Réservez une séance
EN — English
📅
Book a session — Calendly
Perfect blinding is impossible, and that is true of surgery and physiotherapy too. Here is what trials can control, and what the results actually show.

Where this idea comes from

The argument sounds rigorous and is often repeated by people who have read the phrase double-blind but not the methods sections. If the practitioner must know what they are doing, the reasoning goes, the study is unblinded, and an unblinded study proves nothing. The conclusion is delivered with the confidence of a logical proof.
It is also convenient for practitioners who would rather not be measured. The same sentence can be used to dismiss an inconvenient trial, and it appears in marketing material as often as in sceptical commentary. An argument that both sides find useful is usually doing less work than it appears to.

What Chinese medicine actually says

The tradition has no concept of blinding, and it would be anachronistic to expect one. Its method was clinical observation over long periods, with the practitioner as the instrument of both diagnosis and treatment. That is a genuine methodological limitation, not a cover story, and it applies to much of the history of medicine.
What the tradition does insist on is that the individual matters: the same complaint in two people may call for different points. That commitment makes standardisation difficult, and it means a trial testing one fixed protocol is testing something narrower than the practice. Both statements are true at once. A trial that standardises the protocol is therefore testing a simplified version of the practice, which is a real limitation and a manageable one.

What the research says

Blinding is a spectrum rather than a switch. Patients can be blinded with retractable sham needles, outcome assessors can be kept unaware of allocation, and data analysts can be blinded to group. What cannot be blinded is the practitioner, and the risk is that enthusiasm leaks into the interaction and inflates the estimate.
That risk is real, which is why sham-controlled results are the more conservative ones. In the 2018 analysis by Vickers and colleagues, the difference between acupuncture and sham was small but present across chronic pain conditions. The Cochrane review of migraine prophylaxis by Linde and colleagues in 2009 reached a similar conclusion with moderate-certainty evidence.

What you can realistically expect

Questions about evidence are answered specifically at Centre Équilibre: Ghislaine Morin sets out which trials exist, for which conditions, and how strong the findings are. The centre does not claim that blinding is unnecessary, and it does not claim that the question is unanswerable. It describes a modest evidence base and the limits that come with it.
The TCM dispensary she runs at the centre in Grenoble faces a stricter version of the same problem, because herbal formulas are harder to standardise and to blind than needles. That is a reason for caution about claims, not a reason to abandon measurement, and the two are often confused in commercial settings.

When to consult, and when not to

Consult if you accept that the evidence is partial and want a time-limited trial with a clear measure of success. Do not consult anyone who says the question cannot be studied; that claim is false, and it usually protects the practitioner rather than the patient.
Do not accept a session as a substitute for a diagnosis, and do not let a methodological argument, in either direction, decide a medical question. If a treatment is being recommended for a serious condition on the grounds that trials are impossible, that is the moment to ask a doctor instead.

Frequently asked questions

Is double-blind testing impossible for acupuncture?

Perfect blinding is impossible because the practitioner must know what they are doing. Partial blinding is routine: patients can receive retractable sham needles, and assessors and analysts can be kept unaware. Trials are therefore feasible, and they have been done at scale.

What is a sham acupuncture needle?

A device that looks like a real needle and touches the skin without penetrating it, or penetrates very shallowly at non-points. It lets researchers compare needling with the ritual of needling. It is not a perfect control, but it separates the ritual from the puncture.

Do unblinded trials of acupuncture prove anything?

They are weaker evidence, because enthusiasm and expectation can inflate reported benefit. That is why the sham-controlled trials carry more weight, and why their smaller effect sizes are the more credible estimates. Unblinded studies still contribute to safety data and to descriptions of ordinary practice.

Useful links

Sources

If acupuncture works on animals, does that rule out a placebo effect?Is acupuncture just attention? What placebo research can tell us
Loading...
Ghislaine Morin
Ghislaine Morin
Praticienne en médecine traditionnelle chinoise et sophrologue à Grenoble
Sommaire
0%