PainScience.com is “all about the science,” but not actually all: I am realistic about the limits of the science, much of which is clearly pretty junky. (And I don’t just mean blatant pseudo-science in fake alt-med “journals” — I’m talking about the weakness of normal, mainstream science.) This is a huge topic, but here’s one simple example of a common problem: control groups that don’t actually control. This is particularly common in back pain science.
Rather than comparing a treatment to a good, carefully selected placebo, most studies use a comparison to a treatment that is allegedly neutral, underwhelming, or placebo-ish. That makes the results hard to interpret: if they work out about the same, it could mean that the treatments are equally effective … or equally ineffective!
So much back pain science has this problem — or any one of a dozen others — that you can effectively ignore at least 80% of it, because it’s so far from the last word on anything. Data like that mostly just muddies the waters.
Good science is essential to solving these problems, but really good studies are also really difficult and rare. See the next post.