Yawn. Another worthless acupunct–I mean acupressure–study

ResearchBlogging.orgHere we go again.

It seems just yesterday that I was casting a skeptical eye on yet another dubious acupuncture study. OK, it wasn’t just yesterday, but it was less than two weeks ago when I discussed why a study that purported to show that acupuncture worked as well as drug therapy for hot flashes due to breast cancer therapy-induced menopause. Unfortunately, these days these sorts of dubious studies seem to be popping up fast and furious like Whac-A-Mole, so much so that I can’t always keep up with them. So it is again, although this time it’s acupressure, not acupuncture. Unfortunately, this time it’s an experiment on children as well. Why woo-meisters insist on subjecting children to their woo, I don’t know, but they do:

An adhesive bead placed at a pressure point between the eyebrows reduced anxiety by 9% during the preoperative waiting period whereas anxiety continued to rise 2% for children who got sham acupressure (P=0.012), reported Shu-Ming Wang, M.D., of Yale, and colleagues in the September issue of Anesthesia & Analgesia.

However, the randomized trial showed acupressure had no effect on the need for intraprocedural propofol (Diprivan) or the depth of sedation.

Steve Novella‘s not going to be happy about this. The lead author of the study is Shu-Ming Wang, MD in the Department of Anesthesiology at Yale School of Medicine. I feel his pain. I’m lucky in that I have not seen much woo in my particular institution thus far. I can only imagine how irritating it would be to be at an institution that not only produces studies like this one but has earned a prominent place in my Academic Woo Aggregator, largely thanks to its most famous woo-meister, Dr. David Katz. I don’t see him on this article, but I see his influence permeating even departments that should know better.

But my sympathy has lead me to digress.

Let’s get get back to the topic at hand. First, note how little the purported effect was. Second, note that an objective measure did not change, namely the amount of sedation required (more on that later). But you won’t hear that in the credulous article I found about the study:

Acupuncture has been shown to relieve anxiety in adults before surgery, but acupressure — a noninvasive stimulation technique — is likely to be more appealing to children than a technique involving needles, the researchers noted.

These findings are likely to be clinically important since excessive preoperative anxiety contributes to operative delays and heightens the pain response, commented Zeev N. Kain, M.D., of the University of California Irvine and a coauthor on the study.

Yes, it’s all there: Claims that acupuncture definitely works, and a claim that a very small effect is “clinically significant.” For example, later in the article, Dr. Kain claims that a 10% reduction of anxiety is significant in adults and then extrapolates from that that therefore an 11% reduction must be significant in children based on–no evidence.

But let’s get to the paper itself, entitled Extra-1 Acupressure for Children Undergoing Anesthesia. Looking at the paper itself, I find a lot of holes. Being the old fart that I am, I can’t help but liken the number of holes in this study to the number of holes it takes to fill the Albert Hall. Yes, my friends, it’s that bad. Let’s look at their hypothesis first:

We examined whether acupressure in the Extra-1 (Yin-Tang) point would result in decreased preprocedural anxiety and reduced intraprocedural propofol requirements in a group of children undergoing endoscopic procedures.

What I find interesting is that there’s no discussion at all of why they think that. Why? There’s no physiological mechanism discussed, no potential reason discussed, not even a speculation. Why this particular point? It’s just a spot between the eyes. There’s no plausible mechanism or reason to suspect that it would work. Yes, they do cite some papers claiming that acupuncture and acupressure “work” for adults, but neglect to mention that the totality of evidence is at best equivocal. But they give no reason for us to think that sticking a bead on one area of the head would somehow magically decrease postoperative anxiety.

So, how, specifically was this study done? Fifty two children were randomized to either acupressure a the Extra-1 point or “sham” acupressure defined thusly:

  1. Ex-1 Group. Intervention applied at the Extra-1 acupoint, which is located in the midpoint between the eyebrows.
  2. Sham Group. Intervention applied above the lateral border of the left eyebrow. This widely used sham point was selected as it has the same dermatomal distribution as Extra-1 and does not result in any reported clinical effects.

Wow! It sounds all science-y, doesn’t it? They even used the term “dermatomal distribution”! Very nice. There are a couple of problems revealed in the methods, here directly quoted:

  • In order to prevent any possible bias, all participants were informed that the aim of this study was to determine “which acupoint on the forehead is more effective in reducing preoperative anxiety.”
  • The acupuncturist (SMW) applied acupressure beads to the two intervention groups based on a random computer-generated assignment.

The first point is a almost as though authors decided to try to gin up a placebo effect as much as they can. I would have said that the study was to test whether acupressure works. It’s a subtle but important difference. The second part tells you something very, very important, namely that the person giving the treatment was not blinded to the experimental group. Dr. Wang is an acupuncturist who presumably believes in the efficacy of acupuncture and acupressure. It’s not unreasonable to wonder if through body language he telegraphed which group each child was in. If the effect observed had been large, one might argue that that couldn’t account for such a large effect, but the effect reported is small, only 11%. Add the two together, and I worry very much about the validity of the reported results of this study. Further add that it is unclear whether any of the other treatment staff was blinded. Again, when the reported difference is so small, such issues become very important.

