Manual acupuncture versus sham acupuncture and usual care for prophylaxis of episodic migraine without aura: multicentre, randomised clinical trial
BMJ 2020; 368 doi: https://doi.org/10.1136/bmj.m697 (Published 25 March 2020) Cite this as: BMJ 2020;368:m697Linked Editorial
Manual acupuncture for migraine
All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
Designing a sham acupuncture control which is indistinguishable from true acupuncture, but that lacks acupuncture-specific effects, has been the key challenge in designing acupuncture randomized trials.[1] In the recent sham-controlled trial of acupuncture for headache in The BMJ,[2] the sham needle used has a blunt tip that applies pressure to the skin that can feel like a pricking sensation; however, it does not penetrate the skin but instead retracts up into the shaft when pressed against the skin, like a retracting dagger. This sham needle has been successfully validated in a brief crossover trial[3] in which acupuncture-naive volunteers were randomized to a real or sham needle at a single point for two minutes each without manipulation and were then asked if they felt the needle penetrate the skin. However, Xu 2020 used this sham in a much more intensive acupuncture intervention which involved 20 sessions over 8 weeks with 10-14 needles per session in the true acupuncture group and 8 in the sham group. In addition, at each 30-minute session, the needles were manipulated 4 times. This was done in the true acupuncture group to achieve de-qi,[4] which is a composite of unique sensations including aching, soreness, pressure, tingling, and dull pain, and which results from the acupuncturist manipulating the needle at a specific direction, angle and depth. There was an attempt to rotate the non-penetrating needles in the sham acupuncture group to create the appearance of similar interventions (Appendix 2: acupuncture ritual), but it is not just the appearance of needle manipulation but also the sensations resulting from it that can differ between groups. As Professor Edzard Ernst has pointed out in his comment on this trial, attempt to elicit de-qi “by the manual manipulation of needles...is usually successful in the majority of patients. In the control group, where non-penetrating needles were used, no de-qi could be generated. This means that the two groups must have been at least partly de-blinded.”[5] However, blinding was reportedly maintained at the end of the 24-week trial with “...no significant difference between the manual acupuncture and sham acupuncture groups for patients’ ability to correctly guess their allocation status (P=0.891)” (Appendix 3: Supplementary Table E).
Additional supporting detail on the blinding procedure and the assessment of its success would be helpful to further understand how this high level of blinding had been maintained.
First, were challenges encountered in recruiting 150 acupuncture-naive patients in China? Given how well acupuncture is in ingrained in the Chinese culture, one might expect that most patients would have already tried it if they had an interest in it. The study flow chart (Figure 1) suggests that of 60 patients screened and excluded at baseline, only 6 (10%) were excluded because they were not acupuncture naive. Could some patients have only claimed they were acupuncture naive to get 20 free acupuncture sessions?
Second, how were potential participants informed about treatment options during enrollment? If they were informed that they may get assigned to either true or sham acupuncture, this would increase the risk of unblinding because it would likely motivate enrolled patients to look for hints to determine whether or they were getting real or sham treatment. On the other hand, if potential participants were informed that they may get one of two different types of real acupuncture, this would not be fully accurate and would not respect the patients’ right to make an informed decision about participation in this 20-week trial.
Third, what was the specific question patients were asked at the end of the trial to determine if blinding was successful? If patients were only asked whether they thought the needles had penetrated the skin, this single question may not adequately capture the difference between the true and sham acupuncture interventions because in addition to needle penetration, one other major difference between the groups is that de-qi could be elicited only in the true acupuncture group. The authors compare the “intensity of de-qi feelings” with the pulses used in electrical acupuncture. Both are obvious and pronounced sensations felt only by patients in the true acupuncture group, which makes blinding for these acupuncture interventions such a challenge. If patients had been asked whether they experienced de-qi sensations at the end of the trial, one would expect there would have been significant differences between groups.
1. Manheimer EW. Systematic reviews in the field of complementary and alternative medicine: importance, methods and examples concerning acupuncture. PhD thesis, Vrije Universiteit Amsterdam. Available at: https://research.vu.nl/en/publications/systematic-reviews-in-the-field-o...
2. Xu S, Yu L, Luo X, et al. Manual acupuncture versus sham acupuncture and usual care for prophylaxis of episodic migraine without aura: multicentre, randomised clinical trial. BMJ. 2020 Mar 25;368:m697.
3. Streitberger K, Kleinhenz J. Introducing a placebo needle into acupuncture research. Lancet. 1998 Aug 1;352(9125):364-5.
