• Welcome to Phoenix Rising!

    Created in 2008, Phoenix Rising is the largest and oldest forum dedicated to furthering the understanding of and finding treatments for complex chronic illnesses such as chronic fatigue syndrome (ME/CFS), fibromyalgia (FM), long COVID, postural orthostatic tachycardia syndrome (POTS), mast cell activation syndrome (MCAS), and allied diseases.

    To become a member, simply click the Register button at the top right.

PACE Trial and PACE Trial Protocol

user9876

Senior Member
Messages
4,556
I just came across this ruling from the information commissioner about the disclosure of deteriation data.

http://www.ico.gov.uk/~/media/documents/decisionnotices/2012/fs_50451416.ashx

The ruling seems to be that it is in the public interest to disclose the data and Queen Mary's have 35 calender days from the 26th of september to comply.

4. The Commissioner requires the public authority to take the
following steps to ensure compliance with the legislation.
 To provide the requested information to the complainant.

5. The public authority must take these steps within 35 calendar
days of the date of this decision notice. Failure to comply may
result in the Commissioner making written certification of this fact
to the High Court pursuant to section 54 of the Act and may be
dealt with as a contempt of court.
 
Messages
15,786
I just came across this ruling from the information commissioner about the disclosure of deteriation data.

http://www.ico.gov.uk/~/media/documents/decisionnotices/2012/fs_50451416.ashx

So it looks like they didn't want to release it because it would be "misinterpreted" unless we wait until they can apply the appropriate spin :p

There was no good reason to separate it from the initial publication, and if they really gave a damn about "misinterpretations" they would have included this information in it.
 

user9876

Senior Member
Messages
4,556
So it looks like they didn't want to release it because it would be "misinterpreted" unless we wait until they can apply the appropriate spin :p

There was no good reason to separate it from the initial publication, and if they really gave a damn about "misinterpretations" they would have included this information in it.
I think that is the main point of the information commissioners judgement. I suspect there was surprise that you could publish improvements without deteriations.

However, the Commissioner notes that the
improvement rates were published in an article in The Lancet in
March 2005. Therefore, as some of the information is already in
the public domain, in this case, improvement rates, it is important
that balance is provided.
 

Simon

Senior Member
Messages
3,789
Location
Monmouth, UK
That is a great result, and shows that FOI requests can't always be denied by resorting to spurious arguments. I think it's worth giving the Commissioner's full reasoning for coming down on the side of the complainant: (my bolding)
Balance of the public interest arguments

31. It is the Commissioner’s view that the QMUL’s fear of
misinterpretation is not a valid argument. The data, as provided
to the Commissioner could be used at any point post-publication,
in isolation and out of context. In any event, the QMUL would be
able to provide an explanation to contextualise the information if
necessary.

32. There appears to be little evidence to support withholding the
information requested. The Commissioner notes there is likely to
be a public interest in such information and that disclosure would
provide further information to the public debate on this issue.

33. It is noted that although QMUL has indicated its intention to
publish the data, it has yet to determine a date. The
Commissioner appreciates that clinical trial data is often a
sensitive matter. However, the Commissioner notes that the
improvement rates were published in an article in The Lancet in
March 2005. Therefore, as some of the information is already in
the public domain, in this case, improvement rates, it is important
that balance is provided.

34. He has also again considered his Awareness Guidance which
states that the public interest in releasing the information will
often be stronger if the planned date of publication is far in the
future, or where there is no firm indication of a likely date of
publication. In this case QMUL has not provided a firm indication
of likely publication other than it will be in the foreseeable future.

35. It is the opinion of the Commissioner that although section 22
applies with regard to the matter of future publication, QMUL has
failed to provide substantial arguments relating to the public
interest test.

36. In this case the Commissioner considers that the public interest
in disclosure outweighs the public interest in maintaining the
exemption.
37. The Commissioner has therefore determined that the requested
information should be disclosed.
I'm guessing that the fact the PACE team wrongly claimed they didn't have the data in the first place did not help their case.
 

user9876

Senior Member
Messages
4,556
That is a great result, and shows that FOI requests can't always be denied by resorting to spurious arguments. I think it's worth giving the Commissioner's full reasoning for coming down on the side of the complainant: (my bolding)

I'm guessing that the fact the PACE team wrongly claimed they didn't have the data in the first place did not help their case.
They can still appeal which I suspect they will
 

Simon

Senior Member
Messages
3,789
Location
Monmouth, UK
They can still appeal which I suspect they will
I think you are probably right, and they probably have a case re publishing timescale, but most of the points were independent of timescale e.g that improvement rates have already been published.