Finally, one part of this paper that is very important and that I might not have noticed if someone in a certain discussion forum hadn’t mentioned it was the method used to monitor anaesthesia. True, it wasn’t the primary endpoint, but it is important. It turns out that the Bispectral Index (BIS), the method of monitoring patients, is an EEG-linked system that produces a score between 0-100, with 0 representing no brain activity and 100 representing an awake patient. The target was 40 to 60, and, it should be noted, a BIS less than 60 is generally considered adequately anaesthetized for endoscopic procedures. It turns out that BIS is not nearly as well accepted as is implied in this manuscript and has several limitations, such as variation with different anaesthetic agents, a proprietary closed system where the practitioner can’t be sure how the BIS is calculated, difficulty correlating with depth of anaesthesia, and inapplicability in some special patient populations.

However, the biggest, most glaring weakness of this paper is that, although the self-reported anxiety scores were 11% lower in the “true acupressure” group than in the sham, there was no difference in the amount of Propofol sedative necessary to maintaint a BIS between 40-60. Indeed, it’s rather amusing to watch the authors try to rationalize this observation away as being insignificant:

Lack of group differences with regard to intraprocedural propofol consumption should be viewed in light of three previous studies. In a study of adult patients, Maranets and Kain noted that only higher trait anxiety (but not state anxiety) resulted in higher propofol requirements for maintenance of anesthesia. Thus it is not surprising that in this study reduction of preprocedural state anxiety did not result in reduced intraprocedural propofol requirements. Further, it is possible that intraprocedural effects of acupressure were suppressed by propofol general anesthesia. Indeed, a previous functional magnetic resonance imaging study by our laboratory indicated that acupuncture-induced blood oxygen level dependent signals were suppressed by propofol general anesthesia.25 A similar observation was seen using an auditory evoked index monitor. Lu et al. noted that electroacupuncture enhances the sedative effect of propofol in a target plasma concentration of 1.5 !g/mL but not at 2.0 !g/mL.

That’s right. Propofol kept acupressure from working, so the same amount of propofol needed was the same between groups. Funny how that worked out so that its effect perfectly did that, isn’t it? Ain’t always the way it works out in these studies? It’s also funny how the authors dance around how small the effect supposedly observed was, a mere 11%, an effect that could easily be random chance and/or due to lack of blinding of practitioners, and that there is no standard for what is considered clinically significant in children.

This study is yet another example of putting a lot of effort to have all the trappings of science to study a question that, at its heart, is in essence magic. Remember, acupressure is based on prescientific and mystical concepts of how the body works through some unmeasurable “life force” that flows through certain “meridians.” Moreover, like most “positive” studies of woo, it demonstrated a small effect that’s at the edge of significance. I will, however, admit that comments after the Medscape article did give me a chuckle:

Interesting but not real Chinese medicine. Personalization, comprehensiveness and strategy are foundational to classical Chinese medicine. This study says something obvious to anyone at beginners- level of Chinese medicine. This lowest common denominator technique applied generically only hints at the much greater value of real acupressure and acupuncture. Get someone in there who really practices Chinese medicine based on classical principles to design these studies.

The problem is, “real” practitioners of Chinese medicine generally design studies even worse than this one or don’t do studies at all. After all, they believe acupuncture works. What more do they need? Even more amusing:

No real practitioner would design or do a study like this, as it was not structured to be to any real or appropriate treatment. There was NO design to do any anesthesia/sedation, so any outcome there would not be expected. Even the point used would not be on the top of the list for this procedure. The anxiety reducing point that was used (Yintang)is only one point in a complete breakfast…. and would never be used alone. There are more appropriate points that could be used for the desired effect that this study was looking for. This was a poorly designed study, and even then it produced results. I understand that some of the philosophy of this study would be a cheap and easy way anybody, even not trained as was stated in the article, to reduce anxiety in kids in an invasive procedure. There are ways to do that. as most practitioners would know.

In other words, the investigators picked the wrong magic points or didn’t use enough magic points! But even more hilariously, he’s claiming that even a “crappy” study from an acupuncture standpoint produced “results.” Yes, it did: Questionable results that are probably not real.

Just like nearly every study of unscientific “complementary and alternative medicine” therapies. Given that this was a study in which the subjects were children, one question remains to me: How on earth did this study get through the Yale IRB?


Shu-Ming Wang, Sandra Escalera, Eric C. Lin, Inna Maranets, Zeev N. Kain (2008). Extra-1 Acupressure for Children Undergoing Anesthesia Anesthesia & Analgesia, 107 (3), 811-816 DOI: 10.1213/ane.0b013e3181804441