PACE trial: fraud or incompetence?

[Laelia]

@Valentijn I'm just copying this comment you made on a different thread over to this one because it's more on topic here:

"Fraud is a crime. Do you have evidence that someone committed it, other than circumstance? And there are plenty of ways to discuss improper conduct without using a word which labels it as a crime."

It seems that the words "fraud" and "misconduct" are sometimes used interchangeably. For example in the definition used by the Medical Research Council, as posted earlier by @adreno:

"The Medical Research Council (MRC) definition of misconduct and fraud (or a variation of the MRC code) is widely used. This code states the following definition..." (my bold)

The MRC definition is referring to scientific misconduct which, as I'm sure you know, is not a crime. So going by their definition, use of the word 'fraud' when referring to scientific research does not necessarily label it as a crime.

 

[Valentijn]

It seems that the words "fraud" and "misconduct" are sometimes used interchangeably.

Sometimes. But fraud is also a lot more specific, whereas misconduct is a very broad term. Misconduct can indicate inappropriate but legal behavior, whereas fraud is usually only used in the context of illegal behavior.

 

[Laelia]

whereas fraud is usually only used in the context of illegal behavior

Yes, usually but not always. I really only wanted to make the point that use of the word fraud when referring to scientific research does not necessarily label it as a crime.

 

[SilverbladeTE]

it becomes a crime when that research is used to force unsafe treatments on patients and unwitting doctors are made accomplices to this in their genuine ignorance.

If you KNOW a treatment can cause serious harm, even death, but cover that up, and it's forced on millions of people, then it is a very serious crime indeed.
For example, if you altered research on diabetes and said that "eating increasing amounts of sugar is good for you!", that would result in a lot of people dying and suffering serious complications and therefor, be fraud

That is NOTHING like fiddling with a paper on some esoteric part of Particle Physics which has no financial or Human impact.
this is all about research which affects Practical application of the art of Medical Science
A genuine mistake can cause great harm in the application of medical treatments
alas, Medical Profession is littered with historical proof of outright frauds and grotesque wilful stupidity because come folks preferred to be wilfully blind/ignorant to evade responsibility of their actions and from financial/other rewards form these acts.

this sums up the vile system we all really live in, Corporations and those with power evade personal responsibility time and time again for heinous crimes
this is part of the "OPREN SCANDAL"

http://articles.latimes.com/1985-08-22/business/fi-2455_1_liver-failure

Eli Lilly Pleads Guilty of Failure to Disclose Oraflex-Linked Deaths
August 22, 1985|MARLENE CIMONS | Times Staff Writer

WASHINGTON — Eli Lilly & Co. pleaded guilty Wednesday to federal charges that it failed to tell the government about deaths and toxic reactions associated with its arthritis drug Oraflex, subsequently linked to dozens of fatalities in the United States.
The company, which also was charged with failure to label the drug with information describing its potential side effects, was fined $25,000.
Dr. William Ian Shedden, former vice president and chief medical officer of Lilly Research Laboratories, pleaded no contest to separate criminal charges and was fined $15,000.
Under an agreement worked out in advance with the U.S. attorney's office in Indianapolis, where Lilly is based, federal prosecutors filed 25 misdemeanor charges against the company and 15 against Shedden, and both defendants immediately entered their pleas. U.S. District Judge S. Hugh Dillin fined the company and Shedden $1,000 for each count.

"Misdemeanours" how the hell can that be a minor crime, eh?
and that is one of many such scandals
but after that the scumbag pharma corporations brought in draconian "gagging clauses" into their employee contracts in later 1980s, they said it was to prevent industrial espionage, the reality was it was all about silencing whistle-blowers

 

[trishrhymes]

For me the difference might be something like this:

For misconduct you may lose your job.

For fraud you may lose your liberty.

 

[Laelia]

How can manipulating data ever be unintentional? And surely manipulating data is a form of deceit?

On that basis one should only need to prove that data was manipulated in order to prove fraud had taken place. Which doesn't seem that tricky to me.

But perhaps I'm missing something here?

