Designing a double-blind test for acupuncture

[H'ethetheth]

That's a new one on me.
They certainly didn't have any objection to cutting up people when they were alive...

I had thought about that, but this thing about chinese medical practise is obviously just something I heard, and I did recently, though I don't remember where.
Anyway, I wasn't there, and I'm not a historian. So I might as well shut up now...

So that you will be basing your belief that it doesn't work on an objective measure...unlike the anti-vax woos, who don't need no steenkin' proof?

You're right, I spoke hastily (and in a slightly intoxicated state).
I've just been informed I'm not allowed to take it back though, sorry mate.

Now about that test. I see where the problem arises, but I don't see a simple solution. I'll think about it, and see what I can come up with.

 

[CBL4]

Ririon idea of training people to insert needles in the "correct" places and and random places is what I meant. I do not see a need for it to be quick however. It could be repeated weekly for a few months and if neither the acupuncterists or patients had no reason to believe they were using the wrong points, it would continue to be double blind.

You could also have some "real" accupuncturists judge the accuracy of the needle placement via videotape. This would prevent after the fact claims of improper procedures.

CBL

 

[patnray]

The other thing to do is have to try the "true" accupuncture points vs. random points. The people inserting the needles would not be told which are real and which are fake.

CBL

Perhaps the patient could be behind a screen through which the needles are inserted so the accupuncturist doesn't know where on the patient's body the needle was inserted. The practitioner, after making his diagnosis, would request, through a third party, that the patient be positioned so a certain "meridian" is at the screen. Some patients would be instructed to present the requested "meridian" at the screen and others would be instructed to present non-meridian locations.

Could the practitioner tell the difference? Would there be any difference in outcome for the two classes of patient?

 

[crimresearch]

Folks, acupuncture has never been a dart board....needles aren't randomly stuck...the spots are located by, among other things, feeling for a 'depression in the Qi on the meridian'.

The needles are inserted at certain angles, and depths, again by 'sensing' the patient's Qi and it's resistance to the needle.

And then the needles are manipulated at that spot, depth and angle to 'restore proper Qi flow' according to the ailment diagnosed...

It is so subjective as to be practically untestable in the manner suggested.

What you folks are doing is like saying that you can test whether or not auto mechanics 'works' by throwing screw drivers into a running engine, and seeing if the problem is fixed.

If you don't follow the same procedures as the practitioners, you haven't tested their practices...you have only tested your incorrect version of the practice.

By attemtping to test efficacy, you are peering into the wrong end of the horse...
Move to the other end, and look in the mouth....start by testing the practitioner's ability to *diagnose* with replicable results.

 

[Ririon]

Ririon, the problem with your "sham the student" protocol is that acupuncturists claim that over 100 hours of training is needed. Of course, in reality, one only needs a bunch of needles and a gullible patient. However, that would be the excuse they use.

I guess you have to lock them up for a couple of weeks, then...

Anybody have access to 50 eager students, 100 people with the same ailment, as diagnosed by a concensus of acupuncturists and an empty prison wing complete with guards? The patients have to be locked up, too, so they too have no way of finding out if they are given the real thing.

As I said, it gets complicated...

Anyway, this is about the level you would have to take your research to if you wanted to get some credible results to support acupuncture. If the effect was very strong, you would need fewer people, but I think we can safely assume based on the research available, that any effect would be very small compared to placebo, so you would probably need even more people than this to get anything statistically significant.

However, to demonstrate that acupuncture is complete BS, you don't need to go through all this trouble.

A single blind study where the acupuncturists know if they are faking (but communication about this to the patient is minimized) is much easier to conduct, but can not be taken credit for if it shows acupuncture to work. As a preliminary study, that could either disprove acupuncture (beyond any reasonable doubt, to use a legal term) or trigger the whole "sham the student"-study.

Ririon

 

[Asolepius]

This is a fascinating topic, with some insightful posts. As some of you may know from other threads, I was a patient in an acupuncture trial 3-4 years ago. This has now been published, and is the subject of an acrimonious dialogue between myself and the authors. Essentially, the investigators lied to me about the study design when I enrolled. I still have the patient information sheet - the title of the study is different from that on the published paper. The authors' university is now investigating the study, although I am not holding my breath.

But this discussion highlights the subjective nature of the practice, and the difficulty of blinding it. I like the idea of testing for the diagnostic component. I have had acupuncture 3 times, and each time was told to expect some sort of sensation on insertion - this has been termed `de-chi'. The practitioner is supposed to be able to detect this. Well I never felt a thing, so how did the practitioner place the needles correctly? Perhaps we could test the concordance between what the patient feels and what the practitioner feels?