4. Hui KK, Nixon EE, Vangel MG, et al. Characterization of the "deqi" response in acupuncture. BMC Complement Altern Med. 2007 Oct 31;7:33.
5. Edzard Ernst. Does acupuncture prevent migraine attacks? Published Monday 30 March 2020 [accessed May 3, 2020]. Available from: https://edzardernst.com/2020/03/does-acupuncture-prevent-migraine-attacks/
Competing interests: No competing interests
Dear Editor
Xu and colleagues show that manual acupuncture is more effective than sham acupuncture and usual care for the prophylaxis of episodic migraine without aura. The findings provide a recommendable non-pharmaceutical therapy to the migraine sufferers. The strengths of this trial are the success of blinding the participants, the use of non-penetrating sham acupuncture at non-acupoints as control to minimise the physiological effect and the low dropout rate. Although the trial is well-designed, we still have several issues.
The authors formulated standardised step-by-step instructions and operations to use the same rituals in the manual acupuncture and sham acupuncture groups, which contributed to the successful blinding. However, the numbers of stimulation points between groups were different in this trial. Ten obligatory acupoints and 1-2 additional acupoints were punctured in the manual acupuncture group, while only 8 non-acupoints were stimulated in the sham acupuncture group. It introduces the possibility that more acupoints in the manual acupuncture group contributed to the positive result. In addition, several secondary outcomes were reported to be measured per four weeks during weeks 1-20 in both protocol and registration system, while only the data from weeks 17 to 20 were presented in the article. The trend through study period may be more visualized and can help the readers get a fuller picture of the results.
A dose-response relationship was found when acupuncture was used for pain conditions. In this trial, participants received 20 sessions during 8 weeks and manual manipulation for each acupoint was repeated four times. Hence, the dose of manual acupuncture was adequate. However, the manual acupuncture requires a substantial commitment of time from the patients. The time necessary for frequent transportation is quite a barrier, especially for the employed patients. Furthermore, a number of patients hesitate to receive acupuncture treatment due to a phobia of penetrating the skin. The noninvasive acupoint stimulator, which can be manipulated by the patients themselves at home after the training by acupuncturists, may be an alternative option for chronic diseases.
Competing interests: No competing interests
Dear Editor,
Just wondering whether there acupuncturists delivering the treatments had equipoise? This is why we have double-blind studies.
As far as I can tell, the patients were only asked whether they thought the needles were penetrating or not, not whether they received effective treatment or not...
Competing interests: No competing interests
Dear Editor
Two important aspects of this study were that 20 treatments were administered over 8 weeks and that the real acupuncture did not clearly surpass the sham acupuncture until week 13 and especially at week 20.
Many trials done on acupuncture use far fewer treatments and don't usually measure outcomes past 8-10 weeks. This could mean that many trails on acupuncture using sub-optimal dosing and not measuring longer-term outcomes may find real and sham acupuncture to have equal effectiveness - a possible false-negative.
For more on the subject of false-negatives relative to acupuncture treatment dose and the implications on research and clinical findings, see:
"Is acupuncture dose dependent? Ramifications of acupuncture treatment dose within clinical practice and trials" March 2020 issue Journal of Integrative Medicine Research.
https://www.sciencedirect.com/science/article/pii/S2213422020300032
Possible Conflict of Interest: I am co-author of the paper sited here.
Competing interests: No competing interests
Dear editor
Extraordinary claims need extraordinary levels of evidence to change practice. Despite the very positive editorial article, I don’t feel that this hits that bar. The detail on the method is a little difficult to find (I finally found it in one of the appendices online). That method appears to describe two very different groups. The “True” acupuncture has 10 sites which would be perceived by the patient as quite invasive In sensitive parts of the body- face, wrist etc, plus a further two as decided by the acupuncturist depending on symptoms. The “Sham” only had 8 points which were on the patient’s back. I think the patients belief in the process would have been significantly different between the two arms based on the striking differences between the “true” and “sham” groups. I don’t think you can really claim that this therefore provides “good evidence” It appears to me that this can be all explained by a more elaborate placebo effect
Competing interests: No competing interests
Re: Manual acupuncture versus sham acupuncture and usual care for prophylaxis of episodic migraine without aura: multicentre, randomised clinical trial
Dear Editor
This trial was actually measuring the difference in the strengths of the placebo effects when similar treatments (needles) are administered by those who believe in it (acupuncturists performing acupuncture) and those who do not (acupuncturists using sham needles in the "wrong" places).
To have avoided confounding of variables, it would have been more informative to have used the same number of sites in both groups with real needles - but with that design there might have been a problem recruiting acupuncturists.
Competing interests: No competing interests