CORRESPONDENCE The PACE trial of treatments for chronic fatigue syndrome: a response to WILSHIRE et al (2019) Sharpe, Goldsmith & Chalder

The question should be asked further and aiming to justify the £5M price tag. There is no reasonable justification for the price tag other than creating sunk cost momentum. It's no more sophisticated or developed than the SMILE trial with its LP BS or the 3-needle acupuncture trial we saw recently.

The cost justifications are over specialist training and weekly therapy sessions, which is labor intensive. But that's just describing what was done, not a proper justification. There is no reason why PACE couldn't have been done for 1/10th the price tag, there is no specialist knowledge involved or innovative techniques. It was big for the sake of being big, not because anything they did warranted it. There is no costly equipment or process, the training is dubious at best and they even recognize it can't be scaled to a large population, meaning it's all for show.

The only reason PACE stands is precisely because of the price tag. It justifies itself by having been so costly. The 4-day intensive thing that was done (in Norway?) was no more or less advanced or technical or any of that. This is no different than slapping a $40 price tag on a $.2 chocolate with fancy $1 decorations and getting people impressed because it's so luxurious. No one who paid $40 for that would comment it tasted like $.2 chocolate, even though that's exactly what it is.

 

The comparison of CBT and GET was a good thing to do. Unfortunately, there was not much difference between the two so no conclusion about efficacy or otherwise of either can be drawn. If one had been much better than the other it would have been a useful result. As it is the default assumption is that 'modality does not matter'. Or in other words there is no special skill needed in persuading people to fill in forms more positively.

 

Did they use this term? I know Bleijenberg and Knoop did in a Lancet editorial in 2011.

 

You might be right but if I was reviewing I would not raise all this fluff about not giving an opinion. If the referee thought the letter was cogent all he has to say is that it is a well written reply, or maybe point out some language issues.

 

For posterity's sake:

Sharpe was one of the people responsible for creating CFS as a replacement for ME. He invented the Oxford criteria that underlie his idea of a syndrome of chronic fatigue (read that again: a syndrome, by definition a presentation of multiple symptoms, made-up and focused entirely on one symptom that isn't even a main symptom of the disease he equates with chronic fatigue).

Sharpe published several papers arguing they are the same and that ultimately they are the same as neurasthenia or somatisation. His like-minded colleagues did as well.

Sharpe repeated this claim in several newspaper articles that covered PACE and his work in general. The media's confusion is entirely his and come from his own words, not taken out of context.

Authorities officially use CFS because of Sharpe's work promoting his (and colleagues') idea of a replacement for ME that reduces it to chronic fatigue, for which he invented the made-up diagnosis of CFS.

And while we're here, Sharpe is not a "top researcher" in the field and especially not "one of a few". His ideas still have no objective evidence and are a fringe and otherwise inconsequential oddity in the overall field of medical research on ME.

Sharpe deserves all the criticism he gets for these and many more reasons. His work has all the academic substance and real-world value of an inkblot.

 

I also read it this way.

 

I mean... how do you even go from there? Rest is as much a treatment of ME as avoiding an allergen is a treatment for allergies. It's not, it's simply avoiding conditions that make them worse. No patient has ever labeled rest or pacing as treatment. It's rest, which is something sick people naturally do, like when they have the flu.

They've been at it for decades and still talk all sorts of nonsense about the most basic things. And they dare call themselves "top researchers". What hubris.

 

I haven't seen the PACE authors use it. Except they read the editorial before publication so they essentially endorsed the language.

 

I think I remember Sharpe saying so. Looking. It seems to have been in the registered protocol (and I don't think it was removed) but omitted from the papers and that would be why Knoop and his colleague used that term. So basically they sold a strict criterion when registering protocol and just quietly dropped it when they could not meet it.

There's an article from Tuller on the topic: http://www.virology.ws/2015/11/04/t...really-adopt-a-strict-criterion-for-recovery/.

 

I suspect that even if you tried to plot the SF36 PF scores against perceived PF, there would probably be some pretty horrible looking non-linearity. And if you tried to plot perceived PF against true PF, that also would be a pretty quirky characteristic I imagine. So any attempt to plot SF36 PF against true PF would, I think, be horribly non-linear. The likelihood of stats on SF36 PF scores having much useful meaning regards true PF is pretty remote I think.

Absolutely. Although the Lancet never published a letter I wrote regarding this relating to GETSET, I did make it available in S4ME after it was rejected, https://www.s4me.info/threads/my-letter-to-the-lancet-in-response-to-the-getset-publication.1528/. (I realised later I should probably have used the word 'imply' rather than 'infer' in it, but never mind).

 

Here's the extract from the Bleijenberg and Knoop editorial (Lancet 2011):

Interestingly, this is the only paper I can find on which White collaborated with Knoop and Bleijenberg.

Given that recovery was not mentioned in the PACE trial paper published in The Lancet in 2011, I think I'm probably correct on my hunch that either Knoop or Bleijenberg were the most likely peer reviewers of PACE. That they had insider knowledge is not a good look.

 

Yes, I read before that reviewers often then write editorials accompanying articles.
And particularly in this case where the peer review of the paper was fast-tracked, non-reviewers would have had little time to be asked to look over the paper and then write a review of it.

It's also interesting as the authors claimed it was a reviewer who suggested the normal range analysis.

 

Really? Where did they say that?

 

I think there's a lot of truth in that. In engineering if you have something with a non-linear characteristic, you will typically deal with the non-linearity before trying to do anything else with it. e.g. A thermocouple, which provides changing voltage (albeit very small) with changing temperature, but the plot of T against V is very non-linear. So having read in the voltage it then needs to be "linearised" to arrive at the temperature. But thermocouple characteristics are (mostly) pretty sane, whereas the CFQ I think is pretty insane.

 

I seem to remember Ellen Goudsmit mentioning it was her recommendation but I don't remember anyone saying more about it. I found that odd but no one really picked it up so I ignored it but that's weird.

IIRC it was as a comment to Hilda Bastian's PLOS blog post or something like that.

Found it. Don't know if it's really her but this is the comment on Bastian's post (bold mine):

 

There was a “booster” session at 36 weeks so the gap to outcomes at 52 weeks is less than the 6 months gap they claim.

 

Once again, a collection of generalized statements and a repeat that "the big boys told us we could do it sir!"

I have yet to see a proper explanation of how scores of 65 or less on the sf-36 at the start of the trial selected those patients diagnosed with CFS who were more badly affected, but scores of 60 or more contributed towards a definition of recovery. Even worse, a score of 70 would override any doctor's diagnosis that the patient still had CFS, despite the fact that, at the start of the trial, there were many patients diagnosed with CFS whose scores were 70 or more (these were excluded from the trial).

Remember, initially they were going to take folk with scores of 60 or less, but upped it to 65 to increase numbers.

 

@Dolphin - the normal range analysis they are referring to there is about improvement, not recovery. [ETA - sorry - yes, that was recovery - wow! - so he even forgot he was an author on that paper by Bleijenberg and had to be reminded of it - oops!]

@rvallee - I think Ellen made comments on the Protocol?