The biggest and baddest CBT/GET study has finally arrived. The UK took their best shot at behavioral therapies for ME/CFS in their mult-million dollar, multi-center, multi-year, 600 plus patient PACE Trial. This, the biggest, most expensive and rigorous treatment trial for CFS ever, had taken years to plan; it’s program alone was hundreds of pages long.
The stakes were high. The big Belgian government funded CBT study to end all CBT studies failed miserably last year leading CBT advocates to sputter that the trial had not failed but that expectations were simply too high – as if a return to health was an unreasonable expectation.
But this was the UK, where CBT had found it’s greatest home and it’s most prolific and influential practitioners. This was a country where the behavioral paradigm of CFS even had a strangehold on the media, a place where studies showing evidence to the contrary sometimes simply weren’t covered. If anyone could pull it off the UK could.
Peter White, the main author of the study, had been way out in front of the study, declaring that it was a success prior to publication and sending twinges of fear down the spines of a wary CFS community…..and now it was here. Was he right? First let’s look at the outlines of the study.
The Cohort – First they gathered people who met the Oxford Criteria and had a physical function score of >65. People who had psychosis, bipolar disorder, substance abuse, an organic brain disorder or eating disorder were excluded (58 people in total). No exclusions for other disorders were noted. Many people declined to assessed or to engage in a randomized treatment (n=398) (It’s possible that these people included people afraid of being assessed the GET or CBT protocols). A good number of people simply refused to be assessed (n=143). Another chunk of people (n=139) couldn’t follow the protocols for one reason or another (too sick?).
Demographics – The average age was 38; 77% were female, 56% met the London criteria for ME, a whopping 47% were diagnosed with a psychiatric disorder during their clinical interview (depression, anxiety, obsessive compulsive disorder, PTSD), a high percentage of people in the study were on antidepressants (@33%), half had a history of depression, their average duration of illness was 32 months (2 ½ years) and their average body mass index (BMI) was 25.5…(normal BMI = 20-25).
Patient Expectations – Rather surprisingly given the bad publicity GET gets in the ME/CFS Community, a high percentage of the group were confident about how successful the GET protocol would be (70%) as well as pacing (71%). They had lower confidence in CBT (57%) and low confidence in the efficacy of the specialist medical care (41%).
In the end the 641 people were randomly divided into 4 treatment arms; ‘Adapted Pacing’, Cognitive behavioral therapy, Graduated Exercise Therapy (GET) and Specialized Medical Treatment (care from CFS ‘specialists’ only/no behavioral therapies).
They were given 14 therapy sessions over the first 23 weeks and then one at 36 weeks. Their health was assessed at baseline, 23 weeks and 52 weeks using a wide variety of self report measures. An actigraphy test that measured movement over time was included in the original protocol but that test was dropped from the final protocol (Thanks to Dolphin for the info), suggesting that the study authors had little faith that their therapies would produce benefit in that area.
- Adaptive Pacing Therapy – is a form of pacing that was specifically created for this study that focused on adapting to available energy levels and systematically increasing them over time.
- Cognitive Behavioral Therapy – believes that fear responses interact with physiological processes to perpetuate fatigue. Besides addressing cognitive issues (disempowering thoughts and perceptions) they also address sleep and wake issues. CBT therapists believe CFS is reversible
- Graduated Exercise Therapy (GET) – asserts that activity avoidance results in deconditioning and can be reversed using gradual increases in activity which result in less fatigue and fewer symptoms.
- Specialist care – ‘CFS specialists’ in the UK provide advice on rest, avoiding push/crash and symptomatic pharmacological treatment for insomnia, pain and mood
- Measurements – The practitioners measured treatment effectiveness using a broad array of self-report tests including a clinical global health scale, work and social adjustment scale, 6 min walking ability, Jenkins score for disturbed sleep, anxiety and depression scales, symptoms and others.
The goal of the trial was to “was to change the behavioural and cognitive factors assumed to be responsible for perpetuation of the participant’s symptoms and disability.”
Were they successful in altering the factors they believe perpetuate the disorders symptoms? It’s possible they were – that they actually succeeded in altering the factors ie the thoughts and perceptions they believed are so damaging in CFS. Unfortunately doing so had remarkable little effect on their recipients health.
The authors, themselves, noted “these treatments were only moderately effective,” and that research into more effective treatments is needed. We have to understand the context they were speaking in. ‘Effective’ in medical terms means that a treatment ‘moves’ a symptom not that it cures or removes it. Moderately effective means that it moderately ‘moved’ that symptom; ie it made it somewhat better.
The authors did not say CBT/GET/PACING were cures for CFS or even ‘effective’ treatments for it; they said they were ‘only moderately effective’, a decidedly weak assessment of its effects and that assessment is the crux around which we should understand this study.
The therapies did help some -there’s no doubt about that. The great majority of the people using the behavioral therapies (Pacing (88%), CBT (82%), GET(88%) in the study were ‘satisfied with the treatment’. In contrast to that only 50% of the people receiving only ‘specialist care’ were satisfied – suggesting that these therapies did assist these people to some degree.
