Esther12
Senior Member
- Messages
- 13,774
They're all open access.
http://www.thelancet.com/journals/lanpsy/article/PIIS2215-0366(15)00089-9/fulltext
http://www.thelancet.com/journals/lanpsy/article/PIIS2215-0366(15)00085-1/fulltext
http://www.thelancet.com/journals/lanpsy/article/PIIS2215-0366(15)00044-9/abstract
http://www.thelancet.com/journals/lanpsy/article/PIIS2215-0366(15)00054-1/fulltext
http://www.thelancet.com/journals/lanpsy/article/PIIS2215-0366(15)00110-8/fulltext
http://www.thelancet.com/journals/lanpsy/article/PIIS2215-0366(15)00114-5/fulltext
Even by the standards we've come to expect from PACE, the author's reply is weak.
eg:
Old McGrath points out that the did not show mediation, but only an association. In their reply they start by saying "we noted that fear avoidance beliefs mediated both GET and CBT" and then go on to "McGrath reiterates the limitations of this mediator analysis, which we have already acknowledged in the Article." So do they accept that their data does not show fear avoidance beliefs mediated both GET and CBT or not? If you go on ignoring the limitations acknowledged in your article, expect people to go on pointing this out to you.
Will come back to slag it off a bit more soon, but thought I should let others have some fun too.
In non-blinded trials, self-report measures could mislead
George Faulkner
Trudie Chalder and colleagues (February, 2015)1 report that “both self-report and objective measures were used, and both were found to mediate treatment effects, lending credence to the results”. However, little attention seems to have been paid to the differences between self-report and more objective results from this non-blinded trial.
Within the four groups of the PACE trial, all patients received 3–6 sessions of a basic intervention, specialist medical care. This intervention was all that was provided to one group. In the other three groups, participants attended 12–15 sessions of additional therapy. Those providing cognitive behavioural therapy (CBT) and graded exercise therapy (GET) encouraged gradual increases in activity (mostly walking for GET), whereas those providing adaptive pacing therapy (APT), a novel therapy created specifically for the trial, were told that an important consideration in APT is the 70% rule: never going beyond 70% of a person's perceived energy limit. A non-blinded trial assessing such interventions risks distortions, such as response bias, generating misleading results, particularly since patients in the CBT and GET groups were told during treatment that these therapies had already been shown to be effective.
A 2001 systematic review2 commented on the difficulty of interpreting subjective measures in CFS studies and stated that “a more objective measure of the effect of any intervention would be whether participants have increased their working hours, returned to work or school, or increased their physical activities”. After publication of the PACE trial protocol, concerns were expressed about the restricted use of objective outcome measures, and particularly the decision to use actometers to measure activity only at baseline, rather than as an outcome measure.3 Chalder and colleagues had previously argued that “The heart of CBT is a behavioural approach to the impairment of activity that is part of the definition of CFS,”4 and that an increase in activity, “…must ultimately be the aim of any treatment”.5
In response to these concerns, the PACE trial management group stated that “we have used several objective outcome measures; the six-minute walking test, a test of physical fitness, as well as occupational and health economic outcomes”.3 Since then, unpublished actometer results from three previous trials of CBT for CFS have been compiled and released as a meta-analysis that showed that CBT led only to improvement in self-report measures and did not allow patients to increase activity;6 now results1 from the PACE trial have shown that CBT did not lead to improvement in employment, physical fitness, or 6-min walking test outcomes. GET did not lead to improvement in employment outcomes or physical fitness, and the six-minute walking test showed an improvement that was significant but fell short of the criteria used elsewhere in the trial to define a clinically useful difference between means (0·5 of SD at baseline).
It is important to recognise that the design of, and results from, the PACE trial mean that this latest report might merely be providing information about how response bias can change patient self-report measures.
I declare no competing interests.
References
1Chalder, T, Goldsmith, KA, White, PD, Sharpe, M, and Pickles, AR. Rehabilitative therapies for chronic fatigue syndrome: a secondary mediation analysis of the PACE trial. Lancet Psychiatry. 2015; 2: 141–152
2Whiting, P, Bagnall, AM, Sowden, AJ, Cornell, JE, Mulrow, CD, and Ramírez, G. Interventions for the treatment and management of chronic fatigue syndrome: a systematic review. JAMA. 2001; 286: 1360–1368
3Various. Comments on “Protocol for the PACE Trial”. http://www.biomedcentral.com/1471-2377/7/6/comments. ((accessed February 9, 2015).)
