Characteristics of Acupuncture Treatment Associated with Outcome: An Individual Patient Meta-Analysis of 17,922 Patients with Chronic Pain in Randomised Controlled Trials

Chronic pain affects millions of people worldwide, representing one of the leading public health problems and substantially affecting quality of life. Among available treatment options, acupuncture has become an important therapeutic alternative with growing scientific evidence of efficacy. However, one fundamental question remained inadequately answered: are there specific characteristics of acupuncture or of acupuncturists associated with better or worse outcomes in the treatment of chronic pain? This question is particularly relevant because acupuncture is not a standardized intervention — different practitioners may use varied techniques on the same patient, including different acupuncture styles, numbers of sessions, treatment frequencies, and levels of practitioner experience.

To address this complex question, investigators from the Acupuncture Trialists' Collaboration conducted a comprehensive study analyzing individual patient data from 29 randomized controlled trials involving 17,922 patients with chronic pain. The primary aim was to identify whether specific characteristics of acupuncture or of acupuncturists modified the treatment effects on pain outcomes. The methodology was an individual patient data meta-analysis, considered one of the most robust approaches in medical research. The investigators evaluated multiple acupuncture characteristics, including the style used (traditional Chinese versus Western), point prescription (fixed, flexible, or individualized), needle location, use of electrical stimulation and moxibustion, number and frequency of sessions, treatment duration, number of needles used, and acupuncturists' experience.

The included trials covered four chronic pain conditions: osteoarthritis, headache, back and neck pain, and shoulder pain. Comparisons were made between acupuncture and sham controls (sham acupuncture) and between acupuncture and non-acupuncture controls (such as waiting list or usual care).

The main results revealed a surprising finding: there was little evidence that the different characteristics of acupuncture or of acupuncturists significantly modified treatment effects on pain. When acupuncture was compared with sham controls, none of the characteristics evaluated showed a statistically significant association with better or worse outcomes. This included aspects that many practitioners and patients considered important, such as the style of acupuncture (traditional Chinese versus Western), the number or location of needles, the frequency or duration of sessions, the specific patient-practitioner interaction, and even the acupuncturist's experience. When acupuncture was compared with non-acupuncture controls, the investigators found limited evidence that these characteristics modified treatment effect.

The two notable exceptions were that better pain outcomes were observed when more needles were used and when a greater number of treatment sessions were provided. These findings suggest that the "dose" of acupuncture — both in terms of intensity (number of needles) and duration (number of sessions) — may be important for optimizing treatment outcomes.

For patients considering acupuncture treatment, these results carry important and reassuring practical implications. First, there is no evidence that patients receive inferior treatment because of the diversity of practices among different acupuncturists. This means patients can be confident that, regardless of whether the practitioner uses a more traditional Chinese or a more Western approach, the results are likely to be similar. For clinicians, the study suggests there is no need to prescribe a specific style of acupuncture or to seek acupuncturists with a specific number of years of experience.

The findings do indicate, however, that it may be beneficial to consider treatments with a greater number of sessions when clinically appropriate. For acupuncturists, the results support flexibility in clinical practice and suggest that different approaches may be equally valid. The study also offers reassurance that there is no evidence that consensus methods used to determine acupuncture protocols in research systematically underestimate treatment effects.

As with any scientific study, this research has important limitations that should be considered. The total number of included trials was relatively modest, and analyses with individual patient data included at most five trials for some characteristics. This resulted in relatively low statistical power for some analyses, with wide confidence intervals around the central estimates. In addition, heterogeneity of treatment characteristics was limited — for example, nearly 75% of trials involved between 6 and 15 treatments, and no trial administered acupuncture more than twice per week.

The investigators also conducted a large number of statistical tests, which should be considered in interpreting the results. Importantly, future trials designed to assess potentially small differences in outcomes associated with different acupuncture characteristics will likely require very large sample sizes to detect such effects. The findings suggest there is room for diversity in acupuncture practice, without solid evidence that such diversity leads some patients to suboptimal outcomes, offering a balanced and scientifically grounded perspective on the efficacy of acupuncture for chronic pain.