The statistical significance of length of stay was found only when they did multiple pairwise comparisons between the 3 groups, which should require and additional adjustment of the threshold P-value.

Once again, if there’s no clear scientific basis for a CAM claim, I guess the existence of the Tooth Fairy, all the other “research” is just plain irrelevant. Don’t show me that something happens until you show me how.

My dog IS better than healing touch, and I don’t have a dog!

in case anyone wants complete reference. Not indexed in PubMed.

The physical heart is REAL?!

Only if we allow it to be, apparently.

The physical heart is REAL?! Now that’s just crazy talk.

Enjoyed the article (as usual), Dr. Hall. Excellent slicing-and-dicing.

Does that mean touch is quantized? Does the Toucher become entangled with the Touchee?

Can we have spooky touch at a distance? Oh, wait. That’s what they’re doing already.

How about Schrodinger’s Touch – you can’t know if you’ve been touched or not until you look in your wallet.

:^D

Specifically, Sharon McDonough-Means and Iris Bell condicted a study on theraputic touch for stressed neonates. This, I think, if it was conducted correctly, is a brilliant strategy. Babies don’t know or care if you are non-touching them and can’t be influenced by the process. Anyway, the study, ClinicalTrials.gov identifier: NCT00034008, is completed. I can’t find where on the NCCAM site these filed reports are located. But the information on the study protocols are here:

http://clinicaltrials.gov/ct2/show/NCT00034008?term=%28NCCAM%29+%5BSPONSOR%5D+%28completed%29+%5BOVERALL-STATUS%5D+OR+%28NCCAM%29+%5BSPONSOR%5D+%28suspended%29+%5BOVERALL-STATUS%5D+OR+%28NCCAM%29+%5BSPONSOR%5D+%28terminated%29+%5BOVERALL-STATUS%5D&rank=23

Sorry, that’s long!!

Does anyone know how to access the completed study?

I don’t have a citation or anything for you, but you may get what you want. My institution has a med student volunteer organization that brings dogs (who have undergone some sort of training/obedience course especially for the program) into the hospital.
I understand one of the med students is attempting to find funding to study if it has any actual effect beyond a social one. The med students running the group don’t seem to have any illusions about this though, they’re just happy the patients are smiling, they don’t pretend to have treated anyone.

The mock/sham HT group would be no different from the “real” HT group, and the researchers would conclude that the mock/sham HT was also effective, possibly because the practitioners could not help willing good intentions to the subjects or stop their energy fields form positively interacting with the subjects’ energy fields due to their inherently positive intentions.

I have a particular interest in HT, as it was an attempt to get it officially sanctioned in my little rural hospital that turned me from a closet skeptic, to an activist. Yes, we are still (officially) a woo woo free hospital.

I pointed out at the time that all they were really claiming to do was make patients feel more relaxed and less anxious, and we could accomplish the same thing by allowing by allowing pets to visit.

Looking forward to the SBM conference at TAM7. See you on Thursday.

They apparently don’t understand how blinded or double blinded RCTs work, or perhaps they consider them all unethical.

As long as it is disclosed to the study subjects that they are being randomized in to a group that may receive either real or sham/mock treatment, there is no ethical problem. It is not exactly the best ethics in the world intentionally exclude a control (mock/sham treatment) that would increase the quality of the study and its results.

If the practitioners are so convinced of the efficacy of their therapy that they feel it is unethical to offer sham treatment, why do they think it is any more ethical to withhold supposedly effective treatment at all?

…that would incentivize me to get out of there ASAP whether I was fully healed or not.

Me, too.

Their choice of statistical techniques is highly dubious.

Nothing to say of their interpretation of those highly dubious statistics, even if they were valid. Look at yesterday’s post by David G. and consider how prior probability figures in whether or not we are justified in concluding that ’statistical significance’ implies a true hypothesis. From this paper (Table 2) we see that a P value of 0.01 for data from a study of a hypothesis whose prior probability was 1% (much higher than a reasonable prior probability of a hypothesis based on psychokinesis, by the way) will raise the posterior probability to (only) 13%. That is a far cry from from the 99% that most people, even in academic medicine, imagine to be the case.

When you add all the other design problems with this HT study (no blinding, probably no allocation concealment, no credible control group as Harriet pointed out, etc.), it is reasonable to put it in Ioannidis’s category of very high ‘bias’, which makes the result of this study entirely, well, useless–although it could be useful as a disconfirming study, but that would require greater interpretive powers than, apparently, exist in the field of ‘academic CAM.’

Given all that, by the way, here’s another topic for a blog: how the National Library of Medicine chooses ‘CAM’ journals for indexing.