Trial By Error: QMUL and FOI; Nature and Cochrane; the Pineapple Fund

Kalliope said:
I once had a brief encounter with Lillebeth Larun from the Cochrane review group. She held a lecture about GET and CBT at my local hospital and we had a chat in the break. When discussing the PACE-trial, she said that she was one of the few who had access to the data from the trial (and therefore knew there were no serious flaws with it).

If this is true, could Larun be a way in to get the data?
Click to expand...

Kings and Oxford will also have the data. I don't think Larun would be in a position to give data since it would be given to her under a confidentiality agreement. Of course if they believed the data contained personal information they would have breached the data protection act by giving it to her.

But the fact they gave it to her and she managed to look at it suggests that there is sufficient documentation for others to interpret it and QMUL have data scientists working there who could use that information to extract data.
 
Adrian said:
Kings and Oxford will also have the data. I don't think Larun would be in a position to give data since it would be given to her under a confidentiality agreement. Of course if they believed the data contained personal information they would have breached the data protection act by giving it to her.

But the fact they gave it to her and she managed to look at it suggests that there is sufficient documentation for others to interpret it and QMUL have data scientists working there who could use that information to extract data.
Click to expand...
Good points! She did mention it was under a confidentiality agreement, so it would have to be done in line with the university, as they are incapable of providing the data themselves. But more interesting is your other point about this suggesting there is sufficient documentation for others to interpret it.
 
Kalliope said:
Good points! She did mention it was under a confidentiality agreement, so it would have to be done in line with the university, as they are incapable of providing the data themselves. But more interesting is your other point about this suggesting there is sufficient documentation for others to interpret it.
Click to expand...

I think the ICO may be setting a very dangerous precedent in that QMUL are the institution responsible but they are allowing them to localise the problem to a given team and claim that team didn't exist. QMUL clearly have employees with the skills to access the files and do the work.

Can you imagine if a government department were to respond to an FoI request say for data that may be politically damaging and they said but the person doing that has left or we closed that team down.
 
Lucibee said:
But don't we already know this? Didn't they have a laptop stolen from an unlocked office that contained recordings of patients' interviews? Or was that something else?
Click to expand...

It was a stolen tape recorder for PACE there was a previous incident of a stolen computer at Kings which appeared after a statistician challenged stats in Wessely's paper.

My interpretation was not that they had lost the data but that they couldn't interpret the bits.

From my perspective working in IT security and understanding something about how data sets should be managed it is quite an admission for a university to make that a data set from a trial would only be interpretable by a few individuals. It feels totally unprofessional the data schemas should be well documented (people come and go during projects).

To get the data out should be a case of simply extracting columns from data frames (or what ever they are called in stata) or doing queries into the access database that the data was originally kept in. So their claim that the data is not interpretable suggests to me that they have inadequate documentation over which data frames contain which fields along with which columns are which.

I find this completely horrifying. It suggests a huge degree of amateurism. Its also worrying for the data analysis. If it isn't clear which data is in which table or column then how can the people doing the analysis ensure that they are using the right data and hence check the results are correct. Did anyone QA the results.

From a security handling perspective if they can't show professionalism in one area I would worry that they could keep data secure. Academic institutions have a reputation for poor security because the IT departments find it hard to impose policies on academics who claim they should be free to do what they want.

Of course I don't believe they would have been quite so sloppy with their data handling. I think that QMUL have data scientists who could extract the data probably within a few hours!
 
Adrian said:
I think that QMUL have data scientists who could extract the data probably within a few hours!
Click to expand...

If QMUL (and/or the PACE team) are lying about the data being accessible, wouldn't that be perjury if they did so in a court (I can't remember whether this data request reached a tribunal stage)?

Even if they're merely being evasive, perhaps the next FOI request could be to nail down details of exactly what the problem is supposed to be with the data storage/retrieval, @dave30th? Perhaps there are people here like @Adrian with the expertise to help draw up an undodgeable question.
 
Sasha said:
If QMUL (and/or the PACE team) are lying about the data being accessible, wouldn't that be perjury if they did so in a court (I can't remember whether this data request reached a tribunal stage)?
Click to expand...

I think they have got away with drawing a perimeter around the PACE group and treating them as a separate entity. It would be interesting to know the process by which they went about forming their argument. For example, did they just ask White who said no one could interpret it or did they ask someone in their organisation used to handling data (they have trial support groups, data scientists, computer scientists etc) did they ask, do they know about the capabilities of their own organisation.

