The PACE trial results – an analysis by Vivienne Parry, OBE

The Association of Young People with ME (AYME) invited respected journalist and broadcaster Vivienne Parry to write an independent article on the PACE Trial in light of the controversy that surrounds it and the use of graded exercise therapy and cognitive behavioural therapy with this condition.

Vivienne Parry is best known for her broadcasting for Radio 4, where she presents many medical science programmes. In the New Year's Honours 2011, she was awarded the OBE for services for the public understanding of science. Below is Vivienne Parry's analysis of the PACE trial written exclusively for LINK members.

ME/CFS is a disease which has far too little time and money spent on it. And whilst research on its cause, or more likely its many causes, is still in its infancy, the need for better treatments for those who are affected today by this dreadful condition is acute.

Thanks to the PACE trial, the largest ever study of ME/CFS treatments whose results were published in March, we at least now know for sure that there are two treatments – graded exercise therapy (GET) and cognitive behavioural therapy (CBT) – that are safe and moderately effective.

But the trial outcome didn't go down well in the ME/CFS community with the result condemned as ‘disappointing and surprising' by some organisations, although it was welcomed by AYME. I want to take a look at those comments but let me first refresh your memories on what the trial involved and why it was carried out.

Back in 2003, patient groups were concerned that cognitive behavioural therapy (CBT) and graded exercise therapy (GET) might do more harm than good. They favoured pacing (a way of adapting available energy to daily life with the help of a diary) along with specialist medical care but this had not been subject to a large scale trial. The only way to find out for sure which was the best treatment and whether there were any safety problems was to do a large randomized trial of the various treatments. A randomised trial means that neither eligible patients nor their doctors have a choice about their treatment; a computer randomly allocates them to one of those being trialled. This trial could not be blinded (no-one knows who's had what treatment) because the treatments are so obviously different but it was blinded as far as those who were analysing the results were concerned. This was as rigorous a study as it is possible to have.

From the outset, Action for M.E. and a number of patients were involved in the design of the trial. PACE involved 640 patients from England and Scotland. It was carried out by a team of experts and led by researchers from Queen Mary University London, King's College London and the University of Edinburgh. The research proposals were subject to extensive review and scrutiny by independent experts and trial committees. The £5 million cost was funded by the Medical Research Council, the Department of Health, the Scottish government and also by the Department of Work & Pensions.

The trial assessed the safety and effectiveness of four separate treatments: specialist medical care alone, specialist medical care combined with cognitive behavioural therapy (CBT), or with graded exercise therapy (GET), or with adaptive pacing therapy (APT), usually just called pacing. The specialist medical care involves general advice about managing the illness plus normally prescribed medicines for symptoms such as insomnia and pain.

The 640 patients, all of whom had fatigue as their main symptom, were assigned to the four groups in roughly equal numbers of 160 individuals each. All the patients had an assessment of their fatigue and physical function at the start of the trial and a year later at the end of the trial.

The results were that average fatigue and physical function scores had improved more after CBT and GET than after either pacing or standard medical care alone. The PACE trial found that pacing has no benefit but that up to 60% of patients with ME/CFS benefited from CBT and GET. This counts as being moderately effective treatment.

This was a trial for adults so the results may not be applicable to children and it did not include those patients who were housebound. Further research is urgently needed to address the needs of these two groups.

Serious adverse reactions to treatment were recorded in two pacing patients (1%), three CBT patients (2%), two GET patients (1%) and two in the SMC-only group (1%). Let me say something about what adverse reactions mean here. Everyday things – like getting cold or flu, being in a car accident or even falling over the cat and gashing your knee – that happen to occur within a study time frame count as non-serious adverse events but not reactions to trial treatments. Serious adverse reactions were, for example, someone taking to their bed as a response to a trial treatment. Of the 10 serious adverse reactions recorded, nine were only "possibly related" to a trial treatment and only one was "probably" a reaction to a trial treatment and that was an adverse drug reaction in someone in the standard medical care alone group. All in all, this is very convincing evidence that all these treatments are safe.

