What this covers, and what it does not
Knee and hip osteoarthritis are among the most common reasons patients ask about acupuncture in China, after back pain. The appeal is understandable: the drugs have limits, joint replacement is major surgery, and a course of needles sounds like a middle path. This article sets out what the trials show for knee and hip arthritis, what a course in a Chinese hospital involves and costs, and where acupuncture sits relative to injections, physiotherapy and replacement.
It does not diagnose your arthritis, does not tell you when to have surgery, and is not medical advice. Osteoarthritis affects the whole joint — cartilage, bone, ligaments, muscle and the pain system — and a treatment that makes you feel better for a few weeks is not the same as one that changes the disease.
What the trial evidence shows
Knee osteoarthritis: a small, short-term effect that fades
The 2022 Cochrane review by Manheimer and colleagues pooled 16 randomised trials with 3,498 people, twelve of them knee-only. Against sham acupuncture, the short-term pain difference was about 0.9 points on a 20-point scale — statistically significant, roughly a 5% absolute improvement, but below the review's own threshold for clinical relevance. Function showed a similarly small difference. Restricting the analysis to the sham controls judged most believable made the benefit smaller and no longer significant, and by 26 weeks the difference was under a point. Against a waiting list the effect looked clinically relevant, but that comparison mostly captures expectation and natural history.
Hip osteoarthritis: weaker still
A separate 2018 Cochrane review found only six trials with 413 people. In the two sham-controlled trials, 120 participants in total, acupuncture showed little or no difference from sham — about 2 points on a 100-point scale for pain and function. The unblinded trials that added acupuncture to routine care were more positive, but the reviewers concluded that the benefit was probably due at least partly to expectation. The hip review reported that no trial looked at radiographic joint changes, and the trials in both reviews measured symptoms, not structure. Nothing in this literature suggests acupuncture slows cartilage loss.
What the guidelines say, and why they disagree
NICE's 2022 osteoarthritis guideline recommends against acupuncture, citing lack of benefit and the potential for harm. The American College of Rheumatology and OARSI both conditionally recommend it as an adjunct for knee osteoarthritis — a weak recommendation in both cases. This is less a contradiction than a difference in how much weight to give small, short-term, hard-to-blind effects. The honest reading: a course may make your knee feel somewhat better for several weeks, and it will not change the joint.
The interventions, side by side
The table summarises the published evidence. It is not a recommendation for your case.
| Intervention | Evidence quality | Reasonable use | Where to be sceptical |
|---|---|---|---|
| Acupuncture, knee osteoarthritis | Moderate; small short-term effect over sham, below the clinical threshold | Short adjunct alongside exercise and weight management | Expecting cartilage repair or a lasting change |
| Acupuncture, hip osteoarthritis | Low; two small sham trials show little or no difference | Symptom trial if drugs are poorly tolerated | Reading unblinded add-on trials as proof |
| Electroacupuncture | Weaker; small trials, often single-country | Short trial within a broader plan | Large claims from small Chinese trials |
| Acupuncture plus exercise | Exercise has moderate evidence; the needle adds little | Exercise is the intervention that matters | Buying sessions instead of a walking programme |
| Herbal formulas or cartilage-regeneration claims | Not supported | None | Anything promising to regrow cartilage |
| Acupuncture instead of replacement in end-stage disease | No evidence; the harm is delay | Never | Clinics that discourage surgery |
| Acupuncture after knee or hip replacement | Very limited | Comfort during rehabilitation, with the surgical team's agreement | Needling near the wound or on blood thinners |
| Glucosamine, chondroitin and supplements | Weak and inconsistent | Patient choice, low risk | Any claim to slow the arthritis |
What a course in China involves
Assessment first: the physician should examine the joint, look at your X-rays, and ask how the pain behaves — mechanical, rest pain, night pain, swelling. A course is typically 10–20 sessions over 2–4 weeks, three to five times a week, with needles retained 20–30 minutes; electroacupuncture is common for knee pain. Ask for review at six to eight sessions and a stop rule, and make sure the plan includes exercise and weight management, because those carry the strongest evidence in osteoarthritis.
