Another dodgy meta-analysis published in a dodgy journal (except ‘Acupuncture in Medicine’ is from the BMJ-group)

The plethora of dodgy meta-analyses in alternative medicine has been the subject of a recent post – so this one is a mere update of a regular lament.

This new meta-analysis was to evaluate evidence for the effectiveness of acupuncture in the treatment of lumbar disc herniation (LDH). (Call me pedantic, but I prefer meta-analyses that evaluate the evidence FOR AND AGAINST a therapy.) Electronic databases were searched to identify RCTs of acupuncture for LDH, and 30 RCTs involving 3503 participants were included; 29 were published in Chinese and one in English, and all trialists were Chinese.

The results showed that acupuncture had a higher total effective rate than lumbar traction, ibuprofen, diclofenac sodium and meloxicam. Acupuncture was also superior to lumbar traction and diclofenac sodium in terms of pain measured with visual analogue scales (VAS). The total effective rate in 5 trials was greater for acupuncture than for mannitol plus dexamethasone and mecobalamin, ibuprofen plus fugui gutong capsule, loxoprofen, mannitol plus dexamethasone and huoxue zhitong decoction, respectively. Two trials showed a superior effect of acupuncture in VAS scores compared with ibuprofen or mannitol plus dexamethasone, respectively.

The authors from the College of Traditional Chinese Medicine, Jinan University, Guangzhou, Guangdong, China, concluded that acupuncture showed a more favourable effect in the treatment of LDH than lumbar traction, ibuprofen, diclofenac sodium, meloxicam, mannitol plus dexamethasone and mecobalamin, fugui gutong capsule plus ibuprofen, mannitol plus dexamethasone, loxoprofen and huoxue zhitong decoction. However, further rigorously designed, large-scale RCTs are needed to confirm these findings.

Why do I call this meta-analysis ‘dodgy’? I have several reasons, 10 to be exact:

1. There is no plausible mechanism by which acupuncture might cure LDH.
2. The types of acupuncture used in these trials was far from uniform and  included manual acupuncture (MA) in 13 studies, electro-acupuncture (EA) in 10 studies, and warm needle acupuncture (WNA) in 7 studies. Arguably, these are different interventions that cannot be lumped together.
3. The trials were mostly of very poor quality, as depicted in the table above. For instance, 18 studies failed to mention the methods used for randomisation. I have previously shown that some Chinese studies use the terms ‘randomisation’ and ‘RCT’ even in the absence of a control group.
4. None of the trials made any attempt to control for placebo effects.
5. None of the trials were conducted against sham acupuncture.
6. Only 10 studies 10 trials reported dropouts or withdrawals.
7. Only two trials reported adverse reactions.
8. None of these shortcomings were critically discussed in the paper.
9. Despite their affiliation, the authors state that they have no conflicts of interest.
10. All trials were conducted in China, and, on this blog, we have discussed repeatedly that acupuncture trials from China never report negative results.

And why do I find the journal ‘dodgy’?

Because any journal that publishes such a paper is likely to be sub-standard. In the case of ‘Acupuncture in Medicine’, the official journal of the British Medical Acupuncture Society, I see such appalling articles published far too frequently to believe that the present paper is just a regrettable, one-off mistake. What makes this issue particularly embarrassing is, of course, the fact that the journal belongs to the BMJ group.

… but we never really thought that science publishing was about anything other than money, did we?