Barry
Senior Member (Voting Rights)
Which again suggests PACE may not even have been measuring PEM at all. Yet again - we need to see the data!!!
-
New Forum Software Installed (Still a few issues but reopening the forum) More details.
Adverse outcomes in trials of graded exercise therapy for adult patients with chronic fatigue syndrome, 2021, White & Etherington
Barry
Senior Member (Voting Rights)
Interesting that PACE (unless I've missed something, which is possible) only used a 1994 suggestion (cannot call it a definition) of what characterises PEM, rather than the 2007 NICE guideline. I know PACE might have just missed that boat, but weren't some PACE authors involved in that guideline anyway, so must have had good awareness.
Invisible Woman
Senior Member (Voting Rights)
Not to mention that many of us played sports at some time in our lives. Many of those sports are seasonal so there is a period of retraining at the start of the season.
Even those who didn't play seasonal sports would have had to take time out occasionally because of exams, holidays, other illnesses etc and then gone back to their physical exercise or sport of choice.
We know what physical exercise when you haven't done any for quite some time and retraining for seasonal sports feels like.
Having experienced both return to exercise and retraining and how exercise now causes us to feel I think we are better placed to tell them what the differences are rather then the cart before the horse of them telling us.
More than anything else this demonstrates the sheer hubris of them thinking they know best.
Barry
Senior Member (Voting Rights)
See this post:
https://www.s4me.info/threads/adver...ome-2021-white-etherington.20767/#post-347678
pdf attached
Hoopoe
Senior Member (Voting Rights)
Thank you. I somehow missed that post.
Some comments on the CGI scores
The authors say they chose CGI because "Some trials also reported clinical global impression (CGI) change scores of overall health, but rarely interpreted the data regarding deterioration on this measure" which seems arbitrary to me rather than a good reason. They were also given unpublished CGI scores from a trial by Moss Morris et al. CGI scores were not available for all trials. All this seems a little suspicious to me (why not use the studies own data on harms?). The real reason for choosing CGI could be that that this happened to be the outcome that gave the most favorable safety profile across all choices.
Risk of bias
Have they ever admitted before that these studies have such big problems with bias?
The evidence that GET is unsafe
Here the vague possibility that maybe these numerous reports of harm are all due to widespread misdiagnosis is introduced, in the hopes of persuading the reader to believe that maybe if everything was done properly there would be no harm. But to make this explanation work would require believing in too many unlikely things: that somehow, multiple surveys happen to have receive responses in high proportion from patients with illnesses that are misdiagnosed with CFS and despite looking similar to CFS actually have a totally different response to GET. The authors are proposing an absurd scenario here.
"only a third had confirmed CFS" also doesn't mean that only a third had CFS. What is meant is that only a third did not have any exclusionary conditions that would disqualify a patient from a CFS diagnosis with Fukuda criteria. A patient with true CFS and type 2 diabetes would not be counted as confirmed CFS case.
As I recall there is some survey evidence that suggests GET by a trained specialists has the same rate of harm as patients undertaking GET on their own.
Does having post-exertional malaise mean that GET should not be provided?
This is incorrect. The GETSET trial says
But we don't know if the improvement was real or just an artifact of poor controls. It's also not clear whether there was an adequate distinction between postexertional fatigue and postexertional malaise because they seem to be used interchangably while many now view them as two different things.
There is also the usual issue of participant activity levels not being objectively recorded so it's unclear whether they actually increased their activity levels and what effects this had on their symptoms.
I also note the choice of words "no trial evidence" and think of the objective evidence of deterioration in the 2-day CPET literature. Not mentioning this literature is a serious omission in a paper about potential harms of exercise.
What if GET was banned?
They present a false dichotomy of GET equals activity, no GET equals inactivity. Patients are capable of self-managing their own activity levels and GET is not the only possible way to engage with exercise. As far as I know according to the published studies GET doesn't lead to an improvement in activity levels or fitness so one cannot attribute the positive health effects that come with increasing activity levels and fitness to GET.
Some comments on the CGI scores
The authors say they chose CGI because "Some trials also reported clinical global impression (CGI) change scores of overall health, but rarely interpreted the data regarding deterioration on this measure" which seems arbitrary to me rather than a good reason. They were also given unpublished CGI scores from a trial by Moss Morris et al. CGI scores were not available for all trials. All this seems a little suspicious to me (why not use the studies own data on harms?). The real reason for choosing CGI could be that that this happened to be the outcome that gave the most favorable safety profile across all choices.
