Clinical Trial Methodology
Main result
38.7%
Rate of adequate PRO reporting
BMJ Evidence-Based Medicine · Duan et al. · 2026 · 476 participants
Patient-reported outcomes and acupuncture-related adverse events are overlooked in acupuncture randomised controlled trials: a cross-sectional meta-epidemiological study
Weak
Limited
Moderate
Strong
- Studies with PROs296 participants
- Studies with adverse events272 participants
42.9 %
Studies that omitted safety outcomes
1,277
Total adverse events identified
Study design sheet
- Cross-sectional Meta-epidemiological Study
- n=476 studies analyzed
- High Impact — Reporting Quality
- Participants
- 476
- Duration
Cross-sectional analysis of data published over 10 years
Retrospective analysis of 10 years of publications (2014–2024)
- Groups
- Studies with PROs (296)Trials that used patient-reported outcomes
- Studies with adverse events (272)Trials that reported safety data
- OBJECTIVE
- Investigate how patient-reported outcomes (PROs) and acupuncture-related adverse events are reported in randomized clinical trials
- WHO
- 476 randomized acupuncture clinical trials published between 2014 and 2024
- POINTS
- Studies based on traditional medicine with skin-penetrating needling
- Results in numbers
- 38.7 %Rate of adequate PRO reporting
- 42.9 %Studies that omitted safety outcomes
- 1,277Total adverse events identified
- 1.434OR for adverse events (acupuncture vs. control)
- 0 %Serious adverse events reported
- Outcome comparison
- PRO reporting quality (%)
- English-language publications: 50.4 of 100
- Chinese-language publications: 28 of 100
- PRO reporting quality, by registration (%)
- Studies with registration: 50.4 of 100
- Studies without registration: 30.8 of 100
- PRO reporting quality (%)
What does this mean for you?
This study showed that many studies on acupuncture do not adequately report outcomes important to patients or possible side effects. Although serious events are rare, minor effects such as pain and bleeding at the needle site can occur, but they are not always properly documented in studies.
Article summary
Plain-language narrative summary
This cross-sectional meta-epidemiological study analyzed 476 randomized clinical trials (RCTs) of acupuncture published between 2014 and 2024, investigating the quality of reporting of patient-reported outcomes (PROs) and acupuncture-related adverse events. The research was motivated by the need to understand whether acupuncture studies provide sufficient information for patients to make informed decisions about the benefits and risks of the treatment.
The methodology involved a comprehensive search of seven databases (MEDLINE, EMBASE, CENTRAL, CBM, CNKI, Wanfang Data, and VIP Database), including studies that used traditional medicine-based acupuncture as the intervention. The researchers assessed PRO reporting using the CONSORT PRO extension and developed specific criteria to identify acupuncture-related adverse events.
The results revealed significant deficiencies in the reporting of both outcomes. Of the 476 included studies, 296 (62.2%) used PROs as primary or secondary outcomes, a rate higher than in other healthcare fields. The Visual Analog Scale (149 studies, 23.7%) and the Pittsburgh Sleep Quality Index (42 studies, 6.7%) were the most used PROs. However, the quality of reporting was insufficient, with adequate reporting scores ranging from 2.7% to 97.6% across the different items of the CONSORT PRO extension.
The mean score for sufficient reporting was only 38.7%.
Regarding adverse events, 272 studies (57.1%) reported safety outcomes, but 204 (42.9%) completely omitted this information. Among those that omitted it, 41 studies still asserted that acupuncture was safe. A total of 1,277 acupuncture-related adverse events were identified, with tissue injury (hematoma, bleeding) being the most frequent type, followed by tissue irritation (post-acupuncture pain and discomfort) and systemic reactions. Importantly, no serious adverse events were reported.
The analysis revealed that acupuncture had an odds ratio of 1.434 for adverse events compared with the control group, with a higher risk compared with simulated acupuncture (OR = 1.708) and other controls (OR = 4.105), but with no significant difference compared with medications or other active treatments. Adverse event reporting was incomplete: 84.5% of studies did not specify who reported the events, 94.5% did not mention the time of occurrence, and 81.9% did not describe the management of the events.
The study identified factors associated with better adverse event reporting, including presence of funding, registration information, acupuncturist qualifications, and non-significant primary outcomes. English-language studies had better reporting quality than Chinese-language studies, and registered studies outperformed non-registered ones.
The clinical implications are important. Inadequate PRO reporting limits guidance for clinical practice, since researchers may use these outcomes to appear more comprehensive rather than considering their practical implications. Under-reporting of adverse events may be related to the difficulty of reporting harms compared with benefits, unintentional omission because they were not prospectively defined, treating minor events as trivial, or confusion with symptoms of the treated condition.
