Posted tagged ‘neck pain’

Acupuncture doesn’t work, and the media blows it again

October 8, 2012

The Pediatric Insider

© 2012 Roy Benaroch, MD

This blog is here to provide solid, reliable, science-based information. And by science based, I mean I read the actual studies in the actual science literature—sometimes requiring great quantities of coffee, yes, but that’s the price one has to pay. Then I tell you honestly what’s been figured out. Sometimes it’s something new, sometimes it’s something old, and often it’s something that disagrees with what you’re reading in the newspaper. Such is the nature of science: it is not a static body of knowledge, but rather a method of observing and testing the natural world. Progress. Science!

Media reporters are supposed to read and interpret, too. After all, most people don’t have the time/patience/wherewithal/serum caffeine concentration to wade through the arcana of the scientific literature. So the “science media” developed, to interpret what the big-brained eggheads were up to, to explain things so we could all become more knowledgeable.

To my dismay, science reporting is often just… deplorable. Journalists seem to parrot whatever the “message” is supposed to be, whether or not it’s supported by the research. No one talks about who did the study, who influenced the study, what the study’s shortcomings are, or what the study really showed. You get a snappy headline that’s akin to a nearsighted monkey throwing a water balloon at a pile of porcupines. I suppose wherever it pops, that’s the truth. Take the recent reporting of the AAP’s new opinion on circumcision. What the AAP actually said was that the benefits of circumcision are small, but that they outweigh the risks; so insurance should pay for the procedure for parents who choose it for their babies. The AAP did not endorse circumcision for all babies, despite what the headlines proclaimed.

And now, a new example. The headlines: “Acupuncture provides true pain relief in study” (New York Times), “Acupuncture works, one way or the other” (CNN). Even the National Library of Medicine joined the bandwagon: “More Evidence Acupuncture Can Ease Chronic Pain.”

Wow. One might think from these unequivocal headlines that some kind of new study has really proved that acupuncture works. Unfortunately that’s not what the “study” showed.

The paper itself was published last month in The Archives of Internal Medicine. The authors, all members of the Acupuncture Trialists’ Collaboration, re-reviewed data from 29 of the 31 previously published, randomized trials of acupuncture for any of 4 chronic pain conditions, mashing up the aggregate patient sets into one big study. This is called a “meta-analysis”—it’s not new research, per se, but a way of seeing an overview of a number of studies. Of course, if the studies collected were biased or of questionable quality, a meta won’t make them any more reliable; still, a re-publication seems to attract media eyeballs.

So what did the authors say that they found? There are a lot of statistics to wade through, but I’ll simplify: assume that your back hurts, and on a scale of 1-100 you rate the pain a “100”. On average, the included studies found that your pain would be reduced to a score of 60 with no intervention at all—that’s the reduction for people who had no change in anything the were doing for the pain. The people who had sham acupuncture, using fake needles, had a reduction down to 35. And the people who underwent true acupuncture had their pain reduced all the way down to… 30.

Got that? Doing nothing improves your pain by 40 points. Getting fake acupuncture reduces your pain by a total of 65 points, or 25 points more than doing nothing. And real live genuine pointy acupuncture will get your pain down by another whopping 5 points.

Do you think anyone could tell the difference between a pain score of 30 or 35 on a scale of 100? And: do you think these studies, many of whom the authors themselves admit could not be free from bias or well-blinded, could possibly accurately measure or confirm a 5 point difference?

Even more compelling: if you really focus in on the few best recent studies of acupuncture, using fake needles or other tricks to effectively blind both the patients and the acupuncturists so that no one knew whether real acupuncture was being performed, you find one consistent result: Acupuncture is no better than fake acupuncture. It doesn’t matter where you stick the needles, or whether they even puncture the skin. Acupuncture doesn’t do anything beyond what imagination, expectations, and natural healing would have done anyway.

So: a reasonable, truthful headline on this study would have been “Acupuncture no better than fake,” or “Once again, acupuncture fails to prove its worth.” And the media blows it again.

For more detailed reviews of this study, placebo effects, meta-analyses, and the sorry state of science news reporting, visit Science Based Medicine blogs here, here, and here.