Response to the editorial by Dr Geraghty by the PACE Trial team

[Tom Kindlon]

Free full text: http://journals.sagepub.com/doi/full/10.1177/1359105316688953

Response to the editorial by Dr Geraghty

Peter D White, Trudie Chalder, Michael Sharpe, Brian J Angus, Hannah L Baber, Jessica Bavinton, Mary Burgess, Lucy V Clark, Diane L Cox, Julia C DeCesare, Kimberley A Goldsmith, Anthony L Johnson, Paul McCrone, Gabrielle Murphy, Maurice Murphy, Hazel O’Dowd, Laura Potts, Rebacca Walwyn, David Wilks

DOI: 10.1177/1359105316688953 |

First Published January 24, 2017

This article is written in response to the linked editorial by Dr Geraghty about the adaptive Pacing, graded Activity and Cognitive behaviour therapy; a randomised Evaluation (PACE) trial, which we led, implemented and published. The PACE trial compared four treatments for people diagnosed with chronic fatigue syndrome. All participants in the trial received specialist medical care. The trial found that adding cognitive behaviour therapy or graded exercise therapy to specialist medical care was as safe as, and more effective than, adding adaptive pacing therapy or specialist medical care alone. Dr Geraghty has challenged these findings. In this article, we suggest that Dr Geraghty’s views are based on misunderstandings and misrepresentations of the PACE trial; these are corrected.

https://twitter.com/i/web/status/824017724175163392

 

[dyfalbarhau]

Did the article really need all those contributing authors? Especially when it boils down to "the only correct interpretation of PACE is ours"?

 

[Hajnalka]

People with CFS and/or myalgic encephalomyelitis (ME)

Did I miss something?

Nice, to suggest that Keith is preventing research into ME:

Of course, we need further trials, not only of CBT and GET but also other treatments. To this end, we hope that editorials such as that by Dr Geraghty do not discourage others from doing such research.

At least, while they're busy writing passive-aggressive justifications, they can't cause any other harm.

 

[Esther12]

Did the article really need all those contributing authors? Especially when it boils down to "the only correct interpretation of PACE is ours"?

Maybe done to show that those authors were not backing away from PACE (yet).

 

[actup]

Dr Michael Trump ( sorry I mean Sharpe- keep getting those two mixed up with their "alternative facts") has zero evidence to back up his claims. Small wonder I get them confused. I do suspect though that Sharpe has been reading Trump's book, The Art of the Deal.;- /

 

[Esther12]

Their interpretation of Geraghty in point 9 of their response is a bit odd as in point 5 (where it looks like they caught a genuine slip) they do quote him talking about the requirements of their recovery criteria other than the SF36-PF scale. This is one of those things where the language wasn't perfectly clear - I'm looking foreward to a Geraghty response to clarify matters for White et al. (There were a few parts where is seemed White would benefit from some clarification!)

It's hilarous watching them now try to play down their own complaints about vexatious FOI requests. Surely most people in UK medicine are going to have noticed their 'woe is me' weeping on this. White was campaigning to get the FOI Act changed!

 

[Esther12]

I was reminded of White's submission re the FOIA:

The PACE trial is a trial of four treatments for patients suffering from chronic fatigue syndrome (CFS), which is sometimes called myalgic encephalomyelitis (ME). CFS is a controversial condition, and attracts a strong patient activist voice. This group has used the FoI Act many times since this main paper was published, asking for all sorts of data, from the minutes of all meetings overseeing the trial to all of the patient data collected. See the following for a description of activism (http://www.bmj.com/content/342/bmj.d3780)

...

Section 22a of the Act is insufficient protection for science into controversial subjects, and requires that the research is on-going, so is irrelevant to completed research. We need science in the UK to be protected or it will continue to be damaged as this trial has been (other examples include climate change science, and research into the health effects of tobacco). Exempting Universities from the FOIA would achieve that. Exempting scientific research data produced by Universities and other higher educational institutes might be a workable alternative.

http://forums.phoenixrising.me/inde...st-and-calling-for-changes-to-foi-laws.42027/

(He also attempts to imply the participant requests to destroy their data were a result of concern about FOI requests, when that does not seem to have been the case).

 

[Esther12]

I knew there was another thing that pissed me off:

We reject the accusation that our ‘actions have arguably caused distress to patients’, for which Dr Geraghty offers no evidence.

I wonder what Alem Matthees thinks about that. Or any of the 12,000+ people who've signed a petition calling for the retraction of misleading PACE claims. Or me!

 

[Snowdrop]

I knew there was another thing that pissed me off:



I wonder what Alem Matthees thinks about that. Or any of the 12,000+ people who've signed a petition calling for the retraction of misleading PACE claims. Or me!

