Do the effects of acupuncture vary between acupuncturists? Analysis of the Acupuncture Trialists' Collaboration individual patient data meta-analysis
Vickers et al. · Acupuncture in Medicine · 2021
Evidence Level
STRONGOBJECTIVE
Evaluate whether the effects of acupuncture vary between different acupuncturists
WHO
1,206 acupuncturists across 13 studies treating chronic pain
DURATION
Data from studies conducted through December 2015
POINTS
Varied according to each acupuncturist's style and approach
🔬 Study Design
True acupuncture
n=10000
Acupuncture by different practitioners
Control (sham or usual care)
n=10000
Sham acupuncture or standard care
📊 Results in numbers
Statistically significant heterogeneity between acupuncturists
Acupuncturists with results half a standard deviation above/below the mean
Expected by chance
Difference observed vs. expected
Percentage highlights
📊 Outcome Comparison
Variability between practitioners
This study investigated whether different acupuncturists obtain different results in the treatment of chronic pain. Although it found some variations between practitioners, these differences were very small, suggesting that patients can expect similar results regardless of the acupuncturist they choose.
Article summary
Plain-language narrative summary
This pioneering study investigated a fundamental question in acupuncture practice: do different acupuncturists obtain significantly different results in the treatment of chronic pain? The researchers analyzed individual patient data from 39 high-quality clinical studies, involving 1,206 acupuncturists across 13 eligible studies. The methodology used two complementary approaches: a study-level analysis and an acupuncturist-level analysis, comparing results against both sham acupuncture and usual care. The results revealed that, although there are statistically significant differences between acupuncturists (p ≤ 0.003 for sham-controlled groups and p ≤ 0.001 for non-acupuncture controls), the degree of heterogeneity was surprisingly small.
The observed distribution of treatment effects practically overlapped with that expected by chance. For example, in the non-acupuncture controlled studies, 34% of acupuncturists were expected to have effect sizes half a standard deviation above or below the mean, but only 37% were observed — a minimal difference of 3%. This finding has important clinical implications, suggesting that patients with chronic pain in clinical practice should expect results similar to those reported in high-quality studies, regardless of the specific acupuncturist. In addition, the findings do not provide evidence for a need for greater standardization of acupuncture techniques.
An important limitation is that the study included a relatively limited group of acupuncturists, predominantly German physicians with at least 180 hours of training, and is not fully representative of the diversity found in routine practice. Many acupuncturists were actively selected by the researchers based on reputation or clinical experience, potentially reducing heterogeneity. The study contributes significantly to the understanding of inter-practitioner variability in complementary medicine, demonstrating that, unlike some surgical interventions where large variations between practitioners are common, acupuncture shows remarkable consistency of results across different qualified practitioners.
Strengths
- 1Individual patient data meta-analysis of high quality
- 2Large sample of acupuncturists and patients
- 3Robust statistical methodology with two complementary approaches
- 4Clear and practical clinical implications
Limitations
- 1Sample limited primarily to German physician acupuncturists
- 2Active selection of experienced acupuncturists by researchers
- 3Does not fully represent the diversity of clinical practice worldwide
- 4Few data from non-physician acupuncturists
Expert Commentary
Dr. Marcus Yu Bin Pai
MD, PhD · Pain Medicine · Physical Medicine and Rehabilitation · Medical Acupuncture
▸ Clinical Relevance
The question of operator variability is central to any technical intervention, and acupuncture does not escape this discussion. This work, by pooling individual data from approximately 20,000 patients distributed between true acupuncture and controls, offers the largest empirical base available to answer whether the specific acupuncturist determines clinical outcomes. The practical answer is reassuring: heterogeneity between practitioners, although statistically detectable, is clinically negligible. For physicians referring patients with chronic low back pain, osteoarthritis, or tension-type headache, this means that the choice of acupuncturist does not need to be the determining factor in the decision. The results observed in high-quality randomized controlled trials are generalizable to clinical practice, which strengthens the external validity of existing studies and supports the incorporation of acupuncture into multidisciplinary chronic pain protocols.
▸ Notable Findings
The most striking finding of this work is not the statistical significance of heterogeneity between acupuncturists — which does indeed exist, with p ≤ 0.003 — but rather its tiny clinical magnitude. By chance, 34% of acupuncturists were expected to fall half a standard deviation above or below the mean of results; 37% were observed. This 3 percentage point difference is what separates real variation from random variation, and it is clinically irrelevant. The parallel with surgery is instructive: in operative procedures, the learning curve and surgeon experience often explain much greater variations in outcome. In acupuncture performed by trained physicians, this phenomenon simply does not reproduce on the same scale. This suggests that the therapeutic effect of acupuncture depends more on the intervention itself than on idiosyncratic operator characteristics, which has direct implications for how we interpret clinical trials on the topic.
▸ From My Experience
In my musculoskeletal pain clinic practice, this question about variation between acupuncturists comes up constantly — both from residents and from patients who arrive frustrated after previous treatments. What I have observed over the years is consistent with what this work demonstrates: when the physician has solid training and applies points appropriate to the clinical picture, results are quite reproducible across different operators. I usually see noticeable clinical response between the third and fifth session in chronic low back pain and knee osteoarthritis, and we generally work with cycles of 8 to 12 sessions before reassessing. I routinely combine acupuncture with supervised exercise and, when indicated, with dry needling of myofascial trigger points — the sum of effects is clinically and mechanistically coherent. The profile that responds best, in my experience, is the patient with predominantly non-nociceptive chronic pain, without severe central component, and with good adherence to the overall treatment.
Full original article
Read the full scientific study
Acupuncture in Medicine · 2021
DOI: 10.1177/0964528420959089
Access original articleScientific Review
Marcus Yu Bin Pai, MD, PhD
CRM-SP: 158074 | RQE: 65523 · 65524 · 655241
PhD in Health Sciences, University of São Paulo. Board-certified in Pain Medicine, Physical Medicine and Rehabilitation, and Medical Acupuncture. Scientific review and curation of every entry in this library.
Learn more about the author →Medical disclaimer: This content is for educational purposes only and does not replace consultation, diagnosis, or treatment by a qualified professional. Some information may be assisted by artificial intelligence and is subject to inaccuracies. Always consult a physician.
Content reviewed by the medical team at CEIMEC — Integrated Centre for Chinese Medicine Studies, a reference in Medical Acupuncture for over 30 years.
Related articles
Based on this article’s categories