Intended for healthcare professionals


Reanalysis of PACE trial reignites row over chronic fatigue treatment

BMJ 2016; 354 doi: (Published 29 September 2016) Cite this as: BMJ 2016;354:i5230
  1. Jacqui Wise
  1. London

A preliminary reanalysis of the controversial PACE trial into treatment of chronic fatigue syndrome has concluded that the previously reported recovery rates were inflated fourfold and that the recovery rates in the cognitive behavioural therapy (CBT) and graded exercise therapy groups were not significantly higher than in the group of patients who had specialist medical care alone.1

The reanalysis, published on the Virology blog, comes after a five year battle to get the original data released. Alem Matthees, an Australian patient with chronic fatigue syndrome, finally succeeded in getting Queen Mary University of London to release the data under the UK Freedom of Information Act.2 3

However, the team of researchers behind the PACE (pacing, graded activity, and cognitive behaviour therapy: a randomised evaluation) trial have said that their own reanalysis confirmed their original conclusion that CBT and graded exercise therapy, but not adaptive pacing therapy, can moderately improve fatigue and physical function in people with chronic fatigue syndrome.4 5 6 Activists and charities that represent these patients generally support adaptive pacing, which involves a gradual increase in activity and rejects psychological treatments.

The PACE trial, published in the Lancet in 2011, compared the effects of adding cognitive behavioural therapy, graded exercise therapy, or adaptive pacing therapy to specialised medical care in 641 patients with chronic fatigue syndrome.7 A second paper on recovery rates, published in Psychological Medicine in 2013, found that CBT and graded exercise therapy were associated with a significantly higher recovery rate of 22% at 52 week follow-up than the 8% for adaptive pacing therapy and 7% for specialist medical care.8

However, critics of the PACE trial say that the researchers had not analysed the data as set out in the original trial protocol.

Matthees and two academic coauthors, Philip Stark, from the department of statistics at the University of California, and Bruce Levin, professor of biostatistics at Columbia University, New York, have now reanalysed these recovery rates using the thresholds specified in the published trial protocol. They found that the recovery rates were much reduced, with 3.1% for specialist medical care alone, 6.8% for CBT, 4.4% for graded exercise therapy, and 1.9% for adaptive pacing therapy.

Matthees and coauthors said, “The changes made by the PACE trial investigators after the trial was well under way resulted in the recovery criteria becoming too lax to allow conclusions about the efficacy of cognitive behavioural therapy and graded exercise therapy as rehabilitative treatments for chronic fatigue syndrome.” They called for a “thorough transparent and independent reanalysis [to] be conducted to provide greater clarity about the PACE trial results.”

Their reanalysis was limited to the secondary outcome of recovery, not improvement. The authors said that they plan to re-analyse the primary outcome measures of improvement and submit a more complete analysis to a peer reviewed journal in due course.

Peter White, leader of the PACE trial and professor of psychological medicine at Queen Mary University London, said that the reanalysis on the Virology blog was done very well but used more stringent thresholds for recovery than his team had. He told The BMJ, “It is no surprise that this resulted in fewer people apparently recovering. How do they know that they used the right recovery thresholds? Our results were similar to those of two previous studies, which gives us some confidence in the findings.”

White said that in common with many trials the statistical plan was developed after the trial had started but before any data were examined and was ratified by the trial steering committee. One change was to use continuous scoring of each fatigue scale item rather than that bimodal scoring originally proposed. The PACE team has now reanalysed their trial with the primary outcomes in the form outlined in the published protocol. They found that the odds of being improved on the composite outcome of physical functioning and fatigue were between 2.4 and 2.7 times greater with CBT and graded exercise therapy than with adaptive pacing therapy and specialised medical care.4 They plan to do a third reanalysis of the recovery paper in due course.

White said that his team had released this new analysis on the website rather than submitting it for publication to a journal to inform the debate and not cause delay. He said that the more conservative thresholds had made a difference to the results. “However, it made no difference to our conclusion that cognitive behaviour therapy and graded exercise therapy are safe and moderately effective in improving outcomes for chronic fatigue syndrome.”

Responding to criticisms over their sharing of data, White said that they had shared PACE trial data over a dozen times with bona fide researchers, including a Cochrane review group.


  • Editorial The long wait for a breakthrough in chronic fatigue syndrome doi:10.1136/bmj.h2087


View Abstract

Log in

Log in through your institution


* For online subscription