#MEtoo

 

[Snow Leopard]

They seem to be confusing the words "corrected" with "contested".

After reading the reply - it sounds overly defensive and uses a lot of weasel words (deliberate imprecision).

Deny Deny Deny (don't bother admitting when things could have been done better!).

 

[Dolphin]

Peter D White, Trudie Chalder, Michael Sharpe, Brian J Angus, Hannah L Baber, Jessica Bavinton, Mary Burgess, Lucy V Clark, Diane L Cox, Julia C DeCesare, Kimberley A Goldsmith, Anthony L Johnson, Paul McCrone, Gabrielle Murphy, Maurice Murphy, Hazel O’Dowd, Laura Potts, Rebacca Walwyn, David Wilks

Psychol Med. 2017 Jan 23:1-12. doi: 10.1017/S0033291716003615. [Epub ahead of print]
Heterogeneity in chronic fatigue syndrome - empirically defined subgroups from the PACE trial.
Williams TE(1), Chalder T(2), Sharpe M(3), White PD(1).

Acknowledgements
We acknowledge the help of the PACE Trial Management Group, which consisted of the authors of this paper, excluding T.E.W., plus (in alphabetical order): B Angus, H Baber, J Bavinton, M Burgess, LV Clark, DL Cox, JC DeCesare, E Feldman, P McCrone, G Murphy, M Murphy, H O’Dowd, T Peto, L Potts, R Walwyn and D Wilks.

So 2 people, the ones crossed out, are not co-authors of the reply to Dr Geraghty.

 

[user9876]

Did the article really need all those contributing authors? Especially when it boils down to "the only correct interpretation of PACE is ours"?

Yes I take it as a statement from all the authors that they are happy to continue to mislead and lie to patients and other medical professionals. Given the current debate there is no excuse for them not understanding the issues.

 

[Cheesus]

Did the article really need all those contributing authors? Especially when it boils down to "the only correct interpretation of PACE is ours"?

It's an appeal to authority, which is also a position they take in the paper in citing the support of the Lancet, NICE and NHS Choices. Appealing to authority is a poor defence of scientific evidence, which should need no support of authority in order to withstand criticism.

 

[Sean]

On the good side, their stubborn pig-headed refusal to face reality is of the self-garroting kind.

Good chance for Dr G to land some serious blows in his reply.

 

[Cheesus]

On the good side, their stubborn pig-headed refusal to face reality is of the self-garroting kind.

Good chance for Dr G to land some serious blows in his reply.

My understanding is there will be editorials from a broader array of people in response to this reply.

 

[Sean]

My understanding is there will be editorials from a broader array of people in response to this reply.

Sounds like things could get interesting.

 

[trishrhymes]

That many authors for such a load of self justifying waffle looks very silly indeed.

Good to see @Keith Geraghty has them seriously rattled.

I do hope there will be other experts providing a counterbalance. I fear lots of readers will take the power of numbers on trust and believe these self-serving idiots.

 

[user9876]

They make this claims

It followed the consolidated standards of reporting Trials (CONSORT) guidance on how to report and conduct a high-quality trial (http://www.consort-statement.org/).

Outcomes

6a Completely defined pre-specified primary and secondary outcome measures, including how and when they were assessed
6b Any changes to trial outcomes after the trial commenced, with reasons
17a For each primary and secondary outcome, results for each group, and the estimated effect size and its precision (such as 95% confidence interval)

I would argue that the didn't give sufficient reason for protocol changes the one where they say that using continuous scoring rather than binomial scoring as it increases accuracy is unsupported. It is not like saying measure in mm rather than cm as they are different marking schemes with different orderings. So to claim that one marking scheme is more accurate than another they need to go back to show one gives a more accurate rendition of fatigue in patients than the other. No such study was referenced.

Further there are secondary outcomes (such as their recovery ones) that have not been published. They published a post-hoc recovery paper but simply dropped the recovery measures between the protocol and their statistical analysis plan. The question may be did they get explicit approval for this from the ethics committees or did they just get the stats plan approved and the ethics committees failed to notice.

A Research Ethics Committee gave ethical approval, and it was overseen throughout by the independent Trial Steering Committee and Data Monitoring and Ethics Committee; patient members sat on the Trial Management and Steering Committees

PACE does raise serious issues about whether the structure of research governance is sufficient. They designed a protocol that was not capable of answering the questions that they were trying to test because they relied on subjective outcomes where some treatments were aimed at changing subjective feelings about the disease. Then they weakened all the criteria. The fact this was approved by committees raises questions about the committees and what information was provided and what the committees actually approved. We have been told we are not allowed to see this information.