The study has limitations, including possible subjective bias in identifying adverse events, restriction to Chinese- and English-language publications, and inclusion only of two-parallel-group studies. Its strengths include the absence of restrictions by disease type or journal, comprehensive exploration of PROs, and the first comprehensive synthesis of adverse events in acupuncture RCTs.
Strengths
- First comprehensive analysis of PROs and adverse events in acupuncture RCTs
- Large sample of 476 studies with no disease or journal restrictions
- Use of standardized guidelines (CONSORT PRO Extension) for evaluation
- Specific identification of adverse events based on prior knowledge
- Analysis of confounding factors that affect adverse event reporting
Limitations
- Possible subjective bias in identifying adverse events
- Restriction to studies published in English and Chinese only
- Limited to studies with two parallel groups
- Difficulty distinguishing adverse events from symptoms of the treated condition
- Data based only on published information, not on original data

Expert Commentary
Clinical Relevance
For those who prescribe acupuncture in the context of pain and rehabilitation services, this work resolves a recurring practical question: how usable the RCT literature is to guide the risk-benefit conversation with the patient. The answer, based on 476 trials over a decade, is that it is not yet satisfactorily so. With only 38.7% adequacy in the reporting of patient-reported outcomes and 42.9% of studies omitting adverse events entirely, the physician who tries to ground informed consent in the published evidence encounters structural gaps, not merely isolated ones. The clinically relevant finding is that, among the 1,277 adverse events identified, none was serious, and the effect profile consists predominantly of hematoma, local bleeding, and post-needling discomfort — a finding that strengthens the safety conversation with hesitant patients, provided it is communicated with precision.
Notable Findings
The odds ratio of 1.434 for adverse events of acupuncture versus control is the most informative number in the study. It rises to 1.708 when the comparison is made against simulated acupuncture — which makes pathophysiological sense, since real needle penetration produces measurable tissue microtrauma — and reaches 4.105 when the control is another non-active comparator. Against medications or active treatments, the difference does not reach statistical significance, which positions acupuncture within a safety spectrum comparable to conventional interventions in routine use. Another finding worth noting is that 41 studies asserted that acupuncture is safe without reporting any adverse event data — which is not an assertion of safety, but an absence of data. PRO reporting quality varied between 2.7% and 97.6% across the different CONSORT PRO items, indicating that heterogeneity is extreme and prevents reliable synthesis.
From My Experience
In my practice in the musculoskeletal pain clinic, the events this study categorizes as 'tissue injury' — local hematomas and bleeding — are exactly what we most often report in the chart and what most concerns the patient in the first session. I have observed that when the physician anticipates these possibilities in the initial consultation, the rate of early dropout drops visibly. For the patient profile that responds best — localized chronic pain, low tolerance to systemic medications, motivation for multimodal treatment — I typically see measurable clinical response between the third and fifth session, and I structure protocols of eight to twelve sessions before reassessing the indication. What this study reinforces for me is the need to prospectively document adverse events in all cases, something we institutionalized in our service a few years ago precisely because the published literature did not offer that reference with sufficient clarity.
Dr. Marcus Yu Bin Pai · MD, PhD · Pain Medicine · Physical Medicine and Rehabilitation · Medical Acupuncture
PhD in Health Sciences, University of São Paulo. Board-certified in Pain Medicine, Physical Medicine and Rehabilitation, and Medical Acupuncture.
Full original article
Read the full scientific study
BMJ Evidence-Based Medicine · 2026
DOI: 10.1136/bmjebm-2024-113497
Scientific review
Dr. Marcus Yu Bin PaiScientific lead · CRM-SP 158074 · RQE 65523 · 65524 · 655241
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.
- Strong
- Moderate
- Limited
- Weak
See the numbers: This study among the Library’s studies
| Study design | Clinical Trial Methodology | All studies |
|---|---|---|
| Meta-analysis · Systematic Review | 38 | 387 |
| RCT | 46 | 224 |
| Clinical Guideline | 2 | 5 |
| Observational | 5 | 38 |
| Case Series | 0 | 7 |
| Narrative Review | 10 | 155 |
| Preclinical / Animal | 0 | 6 |
| Other | 56 | 203 |
Related articles
Based on this article’s categories
Kim et al. · BMC Complementary and Alternative Medicine · 2012
- Systematic Reviews & Meta-Analyses
- Safety & Adverse Events
- Clinical Trial Methodology
Vickers et al. · Archives of Internal Medicine · 2012
- Systematic Reviews & Meta-Analyses
- Chronic Pain
- Clinical Trial Methodology
Foster et al. · Acupuncture in Medicine · 2021
- Systematic Reviews & Meta-Analyses
- Chronic Pain
- Clinical Trial Methodology
Li et al. · BMC Complementary Medicine and Therapies · 2026
- Headache & Migraine
- Clinical Trial Methodology
- Systematic Reviews & Meta-Analyses