In the past they claimed they would need to hire a statistician to calculate the mean of data. Again they have statisticians in QMUL just not associated to the particular project. The ICO is setting an extremely naive precedent with this ruling.
 
Adrian said:
I think they have got away with drawing a perimeter around the PACE group and treating them as a separate entity. It would be interesting to know the process by which they went about forming their argument. For example, did they just ask White who said no one could interpret it or did they ask someone in their organisation used to handling data (they have trial support groups, data scientists, computer scientists etc) did they ask, do they know about the capabilities of their own organisation.
Click to expand...

Now that the data from a £5m, publicly funded trial are no longer available and for all time, I think we do need to ask QMUL these questions. The situation is completely unacceptable, especially when that trial is at the centre of a storm of controversy.
 
It would also be interesting to know the relevant contractual provisions. One would not expect a project lead to be able simply to walk away without having fully briefed someone, especially where there are ongoing contentious matters with potential costs to the organisation.
 
Further, are they telling us that White inputted the all data entry himself by hand? If not, how did any others know which fields to add the data to?

Was the database tested for integrity by a third party? I'd want to make sure the data in my £5 million database was correct.

Even from the starting point it doesn't make sense. Who makes a database without clear (or deduce-able) labeling? I can see it being viable for a single person project, but not one involving a team.
 
I was pondering the question how could QMUL, implicit in their own claims, miss the facts that one of their Proffessors was retiring and that this would mean that they would not be able to access a significant set data even though they were concurrently fighting several prominent freedom of information requests in relation to that data?

For the cynically minded this raises all sorts of questions about whether or not this is an excuse on the part of QMUL to avoid releasing any further data from this highly controversial study: as people have already raised, what is so peculiar about this data or its storage that only the original investigators whilst in the employ of QMUL can access it, is there no duplicate sets of the data held by the other insitutions involved in the research or by the privileged few other academics that were allowed to see it that could be authorised for release by QMUL, how can further articles and PACE follow ups be planned if the original data is no longer accessible, how can Prof White be simultaneously available to head up this follow up under QMUL and not available to access the data for a freedom of information request, etc

Alternatively one could naively give QMUL the benefit of the doubt and wonder if they were some how taken by surprise by the retirement of Prof White. Can we assume that he was due to retire at 65 years or would that be discretionary for a proffessor? If 65 was his retirement age then it would seem he went a matter of months early, aged 64 yrs, leaving both his academic work at QMUL and his clinical practice at Barts. Is it fair or reasonable to ask why he took early retirement? Some suggest this should be asked, see http://niceguidelines.blogspot.co.uk/2016/09/pace-trials-principal-investigator.html . That Prof White continues to be active on the insurance lecture circuit and is listed as the lead in the ongoing PACE work is at best confusing. I do not know enough myself to fairly make any comments on this early retirement, but certainly believe that QMUL have questions to answer as to why they failed to undertake any contingency planning in relation to this internationally significant data.

(Added shortly after posting: QMUL either have been incompetent or are being untruthful, and the FOI process needs to demand which it is. As a commentor on one of the threads raised, could this in itself be the subject of a FOI request?)
 
Last edited:
Peter Trewhitt said:
I was pondering the question how could QMUL, implicit in their own claims, miss the facts that one of their Proffessors was retiring and that this would mean that they would not be able to access a significant set data even though they were concurrently fighting several prominent freedom of information requests in relation to that data?

For the cynically minded this raises all sorts of questions about whether or not this is an excuse on the part of QMUL to avoid releasing any further data from this highly controversial study: as people have already raised, what is so peculiar about this data or its storage that only the original investigators whilst in the employ of QMUL can access it, is there no duplicate sets of the data held by the other insitutions involved in the research or by the privileged few other academics that were allowed to see it that could be authorised for release by QMUL, how can further articles and PACE follow ups be planned if the original data is no longer accessible, how can Prof White be simultaneously available to head up this follow up under QMUL and not available to access the data for a freedom of information request, etc

Alternatively one could naively give QMUL the benefit of the doubt and wonder if they were some how taken by surprise by the retirement of Prof White. Can we assume that he was due to retire at 65 years or would that be discretionary for a proffessor? If 65 was his retirement age then it would seem he went a matter of months early, aged 64 yrs, leaving both his academic work at QMUL and his clinical practice at Barts. Is it fair or reasonable to ask why he took early retirement? Some suggest this should be asked, see http://niceguidelines.blogspot.co.uk/2016/09/pace-trials-principal-investigator.html . That Prof White continues to be active on the insurance lecture circuit and is listed as the lead in the ongoing PACE work is at best confusing. I do not know enough myself to fairly make any comments on this early retirement, but certainly believe that QMUL have questions to answer as to why they failed to undertake any contingency planning in relation to this internationally significant data.

