A general thread on the PACE trial!

[Barry]

One participant in the original trial has contacted me:

“I was determined to be a part of the...trial because I wanted to get better—so if this ‘treatment’ could make me better I wanted to give it the chance to do so. I was assigned Graded Exercise Therapy. It never occurred to me that it would actually make me more ill. Nor did it occur to me that decline would not be documented, and that despite patients not recovering (or in some cases worsening), they would publish that the treatment was successful...It was stressed that I would only get better if I tried harder, and even though the graded exercise was clearly making me worse, my struggle and pain was dismissed.”

https://hansard.parliament.uk/commo...-4566-940D-249F5026FF73/PACETrialPeopleWithME

[My bold]

We have here a PACE trial participant clearly stating that their decline due to GET was not documented. If verifiable, this is surely a highly unprofessional act, possibly illegal, and at the very least profoundly unethical. I have often wondered whether this kind of thing might be a key reason the PACE authors fight so hard against releasing more data; does it hold evidence not just of poor trial methodology, but also perhaps of gross misconduct. Evidence which they are desperate to suppress?

Tagging @dave30th in case of any use/relevance. To me the text I have boldened seems highly significant, and possibly not homed in on?

 

[Trish]

If verifiable, this is surely a highly unprofessional act, possibly illegal, and at the very least profoundly unethical. I have often wondered whether this kind of thing might be a key reason the PACE authors fight so hard against releasing more data; does it hold evidence not just of poor trial methodology, but also perhaps of gross misconduct. Evidence which they are desperate to suppress?

Given the situation described by the GET participant, there will be no record in the data of that person's accurate response to GET, so no evidence of misconduct will be there in the data to be suppressed.

It would be fascinating if another group of researchers could be permitted to contact all the PACE participants and ask them, in confidence, whether they reported their true symptoms, or what they thought the therapists wanted them to report, or what they had been temporarily brainwashed into thinking.

Isn't there a plan for a 10 year follow up questionnaire based study? If that could be run by an independent team, and made clear to participants that it was independent, maybe some useful feedback could be obtained.

 

[Barry]

Given the situation described by the GET participant, there will be no record in the data of that person's accurate response to GET, so no evidence of misconduct will be there in the data to be suppressed.

My interpretation is a bit different. The participants would not really know what the researchers did or did not record into the raw data, and so I suspect when this participant spoke of it not being "documented", they really meant it was not documented into the trial's publication. The authors would have been able to filter what data they heeded or not after the trial. The authors may not even be sure themselves what incriminating data may lurk in some of the deeper reaches of the raw data - some researchers and assistants may have been more diligent than others at noting events, and so ended up in the database.

 

[Trish]

Any data we would have access to with FOI requests, or data shared with other researchers would, I think, only be at the level of scores on the various questionnaires, ie just numbers or categories (better, much better etc).

In other words, all the raw data anyone would see is the numerical or categorical data used as the basis of the statistical analyses.

I doubt anyone will ever get permission to see the written records of therapists, or to hear the recorded interactions with patients, if any of these still exist. Patient confidentiality would, surely, preclude this.

 

[Valentijn]

In other words, all the raw data anyone would see is the numerical or categorical data used as the basis of the statistical analyses.

I doubt anyone will ever get permission to see the written records of therapists, or to hear the recorded interactions with patients, if any of these still exist. Patient confidentiality would, surely, preclude this.

An exception might be made if there is an official investigation of fraud. Though that would presumably be done by parties appointed by a court or other authority, or under the direction of lawyers party to a lawsuit, subject to strict confidentiality requirements regarding identifiable details of the contents.

 

[alex3619]

[My bold]

We have here a PACE trial participant clearly stating that their decline due to GET was not documented. If verifiable, this is surely a highly unprofessional act, possibly illegal, and at the very least profoundly unethical. I have often wondered whether this kind of thing might be a key reason the PACE authors fight so hard against releasing more data; does it hold evidence not just of poor trial methodology, but also perhaps of gross misconduct. Evidence which they are desperate to suppress?

Tagging @dave30th in case of any use/relevance. To me the text I have boldened seems highly significant, and possibly not homed in on?

Allegations like this arose at the time of the PACE trial, but were difficult to document - I was aware of one such claim. If there was one, there might be more. In this claim their statement that they were worse was deliberately written down wrong, they could see what was written. If there is ever a public enquiry then the opportunity will exist for these patients to give testimony. I think such an enquiry is overdue.

