Traditional medicine got WHO codes, and testing it fairly is the harder part

Key takeaways

• Since 2022, the World Health Organization’s disease classification has included an optional chapter for traditional medicine diagnoses, but WHO says the codes endorse no practice.
• WHO published a research agenda for 2025 to 2034 in September 2026, noting that traditional and integrative medicine gets less than 1% of global health research funding.
• In a pooled analysis of 20,827 patients, acupuncture beat no acupuncture for chronic pain by a moderate margin and beat sham needling by a small one.
• When researchers phoned the authors of 2,235 “randomized” trials from Chinese journals, only 6.8% had truly randomized, and traditional and conventional medicine trials did about equally badly.
• Before trusting a traditional-medicine study, check how patients were assigned, what the comparison group got, and whether the exact herbal product was named and tested.

In a study published in 2009, a team at the Chinese Cochrane Centre did something few journal reviewers ever do. They picked up the phone¹.

They had found 3,137 reports in Chinese medical journals, published from 1994 to 2005, that called themselves randomized controlled trials¹. In that kind of study, chance alone decides who gets a treatment, which keeps hopes and hunches from tilting the results. The team then interviewed the authors of 2,235 of those reports about how they had assigned patients¹.

Only 6.8% had truly randomized¹. Most of the authors who had used the label wrongly, 85.6% of them, did not fully understand what randomization requires¹.

Here is the detail that often gets left out. Trials of traditional Chinese medicine were no less authentic than the conventional medicine trials in the sample¹. The weak spot was research training, not the tradition.

That study still frames the big question. Traditional medicine is now a global research priority. Can the research get good enough to tell you what works?

WHO now gives traditional medicine a center, summits and its own codes

The World Health Organization (WHO) set up its Global Traditional Medicine Centre in 2022, with founding support from India’s government². It is based in Jamnagar, in the Indian state of Gujarat². Our earlier article on traditional knowledge covers the funding and ownership debates around it.

In August 2023, WHO and India held the first WHO Traditional Medicine Global Summit in Gandhinagar, during India’s G20 presidency³. It produced an action agenda called the Gujarat Declaration⁴. A second summit followed on December 17–19, 2025⁴.

WHO says 170 of its 194 member states report use of practices such as herbal medicines, acupuncture, yoga and Indigenous therapies³. At the 2023 summit, John Reeder, who directs a WHO research program on tropical diseases, said this science “should be held to the same rigorous standards as in other fields of health”³.

The quieter change came in medical records. WHO’s newest disease classification, ICD-11, took effect on January 1, 2022, with a traditional medicine chapter included³. ICD is the coding system hospitals, insurers and health agencies use to record diagnoses.

That chapter, Chapter 26, lists conditions that began in ancient Chinese medicine and are now used in China, Japan, Korea and elsewhere⁵⁶. A second module, for Ayurveda, Siddha and Unani diagnoses, has been in the works since at least 2023³.

Its status depends on who is speaking. In February 2025, India’s Ministry of Ayush said WHO had added that module to its 2025 ICD-11 update, a year after it was released for testing in January 2024⁷. WHO’s own ICD-11 questions page, checked in October 2026, still says only that development of the module has started⁵.

A code is a label, not a verdict

A traditional diagnosis can now be recorded next to a conventional one. This is called dual coding. WHO says the chapter is optional, is meant for tracking illness and care, and is never used to record causes of death⁵.

WHO is also clear about what inclusion doesn’t mean. The chapter does not judge or endorse the scientific validity of any traditional practice, and a listed condition says nothing about whether any treatment for it works⁵.

Critics worried anyway. When Nature reported the plan in 2018, scientists it quoted, including Oxford epidemiologist Richard Peto, questioned whether WHO still stood behind evidence-based medicine⁶. Ryan Abbott, a physician on the UCLA faculty, told Nature the listing would help providers convince insurers that traditional care is legitimate medicine⁶.

Both sides have a point. WHO argues that standard coding is a precondition for managing traditional medicine inside health systems³. Codes let researchers count who uses traditional care and what happens to them next.

But a code only names a condition. It can’t show that a remedy helps.

