Guided graded exercise self-help for chronic fatigue syndrome: Long term follow up & cost-effectiveness following the GETSET trial, 2021, Clark et al

The abstract is amazing in its deception. There was no significant time x group interaction on any outcome measure. This is the relevant result in such a study design. The correct interpretation is that the treatment did not work. But the conclusion is worded in a similar way to the PACE long term follow up paper which spun the same sort of situation in a positive light. I can’t scrutinise the paper further since access to it is denied to us patients without institutional access but I’m sure it’s similarly filled with deceptive statements. Of course, merely pointing this out is just taken by these ppl as further evidence of being a vexatious loon who is happy to see another treatment fail and thus needing psychotherapy.

 

"I'm sorry I got caught" is the most pathetic there is but this doesn't explain the journal's inability to perform basic peer review and editing, why it was necessary to push them to do something that is expected to be part of their work to begin with. This is the "my dog ate my homework" of excuses. It's not even how pathetic the excuse is, it's that no one cares that is absurd. This is a pattern with this research, yet it's still held to standards so low that anything continues to pass.

Because none of this is accidental, it's fully deliberate, like exaggerating the cost of "somatization", which is not at all what they evaluated anyway. It's a pattern of willful deceit that the entire profession is complicit in enabling. The difference here is quite literally turning "it works" with "it doesn't work". Blatant deception.

 

https://twitter.com/user/status/1412285468746694656

 

It will be so informative when we will be able to do studies of the participants in those trials to get the real story of what happened.

Presently impossible to do, but a necessity to learn from this disaster.

 

This may be a trivial thought, but I find that the improvement in the “standard medical care” group at follow-up which led it to catch up with the improvement in the GET group is the nail in the coffin for GET. Assuming that it was indeed GET that led to an improvement in the active group, GETSET shows that this improvement is so marginal that doing nothing is just as effective. It is hard to come up with a rationale for recommending GET then.

 

Coming soon in BPSland: akshually controlled studies are bad, only randomized is better.

Probably. Very likely.