We reject the accusation that our ‘actions have arguably caused distress to patients’, for which Dr Geraghty offers no evidence.

Much of the press coverage around PACE with headlines such as get some exercise has caused distress to patients. Also offering treatments that don't work is likely to distress patients especially when the treatments seek to blame patients for not doing enough or telling patients that they just believe they are ill.

The protocol was published some 3 years before the analysis began, and 4 years before the first outcome paper was published (White et al., 2007). The papers reporting the trial findings were peer reviewed before their publication in high-impact journals, such as The Lancet (White et al., 2011).

Interesting that the don't say that the protocol was published after the start of the trial which is what should happen. Or that they deviated significantly from the protocol.

The PACE trial simply confirmed what previous smaller trials had already found (Edmonds et al., 2004; Price et al., 2008): that patients are more likely to get better with either CBT or GET than with other treatments or usual care.

It is clearly false. Their trial showed that by their own definition of recovery that patients were not more likely to get better with CBT or GET. The way they claimed this was to have a recovery definition that allowed someone to get worse with the primary outcome measures.

We reject the suggestion that the fact that we use these therapies for our patients and have tested them in previous trials is ‘a major source of investigator bias’. Clinical research often arises from questions thrown up by clinical practice. The clinicians among us have dedicated their careers to care for thousands of patients with CFS/ME and we always want the best for them. We are therefore obliged to conduct trials to test the effectiveness and cost-effectiveness of treatments that we use. If Dr Geraghty’s proposal, that trials should only be conducted by investigators with no previous experience of an illness and its treatments, was followed, it would prevent any clinician or researcher from attempting to replicate or refute the results of their earlier trials. While steps should always be taken to minimise bias, as we did, this suggestion is not sensible

They failed to take steps to minimise bias as they did not blind assessors and relied on self reported measures rather than objective ones which is likely to add bias. I fail to see a how whether a clinician has 'dedicated their careers to care for thousands of patients with CFS/ME' is reverent to the refudiation of the point about investigator bias. In making this point they are trying to appeal to other doctors emotions rather than making a valid point on the subject.

The reality is that we clearly reported within the paper the numbers of participants who went on to receive the additional therapies (most commonly CBT and GET), which were offered by trial NHS therapists to all participants who needed and wanted further help.

They did report numbers which is how we know their claims that the LTFU data supported their favoured treatments was spin. What they didn't do clearly was report their results clearly in the abstract or press releases. That would have said there was no significance at long term follow up.

 

[user9876]

This is incorrect. Effectiveness was measured by comparing the mean scores for each of the two primary outcomes between treatment groups; the effect sizes varied between 0.5 and 0.8 (moderate effect sizes), depending on the different comparisons (White et al., 2011).

This is very dodgy although perhaps standard practice. To do a comparisons of means this must be a valid measure which requires that the scale is an interval scale. That is an improvement of x from point y is the same as an improvement of x from point z. With the SF36 scale I think this is dodgy basically does a change from being limited a lot to limited a little when walking one block represent the same change in physical function as a change from being limited a lot to limited a little when walking a mile. If they cannot demonstrate that this is the case their analysis is highly questionable. The CFQ is worse since it has different weightings for changes in physical and mental fatigue due to the number of questions about both.

That is of course ignoring the fact that these are not measures of ability but reports. So their claim should be that there is a moderate improvement in reports from patients for techniques that tell patients they will recover if they ignore symptoms and see them as temporary. Put in those terms it is not a good claim especially as it is not supported by the objective measures they did make.

 

[user9876]

The second criticism concerned our secondary analysis paper about recovery (White et al., 2013). Dr Geraghty states that ‘… some trial participants had reached the level required to be classified as improved or recovered at trial entry’. b, which was only one of the criteria necessary to be considered as recovered. To meet the criteria for recovery, a participant also had to have met additional criteria: no longer be considered a case of CFS (using the trial definition of CFS) and rated their overall health as ‘much’ or ‘very much’ better compared to trial entry. No participants met the full criteria for recovery at trial entry.

They are inaccurate here. In there paper they added additional clauses to their caseness definition so that there are a number of people who are rated as still meeting the Oxford criteria by doctors by are not in their recovery figures due to the sf36 or CFQ scores.

Clearly people could not meet the improvement criteria at the start as this is not measureable concept at the start.

What they don't mention is the number that met on or the other primary outcomes at their absurdly low 'normal' levels at the start.

The fact that 1% met all these criteria at the start should have worried them and made them carefully review their criteria. If they had gone through this basic check they would have noticed their stats were dodgy. They have of course never commented on the issue of bad stats in working out the normal range.