Even if they are successful on appeal, I still think it's quite a triumph to win an initial challenge like this - is this the first time a PACE refusal to publish has been over-ruled?
 
Messages
13,774
It would be great to get the recovery rates too. Does anyone know if anyone has put in an FOI request for them?

edit: I want as much data as is possible. It would be great to get all the raw data.

PS: Congratulations and thanks to whoever started this off.
 

Dolphin

Senior Member
Messages
17,567
I gather that the person who achieved this FOI success would prefer to remain anonymous, but it was someone on this forum. Brilliant stuff!

So it just goes to show that we can all have a go at it! Who is next?
I'm in - if they'll give me the Derby winner for 2013.
 

Simon

Senior Member
Messages
3,789
Location
Monmouth, UK
Fun with Cohen's d
Not sure if anyone is still interested in this dicussion about how to measure effect sizes in PACE, but I've not gone throught those videos posted by Dolphin and maybe I can throw some more light on things:
On the video it doesn't use the SD of the individual change or difference scores, but potentially he's wrong, but I'd like a link to be convinced.
The example Graham had mentioned was before and after a weight watchers/similar with no control group so that was what I discussing. The video talks about two sets of SDs while the change one you mentioned would only be one.
When I watched the video just now he used the example of IQ scores in 2 independent groups (not pre/post in the same group) i.e. different from Graham's example. Calculating Cohen's d using a pooled SD (SDs from both groups) is exactly right in that scenario. It's also analagous to the independent t-test rather than the dependent t-test I mentioned in post 1,957 (can't believe we've kept going this long...).

However, the more interesting case is the one you have then discussed where one has four sets of numbers: before and after treatment for groups 1 and 2 (or two sets of change scores), and I'm not sure what is done in that case - you could be right but I haven't thought it through or recall reading much on it.



ETA: thinking about it, perhaps that is right - the change for a control group is like a baseline score.
Aw, looks like you changed your mind after I agreed with you! Which is a shame as I think you were spot on.

Here's another video using an example I think is similar to PACE, with an experimental group on a low fat diet and control group on a normal diet. The difference between the 2 groups is compared but crucially they they compare the difference between the DIFFERENCE scores i.e. change in weight over the trial, not final weights. So means and SDs are of change scores too, i.e. the pooled SD is pooling the SD for the change in weight for the control group and the SD for the change in weight for the low-fat diet group.


note that the video is actually discussing confidence intervals around mean differences but the principle is the same.

Well, that's my take on it, please say if it doesn't make sense. In my stats exam, at one point I calculated that 3x3=6, so I might be wrong on this too.
 
Messages
15,786
The ruling seems to be that it is in the public interest to disclose the data and Queen Mary's have 35 calender days from the 26th of september to comply.

I think this means it'll be available October 31st, assuming that deadline is when it's disclosed to the public or whoever requested it.
 

Bob

Senior Member
Messages
16,455
Location
England (south coast)

Dolphin

Senior Member
Messages
17,567
Fun with Cohen's d
Not sure if anyone is still interested in this dicussion about how to measure effect sizes in PACE, but I've not gone throught those videos posted by Dolphin and maybe I can throw some more light on things:
When I watched the video just now he used the example of IQ scores in 2 independent groups (not pre/post in the same group) i.e. different from Graham's example.
Ok. But the end of video 1 (at around 4:30), after explaining how Glass's delta can be used to compare two independent groups, he then says he has also seen it used to compare before and after (the same group). The difference then with Cohen's d is that it uses both SDs, so if it was used like glass's d, it would similar (using a pooled mean, like it generally does, rather than the SD of the difference).

ETA: with the IQ study, I don't think he clearly says they can't be before and after means e.g. before and after some training method for IQ tests (say).

So until I hear otherwise, I'm going to presume that's how one would use it for examples like Graham's.

Thanks for the other stuff.