Inside sources tell me why the body's new ME guidelines were delayed
I wrote a piece this week about ME/CFS, and the mystery of the new treatment guidelines being shelved. One of the things I found weirdest was that while NICE’s decision to delay publication had (as far as I could tell) been influenced by various clinicians’ groups — notably, some Royal Colleges, including those of psychiatrists, physicians, and paediatrics and child health — no one seemed to be willing to say why.
Since I wrote it, two people with some knowledge of the internal workings of both NICE and the Royal Colleges got in touch. They’re experts in relevant fields, and I know them both and trust them, so I thought I’d try to express the points they made.
First, they both said, NICE created a new definition of the disease when producing the new guidelines. The details aren’t hugely important, but because of the way the NICE evidence evaluation works (a system known as GRADE), the changed definition means that all evidence gathered under the old definition is automatically downgraded by one level, from “good” to “moderate” or from “moderate” to “low” etc.
According to my sources, this was the issue that the Royal Colleges had: that the new definition inevitably rendered all the evidence on GET low-quality. “It was quite well known it would have this effect,” said one, “so the colleges all wrote really quite detailed feedback.” None of that feedback has (yet) been published.
Essentially, the colleges said that this change would be bad for patients. My source made the point that while patients who have suffered for a long time with treatments that don’t work for them are likely to join campaign groups, those who see moderate improvement with GET and return to their lives probably won’t. So patients’ groups systematically tend to end up run by those patients with intractable conditions. I tended to conflate “patients’ groups” with “patients” in the piece, and perhaps it’s not as simple as that.
My sources also wanted to defend the PACE trial somewhat. Yes, they said, it’s true that it was unblinded and had subjective outcomes, meaning that it was open to bias. But given the difficulties, they say, it was probably as good as it could be: ME/CFS can only currently be diagnosed on subjective grounds, so subjective measures were necessary.
PACE has been accused of “outcome switching” — that is, changing what it was looking for mid-stream in order to make its results look better. I’ve even had people tell me it was “fraudulent”. I think that’s unfair: independent reviews say that changes to the outcome measures were reasonable and transparent. But it’s a huge ongoing row — see here; it appears that the changed outcomes did mean that; on one secondary measure, some patients were recorded as improving despite having got worse, although it didn’t affect the primary outcomes.
I should add that I somewhat misrepresented the PACE trial. I said it was CBT/GET vs standard medical care, where in fact there were four arms: standard medical care, CBT, GET, and “adaptive pacing therapy”: that is, planning and limiting your activity according to the restrictions of your disease. Pacing was the ME/CFS patients’ groups preferred method, but it showed no better outcome than standard care. I don’t think it materially changes any of my conclusions, but it is interesting that any bias in favour of CBT/GET didn’t show for pacing therapy. Also, it’s probably relevant that GET didn’t seem to have any worse side-effects than the other therapies, at least in this trial.
My understanding now is that the Royal Colleges were nervous that they were essentially being forbidden from offering GET, even though clinicians found that it was helpful for many patients. They may have also thought that the evidence supporting that ban was inadequate for the reasons discussed above. They suggested that the NICE GRADE system was badly designed for complex problems like ME/CFS. And they put all that in some long feedback, which, they say, was not then reflected in the final document.
And none of this feedback was made public. I am not endorsing any of what was said above, although I profoundly wish someone from NICE or the RCPCH or the various groups I contacted had talked to me, so I could have included it in the piece. But I do think this is the sort of thing that needs to be said out loud, so we can all assess it: as I said in my conclusion, NICE should have published the evidence, or — if the evidence wouldn’t make sense without the guidelines — some sort of explanation of what’s gone on. Somehow, the NICE process has got badly tangled up here. “It’s an unholy mess,” says one. “The colleges don’t want to come out publicly and say NICE’s committee screwed up.”