(Added shortly after posting: QMUL either have been incompetent or are being untruthful, and the FOI process needs to demand which it is. As a commentor on one of the threads raised, could this in itself be the subject of a FOI request?)
Click to expand...
Don't fall into their trap. When you take liars, scammers and people acting in bad faith at their word and try to explain it then your only kneecapping your position which is their goal, find an excuse they can win and keeps you in a feedback loop. They are screwed because they committed fraud and are desperately trying to cover for it by any excuse they can think of that will stick. If A won't work, then B, if B fails then C and so on. If we want the data we don't fall for their flawed but effective traps, we figure out our next step. By law its public data, so if they won't release it and the law has no prescribed penalty then perhaps the courts are needed again, or we shame them in parliament or something else (even high profile) that forces them to live up to their legal obligations.
We can't forget another tactic is their militant patient narrative, they have to vilify those they claim to be advocating for because they know they are lying.

I'm not quite with it so i might not be explaining well (and i don't just mean my reply at your post). This is a classic political maneuver straight from US politics, they are attempting to baffle us and everyone else with bull****, keep us spinning our wheels and going off on tangents until they can invent something indisputable that people (or the governments/courts) will accept.
 
Last edited:
Kalliope said:
I once had a brief encounter with Lillebeth Larun from the Cochrane review group. She held a lecture about GET and CBT at my local hospital and we had a chat in the break. When discussing the PACE-trial, she said that she was one of the few who had access to the data from the trial (and therefore knew there were no serious flaws with it).

If this is true, could Larun be a way in to get the data?
Click to expand...

Well there may not be "serious flaws from the data", but its still a question of what the data shows.

Lets say for example a set of data showed 70% of people to suffer adverse long term affects from a treatment and that data is never published, one could still claim the secret data shows no serious flaws.

It would just show inconvenient truths that don't match predetermined conclusions and paradigms.
 
Last edited:
large donner said:
Well there may not be "serious flaws from the data", but its still a question of what the data shows.

Lets say for example a set of data showed 70% of people to suffer adverse long term affects form a treatment and that data is never published, one could still claim the secret data shows no serious flaws.

It would just show inconvenient truths that don't match predetermined conclusions and paradigms.
Click to expand...
True. But in this case she meant there were no problems with any aspects of the PACE-trial, including its outcome. It proves that GET and CBT are safe and effective treatments.
 
Kalliope said:
Lillebeth Larun
Click to expand...
see:
"At a meeting organized and financed by PACE investigator Peter White, Larun obtained privileged access to data that the PACE investigators have spent tens of thousands of pounds to keep most of us from viewing. Larun used this information to legitimize outcome switching or p-hacking favorable to the PACE investigators’ interests. The Cochrane review misled readers in presenting how some analyses were conducted that were crucial to its conclusions."

https://jcoynester.wordpress.com/tag/lillebeth-larun/
 
Sly Saint said:
see:
"At a meeting organized and financed by PACE investigator Peter White, Larun obtained privileged access to data that the PACE investigators have spent tens of thousands of pounds to keep most of us from viewing. Larun used this information to legitimize outcome switching or p-hacking favorable to the PACE investigators’ interests. The Cochrane review misled readers in presenting how some analyses were conducted that were crucial to its conclusions."

https://jcoynester.wordpress.com/tag/lillebeth-larun/
Click to expand...
Thanks. I see your point more clearly now.
 
Kalliope said:
From one of the comments :laugh:

Thank you David for pointing out about Shrodinger's proffessor, who is both not available to work for QMUL in accessing the PACE FIO data and simultaneously available to head up the PACE follow up under QMUL.
Click to expand...

Continuing the feline theme, may I suggest Simon ‘Macavity’ Wessely and Esther ‘Puss-in-Boots’ Crawley?
 
You must log in or register to reply here.
Back
Top Bottom