Risk of bias
Have they ever admitted before that these studies have such big problems with bias?
The evidence that GET is unsafe
Here the vague possibility that maybe these numerous reports of harm are all due to widespread misdiagnosis is introduced, in the hopes of persuading the reader to believe that maybe if everything was done properly there would be no harm. But to make this explanation work would require believing in too many unlikely things: that somehow, multiple surveys happen to have receive responses in high proportion from patients with illnesses that are misdiagnosed with CFS and despite looking similar to CFS actually have a totally different response to GET. The authors are proposing an absurd scenario here.
"only a third had confirmed CFS" also doesn't mean that only a third had CFS. What is meant is that only a third did not have any exclusionary conditions that would disqualify a patient from a CFS diagnosis with Fukuda criteria. A patient with true CFS and type 2 diabetes would not be counted as confirmed CFS case.
As I recall there is some survey evidence that suggests GET by a trained specialists has the same rate of harm as patients undertaking GET on their own.
Does having post-exertional malaise mean that GET should not be provided?
This is incorrect. The GETSET trial says
But we don't know if the improvement was real or just an artifact of poor controls. It's also not clear whether there was an adequate distinction between postexertional fatigue and postexertional malaise because they seem to be used interchangably while many now view them as two different things.
There is also the usual issue of participant activity levels not being objectively recorded so it's unclear whether they actually increased their activity levels and what effects this had on their symptoms.
I also note the choice of words "no trial evidence" and think of the objective evidence of deterioration in the 2-day CPET literature. Not mentioning this literature is a serious omission in a paper about potential harms of exercise.
What if GET was banned?
They present a false dichotomy of GET equals activity, no GET equals inactivity. Patients are capable of self-managing their own activity levels and GET is not the only possible way to engage with exercise. As far as I know according to the published studies GET doesn't lead to an improvement in activity levels or fitness so one cannot attribute the positive health effects that come with increasing activity levels and fitness to GET.
Last edited:
This is a very important point. I see so many CBT/GET proponents who present it as: "you either start GET and CBT, gradually raise your activity level and recover, or you lay in the darkness 24 hours a day for the rest of your life - the choice is yours!". And most of the time they aren't even challenged on the ridiculousness of this assumption! As if these two extremes are the only possible alternatives.
There's also the alternative of keeping your activity levels moderate and stable over time, without forcing an increase in activity. This would avoid both the detrimental effects of total bedrest and of the possible damage incurred by regularly triggering PEM. The assumption that the alternative to GET is total bedrest 24 hours a day is a strawman argument, but it's one that they get away with suprisingly often.
Lucibee
Senior Member (Voting Rights)
The first thing I thought when I saw the abstract of this paper was this:
Having now seen the paper, I don't think they provide an adequate explanation.
although they do acknowledge it:
but is that just for convenience?
There were more drop-outs in the GET group in PACE, but also more pts who failed to complete the 6mWT at 52 weeks (who didn't drop out). I think I've shown elsewhere (yep, it's here) that it is likely that this was at least partly correlated with lower physical function at 52 weeks.
But the main thing is that White was PI (or at least involved) in these trials. He could have directly asked pts *why* they dropped out, had he wanted to. And he could have directly asked about harms in a follow-up questionnaire. But he hasn't. This is an entirely vague and hand-wavey way of not answering that question.
eta: There's no real substitute for asking pts what they think: https://bmjopen.bmj.com/content/9/6/e021959 and https://www.patientcentra.com/patient-recruitment-insights/why-patients-drop-out-of-clinical-trials
Having now seen the paper, I don't think they provide an adequate explanation.
although they do acknowledge it:
but is that just for convenience?
There were more drop-outs in the GET group in PACE, but also more pts who failed to complete the 6mWT at 52 weeks (who didn't drop out). I think I've shown elsewhere (yep, it's here) that it is likely that this was at least partly correlated with lower physical function at 52 weeks.
But the main thing is that White was PI (or at least involved) in these trials. He could have directly asked pts *why* they dropped out, had he wanted to. And he could have directly asked about harms in a follow-up questionnaire. But he hasn't. This is an entirely vague and hand-wavey way of not answering that question.
eta: There's no real substitute for asking pts what they think: https://bmjopen.bmj.com/content/9/6/e021959 and https://www.patientcentra.com/patient-recruitment-insights/why-patients-drop-out-of-clinical-trials
Last edited:
Sly Saint
Senior Member (Voting Rights)
precisely. Similar scenario in IAPT re 'recovery' (as pointed out by Scott on cbtwatch) why don't they just ask the patients directly.