4Chalder, T, Deale, A, Wessely, S, and Marks, I. Cognitive behavior therapy for chronic fatigue syndrome. Am J Med. 1995; 98: 419–422
5Wessely, S, David, A, Butler, S, and Chalder, T. Management of post-viral fatigue syndrome. Br J Gen Pract. 1990; 40: 82–83
6Wiborg, JF, Knoop, H, Stulemeijer, M, Prins, JB, and Bleijenberg, G. How does cognitive behaviour therapy reduce fatigue in patients with chronic fatigue syndrome? The role of physical activity. Psychol Med. 2010; 40: 1281–1287
http://www.thelancet.com/journals/lanpsy/article/PIIS2215-0366(15)00089-9/fulltext
Omission of data weakens the case for causal mediation in the PACE Trial
Simon McGrath
In the analysis of the PACE trial, Trudie Chalder and colleagues (February, 2015)1 concluded that fear avoidance and exercise tolerance were potent mediators of the effect of cognitive behaviour therapy (CBT) and graded exercise therapy (GET) on chronic fatigue syndrome.2 Such conclusions might be premature.
The study design had several good qualities including a large sample size, the measurement of most candidate mediators mid-treatment, and the control of baseline variables such as age and illness duration to reduce the chance of unseen confounders. The authors rightly stressed the importance of clear temporal separation when establishing the case for causal mediation, which requires that treatments first cause changes in mediators, such as fear avoidance, and the mediators subsequently change the outcomes–in this instance fatigue and physical function.3 Data from mediators at 12 or 24 weeks and outcome data at 52 weeks were used, to “have the maximum possible separation between mediator and outcome measurements”.1
The primary outcomes had likewise been measured at 12 and 24 weeks, and they mirrored the pattern of change in putative mediators.1 Meaningful temporal separation is not established by mere omission of contemporaneous outcome data, and the absence of such separation weakens claims of mediation.
The case for the walking test as a mediator is particularly uncertain. This variable was only measured at 24 weeks when most changes in primary outcomes had already taken place, and most therapy sessions had been completed.2 The study had walking test data for only three-quarters of participants, with the possibility that data was not missing at random, but caused by deterioration of condition and patients being too unwell or unwilling to attempt the test.
This study demonstrated association between putative mediators and outcomes, but not the causal link central to establishing mediation. For example, patients feeling less fatigued could lead to a reduction in beliefs classed as fearful or aversive. In this instance a primary outcome would be the mediator of the belief, a scenario that the report did not explore.
Although the authors acknowledged that simultaneous change in their study's outcomes means mediators and outcomes might affect each other reciprocally, the abstract and conclusions unambiguously assert mediation.1 Claims of causal mediation, and the linked assertion that this strengthens the theoretical treatment model, are premature.
The planned future paper, using all the longitudinal data including all contemporaneous outcome data,2 might yet throw more light on the issue of how the effect of CBT and GET on self-reported fatigue and function are mediated.
I blog from time to time, unpaid, about research on chronic fatigue syndrome or myalgic encephalitis.
References
1White, PD, Goldsmith, KA, Johnson, AL 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: 823–836
2Chalder, T, Goldsmith, KA, White, PD, Sharpe, M, and Pickles, AR. Rehabilitative therapies for chronic fatigue syndrome: a secondary mediation analysis of the PACE trial. Lancet Psychiatry. 2015; 2: 141–152
3MacKinnon, DP. Introduction to Statistical Mediation Analysis. Taylor and Francis, New York; 2008: 8–102 (348–395.)
4White, PD, Sharpe, MC, Chalder, T, DeCesare, JC, and Walwyn, R. Protocol for the PACE trial: a randomised controlled trial of adaptive pacing, cognitive behaviour therapy, and graded exercise, as supplements to standardised specialist medical care versus standardised specialist medical care alone for patients with the chronic fatigue syndrome/myalgic encephalomyelitis or encephalopathy. BMC Neurol. 2007; 7: 6
http://www.thelancet.com/journals/lanpsy/article/PIIS2215-0366(15)00085-1/fulltext
Post-exertional malaise in chronic fatigue syndrome
Frank Twisk
I read with interest a recent contribution by Trudie Chalder and colleagues1 to The Lancet Psychiatry (Feburary, 2015), in which they present a secondary mediation analysis of the PACE trial. Their main finding was that fear avoidance beliefs were the strongest mediator for both cognitive behaviour therapy (CBT) graded exercise therapy (GET) for chronic fatigue syndrome. Both CBT and GET are qualified by the investigators as “moderately effective treatments”.1
However, looking at the data published by the authors in 2011,2 the preposition that CBT and GET are effective interventions for chronic fatigue syndrome (or myalgic encephalomyelitis, which is an equivalent to chronic fatigue syndrome according to the authors) can be disputed.
