PACE trial
The PACE trial was a large and controversial trial which compared the effects of cognitive behavioural therapy, graded exercise therapy, adaptive pacing therapy, and specialist medical care for people with myalgic encephalomyelitis/chronic fatigue syndrome.
The investigators reported that both CBT and GET were "moderately" effective at treating ME/CFS. The results have been questioned due to methodological weaknesses in the study design. For instance, the definition of recovery was weakened after the data was collected, leading to a larger share of recovered participants. Analyses under the original protocol did not find significant effects and concluded that GET is potentially unsafe for patients with ME/CFS. As of 2021, graded exercise therapy and cognitive behavioural therapy are no longer recommended by the National Institute for Health and Care Excellence for people with ME/CFS.
Initial study and findings
The £5 million trial was funded by various branches of the UK government, including the Department for Work and Pensions. Notable authors include psychiatrists Michael Sharpe and Peter White, along with psychotherapist Trudie Chalder.Study design
641 patients were selected and divided into four groups. The first group only received specialist medical care. The second group received SMC and cognitive behavioural therapy. The third group received SMC and graded exercise therapy. The fourth group received SMC and adaptive pacing therapy, not to be confused with pacing.SMC consisted of specialists giving advice to patients, such as "to avoid extremes of activity and rest", and the prescription of drugs to help manage symptoms. CBT was done on the basis of a theory that regards ME/CFS as being reversible and that "fear of engaging in activity" and "avoidance of activity" are linked and interact with physiological processes to perpetuate fatigue. APT consisted of telling patients not to do activity, if it increases their symptoms; however, patients were encouraged to increase activity if their symptoms were manageable. GET was done on the basis of theories that assumed that ME/CFS is perpetuated by reversible physiological changes of deconditioning and avoidance of activity.
Outcome measures were assessed at the beginning of the trial, and then at 12, 24, and 52 weeks. Subjective outcome measures included self-rated fatigue and physical functioning questionnaires. Objective outcome measures were a six-minute walking test, a fitness test, reception of benefits due to illness, and number of days out of work due to illness. Two and a half years later, there was another follow up using only subjective measures.
The study design had various strengths. The large sample size and randomised allocation of patients into groups of near equal size, increased the precision of findings and, therefore, the likelihood of finding statistically significant outcomes. Additionally the subjects received a substantial amount of therapy, over an extended duration, which contributed increased precision in the findings.
Initial findings
The findings were published in 2011 and concluded GET and CBT were "moderately effective" treatments. 52 weeks after the beginning of the trial, self-reported fatigue scores were significantly lower and self-rated physical function scores significantly higher for the GET and CBT groups than for the SMC and APT groups. The average scores on a 6-minute walking distance test were higher for the GET group than for the other groups. However, the mean distance walked was still well below the mean for healthy elderly people and over 30% of patients in the GET group did not do the test at 52 weeks. The CBT group did not perform significantly differently from the SMC and APT groups on this measure. The results from a step fitness at 52 weeks test showed no significant differences in performance across groups.A 2013 paper examined the proportion of patients who could be classified as "recovered" after the trial. A patient was considered recovered if they obtained a specified threshold score on the fatigue and physical function self-report scales, if they rated their health as "much better" or "very much better", and if they also failed to meet the authors' case definition of CFS. According to the primary measures of recovery reported in the paper, 22% recovered after CBT, 22% after GET, but only 8% after APT and 7% after SMC.
Although patients were selected based on the Oxford Criteria, subgroup analysis found equivalent results for patients meeting the Oxford criteria and London ME criteria or the 1994 CDC criteria.
A follow-up conducted 2.5 years after the commencement of the trial reported no significant differences between the various treatment groups on the primary self-report measures. That is, the treatment-specific effects evident at 52 weeks were no longer evident at 2.5 years.
Controversies and criticism
After being published, the trial generated considerable criticism and concerns which were voiced through letters to the editor, patient groups, and groups of researchers and healthcare professionals. Lancet Editor Richard Horton defended the trial, calling the critics "a fairly small, but highly organized, very vocal and very damaging group of individuals who have, I would say, actually hijacked this agenda and distorted the debate so that it actually harms the overwhelming majority of patients."The study continued receiving criticism from within the scientific community. Ronald Davis of Stanford University wrote, "I'm shocked that the Lancet published it The PACE study has so many flaws and there are so many questions you'd want to ask about it that I don't understand how it got through any kind of peer review". In an analysis of the study's design, the mathematician Professor Rebecca Goldin wrote that "There were problems with the study on almost all levels the flaws in this design were enough to doom its results from the start."
Outcome switching
After the initial data was collected, the method for scoring the subjective primary outcome measures of patient reported physical functioning and patient reported fatigue was changed, i.e. it used a more sensitive measure of the same primary outcome. Outcome switching is a practice that can lead to bias and undermine reliability, as studies may cherry pick a statistically significant outcome. According to the authors, the switch was done in order to "improve the variance of the measure" and was "approved by the independent PACE Trial Steering Committee and Data Monitoring and Ethics Committee". An analysis under the original protocol found no statistically significant difference in improvement rates. The improvement rates in the GET and CBT groups were around 20%, rather than the 60% reported with switched outcomes.There was also a change in how recovery was defined. One of the original requirements was that patients would need to score over 85 on a survey of physical functioning, indicative of what healthy working-age people score. This was changed to a score of over 60. This meant that some were above 60 on this metric at the start of the trial, as the criterion for entry was a score of under 65. However, no patients were counted as recovered at trial entry, as they also needed to not meet the Oxford, CDC or London criteria for ME/CFS any more, and rate their overall health as much or very much better in order to be classed as recovered. Recovery rates under the original protocol were much lower and no statistical difference was found between the groups.
Subjective primary outcome measures
There has been considerable criticism due to the original findings only being supported by subjective primary outcome measures. Professor Jonathan Edwards of University College London wrote that " makes non-starter in the eyes of any physician or clinical pharmacologist familiar with problems of systematic bias in trial execution."The authors responded that a number of objective measures were taken, and the GET group had significant improvement in the 6 minute walking time. However, this was the only statistically significant improvement in objective outcome measures. In a reanalysis of the data Wilshire et al. noted that over 30% of the GET group did not complete the walking test, and that the GET group only walked 10% farther than controls after a year of exercise therapy. The other objective outcome measures included a fitness test, employment, and reception of disability benefits. None of these measures yielded a statistically significant improvement. The authors replied that participants may have lost their job after becoming ill, and that getting better and getting a job are not the same. The authors said they chose patient-reported primary outcomes as they thought patients themselves were the best people to judge the state of their own health.
Possible placebo from patient newsletter
The patient newsletter given throughout the study subtly praised CBT, GET, and "activity management" by saying a government committee had approved them as treatments, while critiquing biomedical research in ME/CFS. This raised concerns of a possible placebo effect, especially given the subjective outcome measures.The authors responded that the newsletter's mention of the NICE guidelines included all 3 treatments being compared in the trial. However, the newsletter did not include "adaptive pacing therapy", it did include "activity management" of which adaptive pacing therapy might be classified as a subtype. David Tuller of University of California, Berkeley replied that trial participants would not necessarily be aware that the authors considered "adaptive pacing therapy" as a subtype of "activity management". He also said that, even if all three treatments were included, the positive bias resulting from it will impact the reliability of the trial outcomes.