I think we all know why
Andy
Retired committee member
Trial By Error: More Disinformation from Professor White in Journal of Psychosomatic Research
https://www.virology.ws/2021/06/10/...r-white-in-journal-of-psychosomatic-research/
https://www.virology.ws/2021/06/10/...r-white-in-journal-of-psychosomatic-research/
While the GET group did post a statistically significant improvement in one of the measures—the six-minute walking test—the marginal reported benefits were clinically insignificant.
Not a stats guy, so could be talking complete shite, but...
PACE used four objective outcome measures, for four different arms, which gives 16 separate results, only one of which reached statistical significance.
Using p = 0.05, we would expect 1 in 20 measures to reach significance purely by chance.
1 in 16 is not far from that.
Not a stats guy, so could be talking complete shite, but...
PACE used four objective outcome measures, for four different arms, which gives 16 separate results, only one of which reached statistical significance.
Using p = 0.05, we would expect 1 in 20 measures to reach significance purely by chance.
1 in 16 is not far from that.
Yes they should have done a correction for the number of primary outcomes.
There is also the issue of dropouts.
Yes, and the issue of the six-minute walking test being only partially an objective test. There's plenty of scope for making an extra effort, or not.
Sly Saint
Senior Member (Voting Rights)
"In fact, the PACE trial—..—included four objective outcomes.
None of them matched the subjective claims of improvement.
While the GET group did post a statistically significant improvement in one of the measures—the six-minute walking test—the marginal reported benefits were clinically insignificant.
At the end of PACE, GET participants still performed much more poorly on this measure than healthy women from 70 to 79 years old as well as patients with pacemakers, Class II heart failure, cystic fibrosis and other major health conditions.
They were also no more likely to be working, no more likely to be off social benefits, and no more physically fit than before, as measured by a step-test.
These poor objective results are unmentioned by Professor White in the new paper—a telling omission."
worth repeating.
None of them matched the subjective claims of improvement.
While the GET group did post a statistically significant improvement in one of the measures—the six-minute walking test—the marginal reported benefits were clinically insignificant.
At the end of PACE, GET participants still performed much more poorly on this measure than healthy women from 70 to 79 years old as well as patients with pacemakers, Class II heart failure, cystic fibrosis and other major health conditions.
They were also no more likely to be working, no more likely to be off social benefits, and no more physically fit than before, as measured by a step-test.
These poor objective results are unmentioned by Professor White in the new paper—a telling omission."
worth repeating.
Right, only 3/4 did the six-minute walking test by the end. There is so much to criticize. Hard to include everything in a letter. I decided to focus on factors that would undermine White's claim of "effectiveness" sufficiently to require a correction.
daftasabrush
Senior Member (Voting Rights)
Self-ratings of Clinical Global Impression (CGI) change scores of 6 or 7 (“much worse” or “very much worse”),
So, getting worse does not count as worse in fact? This sounds like bimodal scoring a la PACE trial again.
Null results - are they really what is worth funding? What would be the results for:
No change +
Worse +
Much Worse +
Very Much Worse?
Control intervention
What?? If it's a control then what exactly is the intervention and how do we know it is actually safe?
Remember that Specialist Medical Care (SSMC) in the PACE trial was mostly provided by Simon Wessely, and used as a false comparison?
Probably safe
Authors shoot themselves in the foot here, accepting that GET may not be safe after all. After 14 years of GET "probably safe" is not good enough.
When CFS definitions including more non-ME patients than ME patients...
Low on GRADE
NICE won't be impressed, and we can be pretty sure it's more likely to be very low evidence.
Full FOIA needed for data please
Included studies will not doubt include this gem from the FINE trial:
"The bastards just don't want to get better"
Not exactly something that sounds like a situation involving safe treatment.List of excluded studies, anyone?
daftasabrush
Senior Member (Voting Rights)
Didn't White "retire" from CFS research?
Peter T
Senior Member (Voting Rights)
Not only did he retire (an aside, it would be interesting to measure the rate of articles published pre and post retirement, and similarly with the number of his talks for insurance companies) but also his retirement has been used to justify rejecting freedom of information requests for the PACE data, which can no longer be accessed without him. In totally unconnected news, White with others is still publishing new articles using the PACE data.