First, the participants investigated were those “meeting Oxford criteria for chronic fatigue syndrome”.2 These criteria primarily select people with chronic fatigue, not chronic fatigue syndrome, let alone myalgic encephalomyelitis. As the authors state “the findings can be generalised to patients who also meet alternative diagnostic criteria for chronic fatigue syndrome and myalgic encephalomyelitis but only if fatigue is their main symptom”.2 The criteria referred to, however, are neither the widely accepted Fukuda criteria for chronic fatigue syndrome nor the consensus criteria for myalgic encephalomyelitis.3 The presence of fatigue is not sufficient to meet the diagnosis chronic fatigue syndrome.3 Moreover, the recent criteria for myalgic encephalomyelitis even abolish the “chronic fatigue” criterion, since it is “the most confusing and misused criterion”.4
Second, the effects of CBT and GET are by far insufficient to qualify as “moderately effective”. When looking at subjective measures, such as fatigue and physical functioning, they stated “the size of this effect is moderate”.2 The proportion of participants within abnormal low cutoff scores for both fatigue and physical functioning after treatment was 30% for CBT, 28% for GET, which are not that high compared with 15% for standard care. When looking at the only objective measure used (ie, distance in m walked in 6 min), the improvements with CBT or GET were very minor.5 The thesis that CBT and GET are effective therapies for chronic fatigue syndrome or myalgic encephalomyelitis is not tenable. By definition, exertion causes post-exertional malaise (a prolonged aggravation of typical symptoms such as muscle weakness, pain, cognitive deficits, etc) in all patients with myalgic encephalomyelitis. This odd phenomenon can plausibly be explained by several exercise-derived physical abnormalities. For example, exercise induces a substantial drop in the (already low) oxygen uptake and instigates an increase in pain receptor genes. So, it is conceivable that fear avoidance in patients with post-exertional malaise is a reasonable and learned response, serving as a rational defence mechanism to avoid long-lasting relapses. Moreover, patients with post-exertional malaise are those who simply cannot change their attitude towards fear avoidance.
In conclusion, to resolve the debate about the proposed effectiveness of exercise and fear avoidance in myalgic encephalomyelitis and chronic fatigue syndrome, it is crucial to make a clear distinction between patients with myalgic encephalomyelitis (experiencing muscle weakness, cognitive impairment, but above all post-exertional malaise), and patients with chronic fatigue syndrome and other forms of chronic fatigue, and to employ objective measures to assess symptoms and improvement.
I am associated with a Dutch ME/CFS patient foundation in a voluntary (non-paid) capacity.
References
1Chalder, T, Goldsmith, KA, White, PD, Sharpe, M, and Pickles, AR. Rehabilitative therapies for chronic fatigue syndrome: a secondary mediation analysis of the PACE trial. Lancet Psychiatry. 2015; 2: 141–152
2White, PD, Goldsmith, KA, Johnson, AL 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: 823–836
3Twisk, FNM. The status of and future research into myalgic encephalomyelitis and chronic fatigue syndrome: the need of accurate diagnosis, objective assessment, and acknowledging biological and clinical subgroups. Front Physiol. 2014; 5: 109
4Carruthers, BM, van de Sande, MI, de Meirleir, KL et al. Myalgic encephalomyelitis: international consensus criteria. J Intern Med. 2011; 270: 327–338
5Kindlon, T. The PACE trial in chronic fatigue syndrome. Lancet. 2011; 377: 1833
http://www.thelancet.com/journals/lanpsy/article/PIIS2215-0366(15)00044-9/abstract
Doubts over the validity of the PACE hypothesis
Robert Courtney
The open-label PACE trial1 was designed to compare efficacy of cognitive behavioural therapy (CBT), graded exercise therapy (GET), and adaptive pacing therapy (APT), when added to a standardised intervention resembling usual care (specialist medical care) as treatments for chronic fatigue syndrome.