A new research agenda arrives with a funding gap

On September 16, 2026, WHO published global research priorities for traditional, complementary and integrative medicine for 2025 to 2034⁸. The agenda supports WHO’s 10-year traditional medicine strategy, which our traditional knowledge article describes.

Its starting point is blunt. Billions of people use these practices, WHO says, yet they receive less than 1% of global health research funding⁸. And the studies that do get done have not consistently matched what policymakers need to know, which limits the evidence on safety, regulation, service delivery and paying for care⁸.

So the agenda names priority conditions, groups of people and cross-sector issues where studies would do the most good⁸. WHO built it through consultation in all six of its regions, and it is aimed at governments, funders and researchers⁸.

In the United States, the National Center for Complementary and Integrative Health (NCCIH), part of the National Institutes of Health, sets research goals in multiyear plans. Its most recent plan, for 2021 to 2025, called for studying treatments that combine many parts as whole packages, then testing which parts are necessary⁹.

The plan also favored pragmatic trials, which test care the way it is delivered in ordinary clinics⁹. And it backed treatment manuals written by practitioners and researchers together, so that different clinics deliver the same care⁹.

Four reasons traditional medicine is hard to test

A drug trial is simple in concept. Everyone gets the same pill or an identical dummy, and nobody knows which. Traditional systems often break all three of those rules.

Treatment is personal. A practitioner of Chinese medicine may diagnose a “pattern” and adjust a herbal formula for each patient. A 2017 reporting standard for Chinese herbal formula trials treats fixed, individualized and ready-made commercial formulas as separate cases, each needing its own details¹⁰.

Blinding is hard. You can’t easily hide whether a needle went in. Researchers use sham acupuncture instead, such as devices that poke the skin without piercing it¹¹. But a sham session still involves touch, attention and ritual, so it may not be inert.

Products vary. Herbs differ by species, plant part and processing. The 2017 standard asks trials to report each herb’s name in at least two languages, its source, part used, processing and dose, plus any testing for heavy metals, pesticides and microbes¹⁰.

The trial may not match the tradition. A 2024 study in Frontiers in Pharmacology reviewed 420 published plans for trials of Chinese herbal formulas, acupuncture and moxibustion, a practice that applies heat from burning herbs near the skin¹². Only 4.3% identified patients by traditional pattern diagnosis, and just 17.4% measured any outcome from traditional medicine¹².

That last point cuts both ways. A trial that strips a practice down may underrate it. A trial with no fair comparison may overrate it.

Acupuncture is the best-studied practice, and the comparison is everything

The strongest single analysis comes from the Acupuncture Trialists’ Collaboration, an international research group. In 2018, a team including Andrew Vickers of Memorial Sloan Kettering Cancer Center published an individual patient data meta-analysis in the Journal of Pain¹³.

It pooled 20,827 patients from 39 trials of acupuncture for back and neck pain, osteoarthritis, chronic headache and shoulder pain¹³. The team accepted only trials where the people enrolling patients could not know or steer the next assignment¹³.

Acupuncture beat both comparisons for every condition¹³. Against no acupuncture, the gain was moderate, about half of a standard deviation, a common statistical yardstick. Against sham acupuncture, it was small, about a fifth of a standard deviation¹³.

The benefit held up well, shrinking only about 15% a year later¹³. The authors concluded that placebo effects alone can’t explain the pain relief, and that referral for acupuncture is a reasonable option¹³.

They also found a telling pattern. Trials that used a skin-piercing needle as the sham showed smaller differences, and the comparison group’s treatment explained most of the variation between trials¹³.

NCCIH’s summary points the same way. Acupuncture often looks better against no treatment than against sham, which suggests expectations and the bond with the practitioner add to the benefit¹¹. For some uses, including IVF success, irritable bowel syndrome, carpal tunnel syndrome and hip osteoarthritis, real acupuncture showed little or no edge over sham¹¹.

Policy has followed the pain evidence. A 2017 American College of Physicians guideline lists acupuncture as a first-line option for chronic low back pain¹¹.

