IDET (Intradiscal electrothermal therapy) fits perfectly into the template for a successful placebo treatment described in my earlier blog. It has everything: biological plausibility, high tech equipment, high cost, conflicts of interest, great lab results, encouraging results from early clinical studies, and no advantage over placebo...
The idea behind IDET is to reduce back pain by inserting a probe into the painful disc (the whole idea of the disc causing pain in the first place is another story), and then heating it up to “shrink” the disc, or “seal” the disc, or kill the nerve endings, or [INSERT BIOLOGICALLY PLAUSIBLE MECHANISM HERE]. You don’t have to kill that many animals to be able to show stuff like this in lab experiments.
The brothers who invented IDET wrote some reports about how patients got better with this treatment. Interestingly, about 70% of them got better. 70% is a number that is often seen in placebo treatments, so when I see reports of a series of patients that showed around 70% improvement, I am immediately suspicious, and then I start looking for the real evidence: blinded randomised trials.
But the inventors knew that reports of case series weren’t going to be enough, so they wrote a review article. This article summarised their previous reports, thereby appearing to build up the evidence base. It used unscientific terms that raise my suspicion of ineffectiveness, like “very promising”, and “an attractive alternative”. Apparently, that was all that the FDA needed to approve it. That approval, and the appeal of a new, high-tech treatment with some kind of biological plausibility, and some positive publications (regardless of how weak) is all it takes for gimmicks like this to take off.
Then some researchers did a blinded placebo study where the patients and the people doing the evaluations didn’t know who got the active treatment and who got the placebo (dummy IDET). The results? Absolutely no difference between the two groups, of course. The other placebo study published a year earlier did show a marginal improvement in the IDET group for some of the outcomes, but these results have failed the test of replicability.
Is it so hard to ask for these studies to be done before devices are approved and released?
Yet IDET is still being used widely. Many more case series (without any comparison or control group, let alone a placebo) have been published showing that patients get better after this treatment, normally about 70% of the time. But this kind of evidence is no better than the evidence for a lot of other things that don’t work because the supporters fail to realise that about 70% of people will improve after almost ANY treatment, as long as it is given in a supportive and credible way.
We are asking the wrong questions. We shouldn’t be asking “Do people feel better after this treatment?” We should be asking:
1. “Would the patients have improved anyway, without the treatment?” (i.e., what is the natural history of the condition?), and
2. “Did the patients get better as a direct result of the physical effects of this treatment, or did they get better because of everything else that happened to them, like providing them with an expectation that they would get better?” (i.e., what is the placebo effect for this procedure?).
You are probably thinking that if this thing works in 70% of patients, what’s my problem? Firstly, I don’t want to pay for it (whether in taxes, health insurance premiums, workers compensation premiums, or in the opportunity cost for doctors and patients due to the wasted resources on this ineffective treatment), and secondly, it can be harmful, as demonstrated by reports of massive disc rupture, dead vertebral bone and loss of nerve function as a result of IDET.
And the punch line? The brothers who developed the technology with a start-up company in 1995 (and wrote the supportive articles) sold their company three years later for $300 million. This makes the conflict of interest statement on their 2002 paper one of the great understatements of all time:
“One or more of the author(s) has/have received or will receive benefits for personal or professional use from a commercial party related directly or indirectly to the subject of this manuscript”