 

[alex3619]

The participants would not really know what the researchers did or did not record into the raw data, and so I suspect when this participant spoke of it not being "documented"

See my comment above. At least one did claim to see what was written, and if we could confirm this or arrange public testimony, or maybe some kind of protected testimony, then it would be damning.

 

[alex3619]

What the BPS cult have done is construct a superficial 'logic' that eliminates a priori the possibility that they might be exposed as being wrong. Any time the 'therapy' doesn't work, it is because of the patient's pathological attitude. Being a BPS cultist means never having to admit you are wrong.

This is the definition of nonscience that Popper created in respect to Freudian psychotherapy.

 

[Barry]

Any data we would have access to with FOI requests, or data shared with other researchers would, I think, only be at the level of scores on the various questionnaires, ie just numbers or categories (better, much better etc).

In other words, all the raw data anyone would see is the numerical or categorical data used as the basis of the statistical analyses.

I doubt anyone will ever get permission to see the written records of therapists, or to hear the recorded interactions with patients, if any of these still exist. Patient confidentiality would, surely, preclude this.

I may have originally misunderstood what I read about the PACE trial data, when it said that all patient identifying data would be recorded into a separate database. I assumed this would have been sensibly designed, so that each participant would have been represented in the anonymised database(s) (let's call it db-X) only by some anonymising identification key. That key, along with the personally identifying details (name, address, NI number, etc, etc), I assumed would then be stored into the other, personal details database (let's call it db-Y), that was supposed to be held completely separately.

In which case db-X could hold all the salient participant data, without any personally-identifying data being held in that database. Each participant would only be referenced within db-X by a completely anonymous identification key. So long as db-Y is never released, no one can connect between the key and the person.

 

[Barry]

Re my post #69, I've just realised where my assumptions went wrong, but it should not have had to. The protocol statement regarding data confidentiality was changed.

In here ...

... it said ...


... but in here ...



... it was changed to merely say ...



... thereby precluding easy sharing of trial data.

 

[alex3619]

In which case db-X could hold all the salient participant data, without any personally-identifying data being held in that database. Each participant would only be referenced within db-X by a completely anonymous identification key. So long as db-Y is never released, no one can connect between the key and the person.

To my understanding that is standard practice. To deviate from that would require a good reason, or lack of competence.

 

[Esther12]

... it was changed to merely say ...

It could be that they just condensed things for the publish protocol. That section of the full protocol you quoted does make it sound as if releasing anonymised data should be easy though.

 

[Barry]

That section of the full protocol you quoted does make it sound as if releasing anonymised data should be easy though.

Which is what I've been saying albeit with inadequate clarity, in various posts/threads along the way.

 

[janice]

I feel this thread is going to be very useful.

I've got until 23rd March to get a list of useful stuff to discuss with my MP.

I am very aware that I can get very angry (and possibly "overwrought" ? TeHe ) when I read all this poor science and know how badly it's affected my life and that of so many others.

I am determined to make sure I present a cool and clear argument to my MP and I feel this brilliantly informed forum and thread will help me do just that.

 

[alex3619]

Since its so hard for them to release bits of data, given no staff, ask for the whole anonymized database?

 

[Barry]

Since its so hard for them to release bits of data, given no staff, ask for the whole anonymized database?

Exactly. The key point is 'database', not all the data extracted from the db into excel or something. We do not want the 'full' data that has been extracted via a SQL query of their own construction, because that query could itself filter what we get ... and more crucially, what we don't get. Just get the whole database, and let others run queries into it ... who knows what might turn up.

 

[Dolphin]

Exactly. The key point is 'database', not all the data extracted from the db into excel or something. We do not want the 'full' data that has been extracted via a SQL query of their own construction, because that query could itself filter what we get ... and more crucially, what we don't get. Just get the whole database, and let others run queries into it ... who knows what might turn up.

The more data that is looked for in one data set, the easier it is for them to claim participants are identifiable and hence they shouldn't release the data.

 

[Barry]

The more data that is looked for in one data set, the easier it is for them to claim participants are identifiable and hence they shouldn't release the data.

But they have stated in the PACE protocol that all personally identifying data is confined to a separate, different database. Or is the argument that even within the non-personal-details db (i.e. the main trial db) there can be data that as a whole that could indirectly identify a person to those who know them?

 

[Dolphin]

But they have stated in the PACE protocol that all personally identifying data is confined to a separate, different database. Or is the argument that even within the non-personal-details db (i.e. the main trial db) there can be data that as a whole that could indirectly identify a person to those who know them?

The latter. That's the argument they tried to use to block Alem's request.

 

[Kalliope]

The message is getting through. It is lovely to see the Lancet ridiculed like this