Saturday 16 March 2019

PACE trial for chronic fatigue syndrome (still) being put through its paces: a reply

I'm bringing the paper published by Michael Sharpe and colleagues [1] to your attention today in the interest of balance and peer-reviewed 'right to reply'.

The Sharpe paper concerns the PACE trial, the study which reported that "when added to specialist medical care, cognitive behaviour therapy and graded exercise therapy were more effective in improving both fatigue and physical function in participants with CFS [chronic fatigue syndrome], than both adaptive pacing therapy and specialised medical care alone."

Anyone with a little bit of knowledge about the PACE trial will know that it's a 'contentious' topic within CFS (and ME, myalgic encephalomyelitis) circles. Indeed, the Sharpe paper comes about as a direct result of a reanalysis paper (see here) which reported findings raising "serious concerns about the robustness of the claims made about the efficacy of CBT [cognitive behavioural therapy] and GET [graded exercise therapy]" in the context of CFS/ME. 'Serious concerns' is putting it mildly considering how others have described the PACE trial and some of its tenets (see here).

Sharpe et al, who were authors listed on the original PACE trial paper [2], have had to defend their work/findings before in the peer-reviewed realm (see here). Same as before, the name Carolyn Wilshire is addressed and her teams reanalysis of the PACE trial data [3]. Said data was, I might add, (partially) released only following intervention from the Information Commissioners Office (ICO) here in Blighty (see here). More recent events have similarly reiterated that 'access to raw study data' is something that CFS/ME researchers perhaps need to bear in mind at study conception (see here).

I'm not going to clinically dissect the Sharpe paper in this post because (a) 'interpretation' forms quite a bit of the reply to the Wilshire reanalysis, and (b) your opinion on the scientific quality of the Sharpe reply is most likely going to be shaped by where you stand in terms of the whole CBT/GET for CFS/ME discussion. Indeed, a peer-reviewer of the Sharpe paper also said as much (see here). What I will comment on is how the Wilshire reanalysis paper and the more recent Sharpe reply to the reanalysis paper might further inform research more generally with CFS/ME in mind.

Oh, and it's probably just a coincidence that the Sharpe paper comes out only days after a news headline reads "Online activists are silencing us, scientists say" talking about a familiar topic.


  • Point 1: Design a good trial analysis plan and stick to it. From my 'outsider looking in' perspective, the changes made to "the scoring of the pre-specified outcomes" regarding fatigue and physical functioning in the PACE trial, however innocent they might have been, have created tension. Lots of tension. Such changes, whether agreed by "Trial Data Monitoring and Steering Committees" or not, can be construed in various different ways. It's better not to make such changes in the first place.
  • Point 2: If you are going to study something like physical functioning in relation to CFS/ME, don't just rely on things like questionnaires and Likert scales; use actigraphy too. Self-report is always a good thing but I've never understood why, with the wide range of cost-effective technology out there (available I believe, even in the early 2000s), wearable trackers such as pedometers or similar were not also utilised during such studies (see here). If you're spending £5 million on a trial, a few quid for some pedometers is not exactly going to break the bank and will inevitably bring some further quality data to the table.
  • Point 3: Recovery. As per other discussions (see here), most people would characterise recovery as a complete remission of symptoms and/or return to typical functioning. If you're not going to use this description, don't use the word recovery. Use something else instead. Indeed, use 'partial remission' or 'improvement' if you need to but don't call anything less than the complete remission of symptoms 'recovery'.
  • Point 4: Long-term outcomes. It's probably best to avoid any sweeping statements after the arms of a trial - a "randomised trial" not necessarily a "randomised controlled trial" according to Wilshire et al - have been completed. More so when you're measuring such long-term outcomes via a postal question minus any objective measure(s) (see point 2). It's probably also a good idea to ask patients about their quality of life too and whether that has changed (see here).
  • Point 5: Even if your paper states in no uncertain terms that: "The effectiveness of behavioural treatments does not imply that the condition is psychological in nature" the use of something like CBT for CFS/ME implies that you probably think there is a substantial psychological 'component' to the condition. This is compounded when you're for example, a Professor of Psychological Medicine. If you were pitting CBT in particular against a specific pharmacological or biological intervention 'for CFS/ME' (see here for example), I'd be more inclined to see your view in a more 'rounded sense'. Indeed, if you were to study one or two biological parameters as well as behavioural ones looking for any change following intervention, you might convince more people that psychology is not the primary line you take. And whilst on the topic of psychology and CFS/ME, it's probably also best not to use 'psychobabble' terms like 'deconditioning' in your research. Such terms are pretty much scientifically untestable and, given the recent discussions about the legacy of some adherents to something like psychosomatic research (see here), is likely to be consigned to the scientific dustbin as some later point.

I think I've covered the main points as I see them. Please feel free to agree/disagree as you wish.

End of Line.

Addition: 26 March 2019. Not quite 'End of Line' it seems, as a reply to a reply to a reply emerges [4]. Peer reviewed science is far from slow...


[1] Sharpe M. et al. The PACE trial of treatments for chronic fatigue syndrome: a response to WILSHIRE et al. BMC Psychology. 2019; 7: 15.

[2] White PD. et al. Comparison of adaptive pacing therapy, cognitive behaviour therapy, graded exercise therapy, and specialist medical care for chronic fatigue syndrome (PACE): a randomised trial. Lancet. 2011; 377(9768):823-36.

[3] Wilshire CE. et al. Rethinking the treatment of chronic fatigue syndrome—a reanalysis and evaluation of findings from a recent major trial of graded exercise and CBT. BMC Psychology. 2018; 6: 6.

[4] Wilshire CE. & Kindlon T. Response: Sharpe, Goldsmith and Chalder fail to restore confidence in the PACE trial findings. BMC Psychology. 2019; 7: 19.



  1. Just on the physiological tools if you read the PACE manuals heart rate monitors were used, VO2 max was measured at baseline, actigraphy was used at baseline. STEP and 6 minute walking tests carried out. The OBJECTIVE measures showed little improvement. The PACE authors said it was too onerous for their recovered patients to wear a 28g actimeter at the end of the trial ie once recovered and the data was not kept. The therapy data ie daily HR etc was not compiled. FOI resquests on What do they Know - asked for all sorts of data to no avail. The problem is NOT a lack of data but a lack of preparedness to show it. The objective results remain overlooked as they do not support GET or CBT as effective treatments.

    1. Many thanks for that. Another commentator also mentioned this:


Note: only a member of this blog may post a comment.