SMILE trial data to be released

I have no idea how that's supposed to work. Then what? Plenty of participants to some of those trials have spoken out. That's the end of it. They speak out, we appreciate the contribution to the conversation and add it to the pile of evidence. That's what happens when participants in those trials are identified. They're anecdotes, not much more can be done, the trial data is much more useful.

It's a really bad idea to make strawman arguments like that. It works within the circlejerk, but it makes them look very unprofessional to independent observers. They may as well have been arguing that alien invasion was a likely consequence of releasing the data for all that that argument was grounded in reality.

motivated intruder? They tried that one with the PACE trial...
 
Excellent result. I realized the children will all be adults by now, given 8 years have passed.

I just want to check a few points
- Bristol did not claim researchers might be harassed if data was released?
- there were no attempts at "character assassination" of "activists" like the "borderline sociopath young men" comment from the PACE trial?
- are your views correctly described in the ruling?

Also any witness evidence? I couldn't say any.

Is there a further appeal step they could take before being forced to release the data?
 
Excellent result. I realized the children will all be adults by now, given 8 years have passed.

I just want to check a few points
- Bristol did not claim researchers might be harassed if data was released?
- there were no attempts at "character assassination" of "activists" like the "borderline sociopath young men" comment from the PACE trial?
- are your views correctly described in the ruling?

Also any witness evidence? I couldn't say any.

Is there a further appeal step they could take before being forced to release the data?

1. They did refer to this but it was not part of the exemptions they claimed.

2. No.

3. Not got all the nuance, but it's how people tend to understand what we say.
 
At a very quick glance at that data, there seem to be significant numbers of missing data, data that makes no sense (like someone scoring 100 on SF PF and 0 on CFQ), people who weren't really sick in the first place (over 80 on SF PF), and people who misinterpreted the CFQ (on the 33 point likert scale, a healthy person should score 11, yet some here score much less).
Edit to add: And several people scoring things like 22.22 on SF PF which as I understand it goes in steps of 5.
 
Last edited:
100 on SF36-PF and 0 on CFQ would be someone filling in the forms as positively as possible - that makes sense doesn't it? Or do you mean at baseline?

Sometimes with SF36-PF when there is no answer to a question they use the average score to the answers given rather than counting the missing answer as zero.

I can't see the school attendance data that's based on school records - that was their pre-specified primary outcome, so getting results for that is pretty key. Am I being an idiot or is that data missing?
 
Last edited:
100 on SF36-PF and 0 on CFQ would be someone filling in the forms as positively as possible - that makes sense doesn't it? Or do you mean at baseline?
No, you're right, it's me not making sense. Too much staring at numbers. I'll delete that bit. Though theoretically 0 on CFQ means scoring better than a healthy person on all aspects of fatigue, since 11 is the score for healthy levels of fatigue. That's some brainwashing!
 
@JohnTheJack: Do you know if the data on school attendance provided is that which allows to calculation of the trial's presepecified primary outcome (ie: that from school records, rather than just the participant's self-report)?

I was expecting there to be two sets of school attendance data, but can only see one.

Thanks for all your work on this.
 
Correlations between the different variables may be interesting. I've done some simple plots to show correlation on the non-economic data. Note I've failed to add a title so the first image is for LP and the second for SMC

I'm not sure what to make of them but it looks like the SMC is what you would expect with correlations between the same measures over different time periods. Hads and anxiety scores seem to correlate between them but not as much else where.

Note we expect strong negative correlation between CFQ and SF36-pf as one goes down and the other one goes up.

With LP the baseline vars seem less well correlated with the other values which is interesting as it reflects the improvements in reported scores. Of course we don't know when anyone actually went to the LP course and this makes data hard to interpret.
LPcorr.png SMCCorr.png
 
Back
Top Bottom