CBT and GET were based on a hypothetical model of chronic fatigue syndrome whereby symptoms and disability are assumed to be primarily perpetuated by a fear of symptoms and activity, avoidance of activity, and associated deconditioning (lack of fitness). CBT and GET were intended to address these assumed perpetuating factors and to “reverse”2 the illness, leading to “an increase in activity capacity” and ultimately a recovery. CBT mainly focused on addressing fear and avoidance, and GET on reversing deconditioning. The trial's manual for therapists explains the model of illness for GET: “Participants are encouraged to see symptoms as temporary and reversible, as a result of their current physical weakness, and not as signs of progressive pathology”.1, 2
In the latest analysis of the PACE trial in The Lancet Psychiatry (February, 2015),3 Trudie Chalder and colleagues explain that CBT and GET did not improve the assumed lack of fitness, as assessed via a step test, and that “fitness measures do not appear to mediate the effects of either treatment”. Chalder and colleagues acknowledge that the trial outcomes do not support the hypothetical deconditioning model of GET for chronic fatigue syndrome.3
The PACE trial had two objective measures of physical capacity1, 3 for which data have been published: a step test and a 6 min walking test. CBT did not significantly improve either measure. Chalder and colleagues argue1, 3 that their mediation analysis strengthens the validity of the hypothetical fear-avoidance model of CBT; however, the failure of CBT to significantly improve either of the trial's objective measures of physical capacity suggests that CBT failed to reverse the illness or lead to an increase in activity capacity. The objective outcomes do not support the fear-avoidance model of CBT.
For the trial's self-report primary outcome measures (fatigue and physical function), a moderate effect size was reported for CBT and GET, and (an additional) 11–15% of participants clinically responded to treatment when CBT or GET were added to specialist medical care.1 However, as the PACE trial was open-label, we cannot know if some or all of the reported improvements reflect issues with the trial's methodology and are a result of biases, such as response bias and a placebo effect.4, 5 For example, unlike the specialist medical care and APT groups, participants receiving CBT and GET were told that the therapies were highly effective and were encouraged to see symptoms as temporary and reversible, possibly leading to differing expectations across the various trial groups.2
The PACE trial is to our knowledge the largest investigation of CBT and GET for chronic fatigue syndrome to date. The deconditioning hypothesis was not supported, and the fear-avoidance hypothesis was not supported by the trial's objective outcomes. These factors, along with the disappointing self-report clinical response rates for CBT and GET in an open-label trial, cast substantial doubt over the validity of the fear-avoidance and deconditioning hypothesis for chronic fatigue syndrome.
I declare no competing interests.
References
1White, PD, Goldsmith, KA, Johnson, AL 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: 823–836
2Bavinton, J, Darbishire, L, and White, PD. PACE manual for therapists; graded exercise therapy for CFS/ME. http://www.pacetrial.org/docs/get-therapist-manual.pdf; 2004. ((accessed Jan 16, 2015).)
3Chalder, T, Goldsmith, KA, White, PD, Sharpe, M, and Pickles, AR. Rehabilitative therapies for chronic fatigue syndrome: a secondary mediation analysis of the PACE trial. Lancet Psychiatry. 2015; 2: 141–152
4Kindlon, TP. Objective measures found a lack of improvement for CBT & GET in the PACE Trial: subjective improvements may simply represent response biases or placebo effects in this non-blinded trial. BMJ Rapid Response. 2015; http://www.bmj.com/content/350/bmj.h227/rr-10. ((accessed Jan 21, 2015).)
5Wilshire, CE. Re: Tackling fears about exercise is important for ME treatment, analysis indicates. BMJ Rapid Response. 2015; http://www.bmj.com/content/350/bmj.h227/rr-7. ((accessed Jan 21, 2015).)
http://www.thelancet.com/journals/lanpsy/article/PIIS2215-0366(15)00054-1/fulltext
Methods and outcome reporting in the PACE trial
Sean B M Kirby
In The Lancet Psychiatry (February, 2015), Trudie Chalder and colleagues1 reported a planned secondary mediation analysis of the PACE trial, citing the trial's published protocol. However, it is not clear when the specifics of their published mediation analysis were decided on. Few details on the assessment of mediators are included within the trial's protocol: a planned split-test to assess internal validity is described, and yet results for this were not reported. In 2011, Kimberley Goldsmith gave an oral presentation on still unreleased results for a mediation analysis of PACE, which attempted to account for problems with bias and unmeasured confounders by using an instrumental variables approach. The meeting abstract for this presentation reported that “there was modest mediation of CBT and GET effects (approximately 20% of the total effect).”2 Including the results from this analysis might have aided with interpretation of their recent Article.1
Previous failures to release data, and deviations from pre-specified outcomes, has led to misleading recovery rates for cognitive behavioural therapy (CBT) and graded exercise therapy (GET) being reported within both the lay and academic press.3, 4 The weakness of the PACE trial's post-hoc recovery criteria was criticised for being “contradictory” by an evidence review for the Agency for Healthcare Research and Quality,5 while a commentary6 used problems with the justifications for those post-hoc recovery criteria to illustrate the need for more data to be released, and greater care taken, when studies report recovery rates, to “minimize the potential for misunderstanding”.