So why have so many patient groups including the charity involved in the design and implementation of the trial, condemn the findings? They said that the results were at odds with numerous patient surveys, for instance one for AfME in which some 50% of patients reported harm from graded exercise therapy.

There are two problems here. One is about science. Research is about coming up with a hypothesis and then trying to knock it down. Sometimes what you believe to be the case (the hypothesis) gets overturned. Long held, cherished and utterly plausible ideas are regularly demolished by evidence. This can be incredibly disappointing but you have to move on and ask the next question, not constantly keep asking the same one in the hope that eventually you will get a different answer.

The other problem has to do with surveys which usually involve a self selected sample of the population and can produce highly biased results. For instance, responders might only be those people for whom a particular treatment hasn't worked whereas those for whom it has worked, haven't bothered to be involved. When respondents to one of these surveys were followed up, especially those who reported they had become much worse as a result of graded exercise therapy, it turned out that some had been sent to gyms or other inappropriate places or hadn't had the sort of supervised, careful therapy from experts that was a feature of the trial and which should be offered to ME/CFS patients.

Some people also said that the trial was meaningless because it excluded those with a neurological disease, therefore could not have contained anyone who had ME since this is classified as a neurological disease. This is a bit silly because why would you design a trial that excluded the very patients you wanted to study? The problem arises from a line in the so called ‘Oxford criteria' which were used to assess eligibility for the trial and says ‘those with proven organic brain disease' should be excluded. This means those with conditions like dementia, multiple sclerosis and Parkinsons, not ME/CFS. The choice of the Oxford criteria, rather than the CDC (an American measure) or Canadian criteria was also condemned since some people say that using these allows those with fatigue from causes other than ME/CFS to be included and therefore those that got better were not people who had ‘real' ME. However, the Oxford criteria excludes anyone who has an alternative cause for their ME/CFS symptoms. The researchers also assessed all patients to see if they met alternative definitions of ME/CFS. One definition, according to the London criteria for ME, based on Melvin Ramsay's original description of ME from the Royal Free hospital and the other according to CDC guidelines. The results were the same in these groups. As well as using the Oxford criteria 67% of patients also met the International CFS Criteria and 51% the London ME Criteria.

What does this mean for treatment? CBT and GET are moderately effective for six out of every ten adult patients. They are likely to be effective in children but no one knows this for sure until further research is done. They are safe but only in the hands of therapists who are properly qualified. Finally, the NICE guidelines are not due for revision until 2013 and NICE have recently confirmed that they will not be altering them until then.

The nub of patient reaction to this seems to rest on an implication - which is why so many found some of the newspaper headlines so offensive - that the way they get better must say something about the cause of their illness. So, for instance, if CBT works for you, it must mean that ME/CFS has a psychological cause or if graded exercise therapy works, it must mean you are lazy. I find this completely baffling. I've worked a lot with cancer groups who campaign to have CBT as part of their treatment because it helps them feel better. No-one would dream of suggesting, I hope, that their cancer was a psychiatric disease.

I sit on the Council of the MRC (although I joined it after the PACE trial was approved) and I'm delighted that a further £1.5 million is going to be made available for ME/CFS research by the MRC. Contrary to what is claimed on message boards, there are strong voices, including my own, speaking up for the need for ME/CFS research within the MRC, which must include work on both causes and better treatments. I look forward to telling you the results of this research in the future. 

The therapies are defined as:

  • Cognitive behavioural therapy (CBT) - A clinical psychologist or specially trained nurse, helps the patient to understand how their symptoms are affected by the way that they think about and cope with them and encourages them to try out increasing their activity. 
  • Graded exercise therapy (GET) – A physiotherapist helps the patient to try a gradually increasing tailored exercise programme which takes into account the individual patient's symptoms, fitness, and current level of activity. 
  • Adaptive pacing therapy (APT) – An occupational therapist helps the patient to match their activity level to the amount of energy they have, aiming to help the patient adapt to the illness rather than assuming they can gradually do more.