Since 2020, Medicare has covered up to 12 acupuncture visits for chronic low back pain, plus 8 more for patients who improve¹¹. Use by U.S. adults more than doubled, from 1.0% in 2002 to 2.2% in 2022¹¹.

Herbal trials have a quality problem, and fixes arrive slowly

The 2009 phone study also showed where solid trials came from. Every trial run to get a new drug onto the market was properly randomized¹. Trials at university-affiliated hospitals did far better than those at smaller hospitals¹.

Since then, researchers have written reporting standards. The main one, CONSORT, is the checklist journals use for trial reports. Versions for herbal and acupuncture trials already existed when a team added one for Chinese herbal formulas in 2017¹⁰. Writing in the Annals of Internal Medicine, its authors called the general reporting quality of these trials disappointing¹⁰.

Reporting is still weak. In the 2024 review of 420 trial plans, the average plan reported only 35.4% of the recommended items¹². Herbal formula plans did worst, at 20.4%, and only 2.1% of all plans described the funder’s role¹².

NCCIH’s summaries reflect the gap. Chinese herbal products have been studied for stroke, heart disease, mental disorders and respiratory illness, but many of those studies were too poor to support firm conclusions¹⁴. For Ayurveda, NCCIH says only a few well-designed trials and reviews suggest its approaches work¹⁵.

Some results are intriguing. A 2013 trial of 440 people with knee osteoarthritis found that two Ayurvedic plant-extract formulas eased pain about as well as glucosamine or the prescription pain reliever celecoxib¹⁵. One trial is a start, not a verdict.

Safety research matters as much as benefit. NCCIH reports that some Chinese herbal products have contained undeclared drugs such as the blood thinner warfarin, as well as heavy metals, pesticides or the wrong herb¹⁴. In a 2015 survey of Ayurvedic product users, 40% had high blood lead levels, and about one in four products tested contained high lead¹⁵.

What the evidence shows for headline practices

Here is how some of the most studied practices rate on our four-level evidence scale.

Practice and use	Evidence level	What the research shows
Acupuncture for chronic back, neck, shoulder, headache and osteoarthritis pain	Strong evidence	Moderate benefit versus no acupuncture, small benefit versus sham, lasting at least a year13
Tai chi for balance in older adults	Promising but early	NCCIH says it may improve balance and ease knee osteoarthritis pain14
Ayurvedic formulas for knee osteoarthritis	Promising but early	One 440-person trial found results similar to standard treatments15
Chinese herbal products for stroke, heart and lung disease	Not enough evidence	Many trials too flawed for firm conclusions14
Acupuncture for IVF success or irritable bowel syndrome	Evidence says no, or unsafe	No clear advantage over sham treatment11
Ayurvedic products not tested for heavy metals	Evidence says no, or unsafe	In a 2015 survey, 40% of users had high blood lead15

Eight questions to ask about any traditional-medicine study

Use this checklist when a headline, practitioner or product page cites a study.

1. Who decided which patients got treated? Look for a computer, a random-number table or a central office. Assigning patients by turns or by admission order is not randomization, and most self-described trials in the 2009 study fell short¹.
2. What did the comparison group get? Nothing, usual care, sham treatment or a drug? Comparisons against no treatment tend to look bigger¹³.
3. Was the exact product named? Look for species, plant part, preparation, dose and batch testing¹⁰. “A traditional formula” is not enough.
4. Was the trial registered before it began? A public registry such as ClinicalTrials.gov lets you check whether the goals changed midstream.
5. How many people, and for how long? A 30-person, four-week trial is a pilot.
6. Who paid, and did they help run it? Recent trial plans rarely said¹².
7. Did it measure what you care about? Pain, function and daily life matter more than a lab marker.
8. Has another team repeated it? One positive trial from one center is a lead, not proof.

Testing a tradition fairly is a way of honoring it

Many readers value traditional systems because they treat the whole person, take time to listen, and draw on generations of careful observation. Those are real strengths. Conventional medicine has started to borrow them.

Fair testing protects those strengths. A weak trial can make a useless remedy look good, which wastes money and can delay care that works. A careless trial can also make a useful practice look worthless, because it ignored how skilled practitioners work¹².