But when asked a more piercing question – did the treatments have significantly positive effects – at their best only about 40% of the participants said they did (for CBT) – meaning that almost 2/3rds of them received minimal or negative benefits from them.
Indeed, while most people improved the tests revealed that few improved really significantly and none suggested a return to health had been achieved. The Chalder Fatigue Scale test is a case in point.
Chalder Fatigue Scale – contains 11 questions
• Do you have problems with tiredness?
• Do you need to rest more?
• Do you feel sleepy or drowsy?
• Do you lack energy?
• Do you have trouble starting things?
• Do your muscles have less strength?
• Do you feel weak?
• Do you have difficulty concentrating”
• Do you find it more difficult to find the correct word?
• Do you make slips of the tongue when speaking?
• How is your memory?
and asks if the factor is less than usual in strength (0 pts)/ no more than usual (1 pt)/ more than usual (2 pts) / much more than usual (3 pts). A top fatigue score would be 33 pts and a completely unfatigued score would be 0.
Chalder Fatigue Scale Results
AP CBT GET Specialist Care
- Baseline 28.5 28.2 28.2 28.3
- 24 Weeks 23.2 21.5 21.7 24.0
- 52 Weeks 23.1(-5) 20.3(-8) 20.6 (-8) 23.8 (-5)
The best these therapies were able to do was to move the participants from an average answer that was somewhere between ‘more than usual’ to ‘much more than usual’ to about ‘more than usual’. Thus, the therapies did help but they still left the participants with ‘more than usual’ fatigue/cognitive issues.
Clinically Useful (?) – The criteria for what constituted a ‘clinically useful result’ was quite low. For instance a person improving by simply two points on the Chalder Scale (0-33) and 8 points on the Physical Function Scale (0-100) was deemed to have a clinically useful result. Thus a person who moved from ‘much more than usual’ to ‘more than usual’ on just two of the answers on the Chalder Scale would have a ‘clinically useful’ result. To put it another way – all it took was a movement of less than 10% (scale-wise) on two questions for the therapies to be determined ‘clinically useful’
Yet even with these lax critieria only 42% of Adaptive Pacing, 59% of CBT, 61% of GET and 45% of Specialist were able to meet them. To put it another way even with these low criteria fully 40% of the CBT participants ( the most effective therapy) did not achieve what the authors determined was a ‘clinically useful ‘result.
‘Global Impression of Health’ – is a measure that appears to have simply asked the participants how their overall health had changed – did they see a positive change? Was the change either way minimal or had they seen a negative change? Once again the protocols largely flopped. When asked if they’d seen a positive change at best only 41% stated they had and 60% reported they had not.
Work and Social Adjustment – is one of the more important scales because it’s more of a functional assessment. This self report scale (0-8) asks how impaired one’s ability to work, to do home chores, engage in social and leisure activities (reading, etc.) and relationships is. A complete score will range from 0 (not impaired at all) to 40 (completely impaired).
AP CBT GET Specialist
- Baseline 27.9 27.4 27.3 26.9
- 52-week score 24.5 (+3) 21.0(+6) 20.5 (7) 23.9 (+3)
Once again CBT and GET were better than ‘Adaptive Pacing’ but once again we have to ask how how good were they? In the most effective treatments the participants went from an average of a five to a four – or from reporting they were in the first tier of being ‘very severely impaired’ to being almost very severely impaired.
Because the scale doesn’t differentiate where the improvements came; it’s possible that most of the improvements came in the social and leisure activities; ie they reported they were doing better with reading, writing, etc. and relationships and had little or no improvement in their ability to work.
In any case the treatments left them on the edge of or in the state of being “very severely impaired’ – hardly a result to cheer about.
Baseline Walking Test – was another measure of functionality. It simply asked the participants to ‘walk as far as they can’ in six minutes and then measured in meters how far they went. The best outcome was in GET but once again one has to ask how much meaningful improvement the participants saw.
AP CBT GET Specialist
- Baseline 314 333 312 326
- 52 weeks 334 354 379 348
According to Walking.About.com http://walking.about.com/od/measure/f/howfastwalking.htm these therapies moved the participants from walking at a ‘puttering around the house’ rate to just above a puttering around the house rate. None of them assisted the participants to come close to even maintaining an ‘easy health walk’ for all of six minutes.
- Puttering around the house – 320
- Easy Health Walk – 500 meters
- Brisk Walk – 650 meters
- Fast Walking – 800-1,000 meters.
The other tests were much of the same; some improvement in qualilty of life was seen but there was no evidence that core features of the illness had changed.
No Difference between Definitions – One surprising outcome was the lack of difference the type of diagnosis played in how well the patients responded to therapy. Advocates for years have argued that the Oxford definition selects for a different kind of patient than the Fukuda definition and that an ME definition focusing on post-exertional malaise and acute onset selects for an entirely different patient and they may be right but all these groups responded almost identically to the therapies.
The fact that the treatments were not particularly effective for anyone suggested they were generic treatments that might have produced the same results for anyone with a chronic illness and thus may have said little about the differences between the definitions.
