Madsen, Gøtzsche, Hróbjartsson. Acupuncture treatment for pain: systematic review of randomised clinical trials with acupuncture, placebo acupuncture, and no acupuncture groups. BMJ. 2009.
https://pubmed.ncbi.nlm.nih.gov/19174438/
Conclusion: “A small analgesic effect of acupuncture was found, which seems to lack clinical relevance and cannot be clearly distinguished from bias. Whether needling at acupuncture points, or at any site, reduces pain independently of the psychological impact of the treatment ritual is unclear.”
Methodology:
Databases: Cochrane Library, Medline, Embase, Biological Abstracts, PsycLIT (all trials published before Jan 1, 2008).
Included trials that labelled intervention “acupuncture” (traditional, electro-acupuncture).
Excluded transcutaneous electrical nerve stimulation and manual acupressure.
Placebo = insertion of needles into non-acupuncture points or use of non-penetrating needles.
Pain self-reported by patients on a visual analogue scale or another ranking system.
Comparison made to first time point after end of treatment.
Data Analysis:
Standardized mean difference: difference between the means divided by the pooled std. deviation.
In applicable cases, combined results multiple acupuncture groups (low and high frequency) into a weighted mean and a pooled variance.
Pooled standardized mean differences from the trials, compared effect of acupuncture with that of placebo acupuncture and the effect of placebo with that of no acupuncture.
Placebo interventions ranked from 1-5, where 1 represented a placebo treatment that most likely could produce physiological effects. (factors: point of insertion, needle size, depth of insertion, penetration of skin, achievement of Qi, manual stimulation)
Results:
Total 13 eligible trials of acupuncture for pain (3025 patients).
Clinical conditions: knee osteoarthritis, tension type headache, migraine, low back pain, fibromyalgia, abdominal scar pain, postoperative pain, procedural pain during colonoscopy.
Tx duration: 1 day ~ 12 weeks.
Placebo: 2/13 non-penetrative needling. 7/13 superficial needling at non-acu-points with fine needles, 4/13 other forms for penetrative needling.
P<0.001 significant difference between acupuncture and placebo pool standardized mean difference.
P<0.001 significant difference between placebo and no acupuncture.
Discussion
Small difference between acupuncture and placebo, moderate difference between placebo and no acupuncture. Placebo effect varied considerably.
However, the magnitude of pain reduction comparing acupuncture and placebo is below a clinically relevant pain improvement.
Weakness:
Trials had different primary outcomes (e.g. days with headache, # of analgesic doses).
Heterogeneity of results “remain unexplained and calls for further studies on the underlying mechanisms of the effects of placebo acupuncture”.
Future Suggestions:
Try to reduce bias by ensuring blinding when possible. E.g. have the needling done by acupuncture naïve clinicians blinded to the hypothesis.
Try to separate the effects involved: physiological effect of needling at acupuncture sites or at other sites and the psychological effect of treatment ritual or patient-provider interaction.