any noticeable benefit to patients. That being the case it’s hard to see why acupuncture is still used. Certainly such an accumulation of negative results would result in the withdrawal of any conventional treatment.

Specific conditions

Acupuncture should, ideally, be tested separately for effectiveness for each individual condition for which it has been proposed (like so many other forms of alternative medicine, that’s a very large number). Good quality trials haven’t been done for all of them. It’s unlikely that acupuncture works for rheumatoid arthritis, stopping smoking, irritable bowel syndrome or for losing weight. And there is no good reason to think it works for addictions, asthma, chronic pain, depression, insomnia, neck pain, shoulder pain or frozen shoulder, osteoarthritis of the knee, sciatica, stroke or tinnitus and many other conditions (Ernstet al., 2011).

In 2009, the UK’s National Institute for Clinical Excellence (NICE) did recommend both acupuncture and chiropractic for back pain. This exercise in clutching at straws caused something of a furore. In the light of NICE’s judgement the Oxford Centre for Evidence-based medicine updated its analysis of acupuncture for back pain. Their verdict was

“Clinical bottom line. Acupuncture is no better than a toothpick for treating back pain.”

The paper by Artus et al. (2010) is of particular interest for the problem of back pain. Their Fig 2 shows that there is a modest improvement in pain scores after treatment, but much the same effect, with the same time course is found regardless of what treatment is given, and even with no treatment at all. They say

“we found evidence that these responses seem to follow a common trend of early rapid improvement in symptoms that slows down and reaches a plateau 6 months after the start of treatment, although the size of response varied widely. We found a similar pattern of improvement in symptoms following any treatment, regardless of whether it was index, active comparator, usual care or placebo treatment”.

It seems that most of what’s being seen is regression to the mean. And that is very likely to be the main reason why acupuncture sometimes appears to work when it doesn’t.

Although the article by Wang et al (2012) was written to defend the continued use of acupuncture, the only condition for which they claim that there is any reasonably strong evidence is for post-operative nausea and vomiting (PONV). It would certainly be odd if a treatment that had been advocated for such a wide variety of conditions turned out to work only for PONV. Nevertheless, let’s look at the evidence.

The main papers that are cited to support the efficacy of acupuncture in alleviation of PONV are all from the same author: Lee & Done (1999), and two Cochrane reviews, Lee & Done (2004), updated in Lee & Fan (2009). We need only deal with this latest updated meta-analysis.

Although the authors conclude “P6 acupoint stimulation prevented PONV”, closer examination shows that this conclusion is very far from certain. Even taken at face value, a relative risk of 0.7 can’t be described as “prevention”. The trials that were included were not all tests of acupuncture but included several other more or less bizarre treatments (“acupuncture, electro-acupuncture, transcutaneous nerve stimulation, laser stimulation, capsicum plaster, an acu-stimulation device, and acupressure”). The number needed to treat varied from a disastrous 34 to a poor 5 for patients with control rates of PONV of 10% and 70% respectively.

The meta-analysis showed, on average, similar effectiveness for acupumcture and anti-emetic drugs. The problem is that the effectiveness of drugs is in doubt because an update to the Cochrane review has been delayed (Carlisle, 2012) by the discovery of major fraud by a Japanese anesthetist, Yoshitaka Fujii (Sumikawa, 2012). It has been suggested that metclopramide barely works at all (Bandolier, 2012; Henzi, 1999).

Of the 40 trials (4858 participants) that were included; only four trials reported adequate allocation concealment. Ninety percent of trials were open to bias from this source. Twelve trials did not report all outcomes. The opportunities for bias are obvious. The authors themselves describe all estimates as being of “Moderate quality” which is defined this: “Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate”. That being the case, perhaps the conclusion should have been “more research needed”. In fact almost all trials of alternative medicines seem to end up with the conclusion that more research is needed.

Conclusions

It is clear from meta-analyses that results of acupuncture trials are variable and inconsistent, even for single conditions. After thousands of trials of acupuncture, and hundreds of systematic reviews (Ernst et al., 2011), arguments continue unabated. In 2011, Pain carried an editorial which summed up the present situation well.

“Is there really any need for more studies? Ernst et al. (2011) point out that the positive studies conclude that acupuncture relieves pain in some conditions but not in other very similar conditions. What would you think if a new pain pill was shown to relieve musculoskeletal pain in the arms but not in the legs? The most parsimonious explanation is that the positive studies are false positives. In his seminal article on why most published research findings are false, Ioannidis (2005) points out that when a popular but ineffective treatment is studied, false positive results are common for multiple reasons, including bias and low prior probability.”

Since it has proved impossible to find consistent evidence after more than 3000 trials, it is time to give up. It seems very unlikely that the money that it would cost to do another 3000 trials would be well-spent. 

A small excess of positive results after thousands of trials is most consistent with an inactive intervention. The small excess is predicted by poor study design and publication bias. Further, Simmons et al(2011) demonstrated that exploitation of "undisclosed flexibility in data collection and analysis" can produce statistically positive results even from a completely nonexistent effect. With acupuncture in particular there is documented profound bias among proponents (Vickers et al., 1998). Existing studies are also contaminated by variables other than acupuncture – such as the frequent inclusion of "electroacupuncture" which is essentially transdermal electrical nerve stimulation masquerading as acupuncture.

The best controlled studies show a clear pattern – with acupuncture the outcome does not depend on needle location or even needle insertion. Since these variables are what define "acupuncture" the only sensible conclusion is that acupuncture does not work. Everything else is the expected noise of clinical trials, and this noise seems particularly high with acupuncture research. The most parsimonious conclusion is that with acupuncture there is no signal, only noise.

The interests of medicine would be best-served if we emulated the Chinese Emperor Dao Guang and issued an edict stating that acupuncture and moxibustion should no longer be used in clinical practice. 

No doubt acupuncture will continue to exist on the High Streets where they can be tolerated as a voluntary self-imposed tax on the gullible (as long as they don’t make unjustified claims).