 

[Molly98]

For me the difference might be something like this:

For misconduct you may lose your job.

For fraud you may lose your liberty.

and for both you loose both with brass knobs on

 

[alex3619]

How can manipulating data ever be unintentional? And surely manipulating data is a form of deceit?

Manipulating data is a routine occurrence in science. The issue is between appropriate manipulation of data and manipulation that leads to deliberate or unintentional bias. Its only deliberate manipulation for bias that is possibly fraud.

Take p-hacking. If you have an idea about something and need to restructure the data, and suddenly find something interesting, is that misconduct or fraud? Yet if you start a program of systematic manipulation to derive a better p value, a value for statistical significance, that is called p-hacking. Its inappropriate. Its not clearly fraud, but it is scientifically dubious.

Intent is always hard to prove.

We know there was at least one instance of deliberate and inappropriate manipulation of data to produce a biased finding in PACE that we can probably prove, involving the mathematically inappropriate calculation of normal using SF36PF data. The others are all conjecture unless we can find hard evidence. Once might be enough to demand some kind of action, but who do we demand it of, who is responsible for rectification of the scientific record, etc. ? Its not like we have a formal court of good science. Hence the calls for retraction from the publishers.

 

[Laelia]

Manipulating data is a routine occurrence in science. The issue is between appropriate manipulation of data and manipulation that leads to deliberate or unintentional bias. Its only deliberate manipulation for bias that is possibly fraud.

Yes, I see what you're saying, I didn't phrase that very well! I was of course referring to the innapropriate manipulation of data in my previous post

I'm having some problems with the multiquote feature so I will respond to your other comments in a separate post.

 

[Laelia]

We know there was at least one instance of deliberate and inappropriate manipulation of data to produce a biased finding in PACE that we can probably prove, involving the mathematically inappropriate calculation of normal using SF36PF data.

That's good to hear. Perhaps we're not that far off proving that scientific misconduct took place in that case. Do you think that this the strongest piece of evidence we have for proving scientific misconduct?

Once might be enough to demand some kind of action, but who do we demand it of, who is responsible for rectification of the scientific record, etc. ? Its not like we have a formal court of good science. Hence the calls for retraction from the publishers.

I notice you mentioned the Andrew Wakefield case earlier in this thread. I believe it was the General Medical Council (GMC) that conducted the enquiry in this case and found him guilty. Perhaps we could press the GMC for an inquiry into the PACE trial?

 

[alex3619]

Do you think that this the strongest piece of evidence we have for proving scientific misconduct?

Its the one that possibly shows misconduct, the others could just be really really bad science.

Perhaps we could press the GMC for an inquiry into the PACE trial?

That would be the case, but given the nature of the GMC I suspect we would require a complaint from a UK doctor in order to have any weight.

I never bother with multiquote, I just keep quoting the bits one at a time until I am done. I find that easier.

 

[Laelia]

Its the one that possibly shows misconduct, the others could just be really really bad science.

What about all the other pieces of evidence people have put forward in this thread? Surely some of these things are relevant and would have to be taken into account?

That would be the case, but given the nature of the GMC I suspect we would require a complaint from a UK doctor in order to have any weight.

There must be a doctor in the UK who would support us with this. What about Dr Myhill for example?

 

[Laelia]

I never bother with multiquote, I just keep quoting the bits one at a time until I am done. I find that easier

I'm getting mixed up, it's not the multiquote I'm having problems with, it's the ordinary quote button. I do the same as you, quoting bits one at a time but sometimes when I select the text I want to quote the reply button does not pop up. Not sure why that is.

 

[alex3619]

What about all the other pieces of evidence people have put forward in this thread? Surely some of these things are relevant and would have to be taken into account?

First we have to establish misconduct, and then the other things can be taken into account with the new perspective. This is not just a matter of probabilities or reason, we will be trying to establish things in a setting in which the investigators have allegiances, friends, allies and in a cultural setting of protecting the medical profession. This is as much politics as science or reason, and probably more about politics than science or reason. The scientific route mostly just leads to retraction. Making a complaint with the GMC or looking for broader correction of the things that are wrong will be heavily involved in UK politics. We have seen how that goes.