The best new research tries to do both jobs: deliver care the way it’s really practiced, use honest comparisons, and describe every product in full⁹¹⁰. That is what WHO’s own research leaders asked for³.

Some practical safety habits help while the science catches up:

• Tell every clinician and pharmacist about the herbs, formulas and treatments you use.
• Choose a licensed acupuncturist. The FDA requires acupuncture needles to be sterile and labeled for single use¹¹.
• Know that serious acupuncture injuries are rare but include infections and punctured organs, linked to nonsterile needles or poor technique¹⁴.
• Never drop a proven treatment for a traditional one without your clinician’s input.
• For a possible poisoning, call Poison Control at 1-800-222-1222. Report problems with herbal products to the FDA through MedWatch or the Safety Reporting Portal.

Questions readers ask

The bottom line

Traditional medicine has moved from the margins to WHO’s agenda, with its own center, summits, research priorities and diagnostic codes. None of that settles whether a given remedy works. Codes and strategies create the conditions for research, and the research itself is still catching up.

The evidence so far is mixed in a useful way. Acupuncture offers modest, lasting relief for several kinds of chronic pain, with a smaller edge over sham than many people expect. Most herbal formulas have not been tested well enough to judge. The fairest way to honor a tradition is to test it the way it is practiced, with honest comparisons, and to ask the same hard questions of every study you read.

Related reading

• An ancient recipe led to a Nobel prize, and new treaties ask who benefits
• Plants gave us Taxol and digoxin, yet the herb isn’t the drug
• Tai chi, prayer and journaling: what research can measure and what it can’t
• Who paid for the study matters, even when the science is real
• What complementary care means, and why the word misleads

Sources

1. Wu T, Li Y, Bian Z, Liu G, Moher D. “Randomized trials published in some Chinese journals: how many are randomized?” Trials, 2009. doi.org
2. World Health Organization, Global Traditional Medicine Centre. “Overview.” WHO, accessed October 2026. who.int
3. World Health Organization. “WHO convenes first high-level global summit on traditional medicine to explore evidence base, opportunities to accelerate health for all.” News release, August 10, 2023. who.int
4. World Health Organization, Global Traditional Medicine Centre. “Global Summit and Collaborations.” WHO, accessed October 2026. who.int
5. World Health Organization. “Frequently Asked Questions: Traditional Medicine in ICD-11.” WHO Classifications, accessed October 2026. who.int
6. Cyranoski D. “The big push for Chinese medicine.” Nature, September 27, 2018. media.nature.com
7. Press Trust of India. “WHO updates ICD framework, includes traditional medicine in health module.” Business Standard, February 19, 2025. business-standard.com
8. World Health Organization. “WHO global research priorities and agenda for traditional, complementary and integrative medicine 2025–2034.” WHO, September 16, 2026. who.int
9. National Center for Complementary and Integrative Health. “NCCIH Strategic Plan FY 2021–2025: Objective 1.” NIH, accessed October 2026. nccih.nih.gov
10. Cheng CW, Wu TX, Shang HC, Li YP, Altman DG, Moher D, Bian Z. “CONSORT extension for Chinese herbal medicine formulas 2017: recommendations, explanation, and elaboration.” Annals of Internal Medicine, 2017. scholars.hkbu.edu.hk
11. National Center for Complementary and Integrative Health. “Acupuncture: Effectiveness and Safety.” NIH, updated October 2022. nccih.nih.gov
12. Zhang L, Li H, Hu L, et al. “Reporting characteristics and quality of randomized controlled trial protocols in traditional Chinese medicine: a cross-sectional study.” Frontiers in Pharmacology, 2024. frontiersin.org
13. Vickers AJ, Vertosick EA, Lewith G, et al. “Acupuncture for chronic pain: update of an individual patient data meta-analysis.” Journal of Pain, 2018. pubmed.ncbi.nlm.nih.gov
14. National Center for Complementary and Integrative Health. “Traditional Chinese Medicine: What You Need To Know.” NIH, updated April 2019. nccih.nih.gov
15. National Center for Complementary and Integrative Health. “Ayurvedic Medicine: In Depth.” NIH, updated January 2019. nccih.nih.gov