There are two delivery routes. Dedicated traditional Chinese medicine hospitals (中医院) have their own wards and clinics. The alternative is the TCM department (中医科) or integrated medicine department (中西医结合科) of a general Grade 3A public hospital — the kind this site coordinates. The directory's 30 hospitals are general and specialist public hospitals of Western medicine; a few run TCM or integrated departments, and China-Japan Friendship Hospital in Beijing is the main hospital of the National Center for Integrated Chinese and Western Medicine. None of the 30 is a dedicated TCM hospital. If you want one — Guang'anmen in Beijing, Longhua in Shanghai — that is a different referral route, outside this directory.
The consultation is in Chinese in an ordinary clinic, and arthritis history is mostly verbal: where it hurts, when, what helps, what you can no longer do. A bilingual coordinator matters more here than for a surgical booking, where decisions rest on imaging you can both look at. Ask for the record, the prescription and an English summary.
What a course costs, and how the bill works
Billing in Chinese public hospitals is per item; there is no package price. Registration, the consultation, each session, electroacupuncture, cupping and moxibustion, tuina, and any herbal prescription and its decoction fee are separate lines. Ask for an itemised estimate before you start.
Sessions in an ordinary TCM department cost about $15–50; international departments and private clinics charge $60–150. Ten to twenty sessions therefore cost roughly $250–800 in standard care, or $900–2,500 in an international department. Add a consultation at $7–42, imaging at $42–111 if it is needed, and herbs at $5–15 a day. A steroid injection costs a fraction of a 20-session course, and a knee or hip replacement in China runs about $7,600–11,800 on this site's numbers. Insurance often excludes acupuncture for arthritis, so plan to pay and claim.
The trip arithmetic matters: two to four weeks of flights and accommodation usually cost more than the treatment, so the journey makes sense only if you also want an expert read of your imaging or a second opinion on surgery.
Where it fits: before, after or instead of an operation
Before surgery, acupuncture is a reasonable symptom-control option while you do the things that change the disease course: exercise, weight loss, physiotherapy, and painkillers or injections as needed. It does not alter cartilage, alignment or the progression of osteoarthritis, and no trial has shown that it does.
After knee or hip replacement it can help with pain and stiffness during rehabilitation, with the surgical team's agreement. Two cautions: for several weeks after arthroplasty most patients take blood thinners to prevent clots, so needling carries bruising and bleeding risk; and any needle near a new joint is a theoretical route for infection, so ask the surgeon first.
Instead of an operation is the wrong choice when the arthritis is end-stage: bone-on-bone, night pain, stiffness, walking limited to a few minutes. Replacement is the intervention that reliably relieves that. It is also wrong when the diagnosis is not osteoarthritis — a hot, swollen joint needs urgent assessment for infection or gout; night pain with weight loss needs investigation; a young patient with inflammatory features needs rheumatology; and a sudden inability to weight-bear after a fall may be a fracture. The harm in this field is delay: twenty sessions that postpone a diagnosis or a needed operation cost more than they buy.
How to tell a department that helps from a clinic that sells hope
These questions separate a clinical service from a sales operation:
- It asks for your X-rays or MRI before the first session, and someone reads them.
- It gives a written plan with a review date and a stop rule.
- It says plainly that acupuncture will not repair cartilage or stop arthritis progressing.
- It refuses to promise a result, a percentage or a timeline.
- Its needles are sterile, single-use and opened in front of you.
- It asks about anticoagulants, implants, diabetes and previous surgery.
- A named licensed physician is accountable for the plan.
- It bills per item with receipts and hands you the records.
- It refers you to an orthopaedic surgeon when the joint is end-stage, instead of adding sessions.
Walk away from package prices quoted before assessment, from cartilage-regeneration claims, and from anyone who tells you to cancel a recommended replacement.
What this means in practice
For knee osteoarthritis, acupuncture gives a small, short-term improvement over sham that most people would barely notice; for hip osteoarthritis the sham-controlled evidence is essentially negative. It does not slow the disease. It is a low-risk adjunct alongside exercise and weight management, and a poor substitute for replacement when the joint is worn out.
MediTour Care coordinates appointments, interpretation and travel; it is not a medical provider and does not give medical advice. Whether acupuncture belongs in your treatment, and when an operation is the better answer, is a decision for you and the clinicians treating you.