Few Negative Effects from GET – another big surprise was the few negative effects from GET given how many people experienced negative effects in other surveys. Looking at the results, though, one wonders how much ‘GO’ there was in the GET program.
Results from GET should have shown up most in the 6-minute endurance test and the GET recipients did walk more than the other participants -but they still walked at an astonishingly low pace. When asked to walk as far as they could the GET recipients only walked at a bit above a ‘putter around the house pace’ and nowhere near a ‘easy health walk’ pace. This suggests that the most important aspect of any study – increased physical functionality was not achieved.
As noted earlier an actigraphy test originally included in the PACE protocol was not included in the project presumably because the study authors lacked faith in their therapies to have an effect in this area.
Patient Cohort – Another consideration involves the types of patients involved in the survey. More people were eligible for the study than participated in it. Almost 400 people (n=398) did not want to participate in the randomized treatments (probably did not want to try GET) and 143 refused to be assessed. Another big chunk of people (n=139) couldn’t follow the protocols for one reason or another (too sick?). All told more people were eligible for the program and chose not to participate in it (681) than participated in it (641). Plus a significant subset of the community was too ill to participate in the study.
Most people also looked forward to the GET program – a surprising finding for many in the CFS community who have tried to exercise again and again and again fruitlessly. It’s possible, therefore, that the strongly exercise intolerant subset of CFS patients avoided the study for fear of being placed in the GET program and that the study ended up top heavy with more exercise tolerant patients who were more likely to be able to withstand the rigors of walking.
Secondary Comparisons – the authors again and again noted how CBT and GET resulted in significantly improved results for CFS patients compared with Adapative Pacing (4/8 tests) and Specialist Medical Care (6/8). CBT and GET, however, only looked good compared with the meager results of those two protocols.
None of these treatments returned their health to the participants or close to it. They were still highly bothered by fatigue, had high levels of symptoms and lacked endurance. If the objective was to get patients well -and the stated objective was to remove the factors that perpetuated their symptoms (ie that caused their condition) – the trial failed spectacularly.
This is the second government sponsored large behavioral to under perform in the last year and should cause officials to reconsider how they are using precious research dollars. http://www.ncbi.nlm.nih.gov/pubmed/20418251
The main conclusion to be drawn from the PACE trials is that CBT, GET and “Adaptive Pacing’ are, like other many therapies, simply symptom based therapies that can improve symptoms in some people (and not in others). The authors noted these therapies should never be considered primary treatments for CFS but as adjunct treatments a CFS specialist might use if appropriate.
CBT is mainly used in mood disorders but it’s been used to reduce stress in many disorders from heart problems to rheumatoid arthritis to asthma can be triggered by stress. It has been shown to be helpful in all of them.
A 2010 study, for example, found CBT was effective in reducing the frequency of fatal and non-fatal heart attacks. Over 4 years the rate of cardiac events was reduced by 41% in this population. CBT was found to be ‘effective’ in reducing the distress of chronic tinnitus and depression in diabetes. CBT has been shown to be helpful in decreasing pain and increasing range of movement in osteoarthritis. It’s time for chronic fatigue symptom to be taken out of the ‘mood disorder’ context and placed in the context of disorders in which stress reduction therapies such as CBT can be helpful, at least for some patients.
Central Theses of CBT and GET Disproved – The thesis that chronic fatigue syndrome results from deconditioning or fear avoidance or negative thoughts was not borne out by this study. The study suggests that that these factors may play a role for but if they do it’s only a subordinate one and only affects some patients.
Any kind of progress is helpful in such a difficult to treat disease. This was evident in the disconnect between the patients expression of ‘satisfaction’ (most were satisfied) and their reports of significant progress (most did not feel they achieved significant progress). ‘Some Help’ is not how CBT is is often characterized, however, and the fact that it has often characterized as a cure by CBT therapists has deformed the field and the public understanding of the disorder and caused resources to be misapplied.
A Comparison – Compare these results to a study Dr. Holtorf referred to at the Fibro-Fatigue Centers – a group of Centers in the US that is hardly renowned for its effectiveness at treating ME/CFS. Even so the the FFC study found that by the fourth visit a largely hormone based protocol (in 4,000 patients) resulted 85% of patients reporting they were ‘improved’, 56% reported ‘significant improvement’ and 40% ‘substantial improvement’. By the 7th visit 62% reported significant improvement and 46% substantial improvement. These numbers are far higher than the PACE trial reports yet these therapies have received little recognition by a medical establishment that has stubbornly clung to their behavioral paradigms.
The Next Trial – The PACE trial set a good example of what can be learned by comparing treatment effectiveness. We learned that behavioural therapies plus CFS specialists were better than CFS specialists alone in the UK and that neither were effective at treating CFS. A study examining the effectiveness of a wide swath of physicians in the US – including ME/CFS specialists (eg. Drs. Peterson, Holtorf, Cheney, Bateman, Klimas, Bateman, Lapp, Enlander) alternative practitioners and MD’s, in combination with or without behavioral therapists, would tell us a great deal about the state of art of ME/CFS treatment.