 

[Laelia]

First we have to establish misconduct

Oh I was thinking we had already established that, but perhaps you mean that we can establish it further?

First we have to establish misconduct, and then the other things can be taken into account with the new perspective. This is not just a matter of probabilities or reason, we will be trying to establish things in a setting in which the investigators have allegiances, friends, allies and in a cultural setting of protecting the medical profession.

Are you saying that it's too early to put a case together to put forward to the GMC? Because we are still trying to establish the facts and the data is still being scrutinized?

 

[Laelia]

There must be a doctor in the UK who would support us with this.

On second thoughts I'd be surprised if we couldn't gather a whole army of UK doctors to support us in this.

[Edit: To support us in pressing the GMC for an inquiry into the PACE trial that is]

 

[RogerBlack]

it becomes a crime when that research is used to force unsafe treatments on patients and unwitting doctors are made accomplices to this in their genuine ignorance.

To restate this in a somewhat different manner - it becomes a crime when it is legally a crime, not when we would wish so.

I note that Andrew Wakefield, despite having his key paper retracted for fraud, despite doing procedures on children for which he had no authority has only suffered loss of his medical licence and no criminal prosecution, despite, reasonably arguably causing many thousands of deaths.

Unfortunately prosecution for harms to patients they did not personally cause are almost impossible due to the multiple layers between them and the patient.
Prosecution for patients directly harmed by PACE is slightly less unlikely.

Prosecution for direct harms would require unearthing of some statement (or him coming out and saying) that he intentionally harmed people with CFS/ME because he hates them.
Prosecution for indirect harms would require unearthing a plan to intentionally harm patients through altering regulations and policies in ways he knows would be harmful.

Being negligent as to possible harm might barely possibly be enough for direct harms, but indirect ones are just too far removed from the trial.

I use 'he' here, when the additional problem with criminal action would be that there are many authors of the PACE trial, and for the most serious charges, there would need to be one or several people knowing that it was a fraud and intending to hurt patients with no reason, and acting to convince the others who would not have gone ahead otherwise.

Knowing that some patients might be made worse, in the belief that you're doing the whole group a favour is medicine as usual.

 

[Laelia]

There is a good argument for the criminalisation of scientific misconduct:

https://www.newscientist.com/articl...to-criminalise-serious-scientific-misconduct/

 

[RogerBlack]

There is a good argument for the criminalisation of scientific misconduct:

https://www.newscientist.com/articl...to-criminalise-serious-scientific-misconduct/

Indeed it is.
The problem with PACE is more subtle though.
I question if anyone on the PACE team believes they are doing anything wrong.
Their aim is not to mistreat patients, but to make others see the world as they believe it to be.

The right way to fix this is not to try to prosecute people after the fact for the narrow gap between reasonable error and unreasonable error, but:
Ensure that peer review looks at all of the paper, not just the bits the reviewers happen to pick up on.
Look at the statistics to see they are robust - without implicit assumptions that SF36 is linear, or misstatements about the population levels.
Examine trial protocols and adherence to them beforehand for risk of bias.
Ensure pre-specified analysis are done, and put up front in the abstract, before any post-hoc stats-hacking.
Check any measures claimed are reasonable for the disease state, and only accept surrogates unwillingly. (for example, claiming your patients are fully recovered, when they are still bedbound because you've convinced them that they're better, rather than looking at employment and other measures).

These measures, adhered to, would mean that PACE would have come out with the headline 'CBT and APT are not significantly better than SMC' - perhaps with a later authors interpretation section with major red flags on it. This would have shut down all of the later CBT trials before they started.

 

[Laelia]

The problem with PACE is more subtle though.
I question if anyone on the PACE team believes they are doing anything wrong.
Their aim is not to mistreat patients, but to make others see the world as they believe it to be.

I think this still qualifies as scientific misconduct because there is intention to deceive. There does not need to be the intention to mistreat patients. If I have understood this correctly (and please someone correct me if I haven't) then on that basis you are arguing against the current definition of scientific misconduct? Is that correct?