FOI request re PACE Cost-Effectiveness Analysis

[JohnTheJack]

I have received a reply from the MRC. They tell me to address all my questions to the CSDR.

I went to the site and poked around a bit and now have a better understanding of what has happened. The MRC do not hold the data. This organization does. It's been set up by a group of organizations that fund research, including the MRC and Wellcome.

The decision to grant access is ultimately decided by these five on the 'Independent Review Panel'.

It does seem that patient-level data are available, but as has been pointed out, only under quite strict conditions.

As usual with the smoke and mirrors of PACE, there is less to this than meets the eye, and it doesn't really seem to be of much use to patients.

 

[Trish]

Thanks for all that effort, @JohnTheJack. I wonder if some researchers would be interested in getting hold of the data and producing a paper, for example on the outcomes of the objective tests in PACE - employment data and step test. @Carolyn Wilshire would be the obvious person if she is available and thinks it's worth doing.

 

[JohnTheJack]

Thanks for all that effort, @JohnTheJack. I wonder if some researchers would be interested in getting hold of the data and producing a paper, for example on the outcomes of the objective tests in PACE - employment data and step test. @Carolyn Wilshire would be the obvious person if she is available and thinks it's worth doing.

That would be very interesting, though of course would need researchers with sufficient time to devote to it.

 

[Esther12]

Is there anyway of us seeing what data is available? eg: is the LTFU employment data included? I couldn't see how to find that out, but surely you must be able to find out what's available before starting to devise a proposal?

 

[Graham]

You have to apply to the IRP for access, having submitted a research plan, details of sponsorship etc and presumably agreeing to all the T&Cs. This is designed to exclude the likes of us having a go.

Possibly not. I have been pushing for some detailed data from Bristol for one of the Crawley studies, and they have produced an agreement to sign about not making the data public. I haven't read it yet, so can't be sure, but it looks as though they are recognizing patient researchers.

 

[Adrian]

Possibly not. I have been pushing for some detailed data from Bristol for one of the Crawley studies, and they have produced an agreement to sign about not making the data public. I haven't read it yet, so can't be sure, but it looks as though they are recognizing patient researchers.

The control through this service is quite limited in that they don't provide the data but they provide access to an online stats service so you can compute stats on the data (assuming SAS does what you want). So the researchers never get the data. I don't know if they restrict what can be done or which data is accessable by looking at the analysis plan but they also suggest this.

I'm not sure how much milage there really is from the PACE data (apart from the step data and the long term follow ups) since the methodology is so poor (and I think widely accepted as poor).

 

[Graham]

I'm not sure how much milage there really is from the PACE data (apart from the step data and the long term follow ups) since the methodology is so poor (and I think widely accepted as poor).

Personally, I think we have pretty much nailed the PACE coffin: any further analysis would just be repeating what a bad trial it was: unless, of course, there is some useful and proper analysis of solid data possible along routes that they did not consider (e.g. length of illness vs performance on step test – do folk deteriorate over time? Or step data vs time delay to diagnosis - is a late diagnosis detrimental? Actually this is more of a three-way analysis).

 

[rvallee]

Personally, I think we have pretty much nailed the PACE coffin: any further analysis would just be repeating what a bad trial it was: unless, of course, there is some useful and proper analysis of solid data possible along routes that they did not consider (e.g. length of illness vs performance on step test – do folk deteriorate over time? Or step data vs time delay to diagnosis - is a late diagnosis detrimental? Actually this is more of a three-way analysis).

Pretty much. When you reach the stage of "we changed our methodology and hid our intent to give ourselves better results and it worked so suck it" and it's accepted as a serious academic argument then nothing matters past this point. It reinforces long-held beliefs and that's all that matters, doing anything else would threaten the magical psychology bandwagon, which can't be stopped without acknowledging that harm was, in fact, done on a massive scale, with institutional support (and funding) and informed of the basic facts, all denied because ultimately facts are meaningless in the BPS-ME model.

At some point people will claim they knew all along that it was bad and cite all the reasons we scream about as evidence but for now this is a political/ideological decision and facts don't affect those. In a way SMILE solidified that, yet again acknowledging a pattern of deceit that violated requirements and "zero tolerance" policies, this area of research is exempted from all rules and ethics, even with repeat offenses that are indefensible on their own, hand-waved away with a "reader beware, don't ask me I'm just the editor".