Handoll and Hanchard argued that a cardinal rule for investigating whether a medical innovation truly works is “the need to separate the clinical evaluation of innovations from their innovators, who irrespective of any of their endeavours to be ‘neutral’ have a substantial investment, whether emotional, perhaps financial, or in terms of professional or international status, in the successful implementation of their idea.”7 There seems to be growing recognition that results from the PACE trial have been presented in a way which has exaggerated the benefits of CBT and GET to patients.2, 5, 6
To allow more informed judgements to be made about the efficacy of the treatments assessed in this large, expensive, and supposedly definitive trial, it would help to release the results for all of the outcomes as laid out in the published trial protocol, so that other researchers, and clinicians and patients, can decide for themselves whether it is the initial or the more recently devised outcomes that are of greater value.
I declare no competing interests.
References
1Chalder, T, Goldsmith, KA, White, PD, Sharpe, M, and Pickles, AR. Rehabilitative therapies for chronic fatigue syndrome: a secondary mediation analysis of the PACE trial. Lancet Psychiatry. 2015; 2: 141–152
2Goldsmith, K, Chalder, T, White, P, Sharpe, M, and Pickles, A. How do treatments for chronic fatigue syndrome work? Exploration of instrumental variable methods for mediation analysis in PACE—a randomised controlled trial of adaptive pacing therapy, cognitive behaviour therapy, graded exercise therapy, and specialist medical care. Trials. 2011; 12: A144
3Hope, J. Chronic fatigue victims ‘suffer fear of exercise’: Patients are anxious activities such as walking could aggravate the condition. London: Mail Online. http://www.dailymail.co.uk/health/a...e-anxious-activities-aggravate-condition.html; 2015. ((accessed Feb 15, 2015).)
4BMJ. Short Cuts: ‘All you need to read in the other general journals’. BMJ. 2011; 342: d1168
5Smith, MEB, Nelson, HD, Haney, E et al. Diagnosis and treatment of myalgic encephalomyelitis/chronic fatigue syndrome. Evidence Report/Technology Assessment No 219. Agency for Healthcare Research and Quality, Rockville, MD; 2014http://www.effectivehealthcare.ahrq.gov/reports/final/cfm. ((accessed Feb 10, 2015).)
6Matthees, A. Assessment of recovery status in chronic fatigue syndrome using normative data. Qual Life Res. 2014; Sept 23; DOI: http://dx.doi.org/10.1007/s11136-014-0819-0
7Handoll, H and Hanchard, N. From observation to evidence of effectiveness: the haphazard route to finding out if a new intervention works. Cochrane Database Syst Rev. 2014; 5: ED000081
http://www.thelancet.com/journals/lanpsy/article/PIIS2215-0366(15)00110-8/fulltext
Methods and outcome reporting in the PACE trial–Author's reply
Trudie Chalder, Kimberley A Goldsmith, Peter D White, Michael Sharpe, Andrew R Pickles
We did a randomised controlled trial (n=640) and found two rehabilitative treatments, graded exercise therapy (GET) and cognitive behaviour therapy (CBT), added to specialist medical care to be superior to adaptive pacing therapy added to specialist medical care and specialist medical care alone in improving fatigue and physical functioning for patients with chronic fatigue syndrome.1 These findings were robust irrespective of how we defined the illness.1 In a mediational analysis2 we examined how the effective treatments worked. We noted that fear avoidance beliefs mediated both GET and CBT.2 This result was pertinent for both physical functioning and self-reported fatigue. It is also consistent with a review3 of the role of beliefs in chronic fatigue syndrome and fibromyalgia, which suggested that fear and avoidance of movement were associated with poorer outcomes.
Most correspondents make criticisms of the trial as a whole, which we have already addressed in detail, and which might reflect the apparent campaign to bring the robust findings of the trial into question.4, 5 McGrath reiterates the limitations of this mediator analysis, which we have already acknowledged in the Article.2 We intend to further explore multiple mediator effects and the potential effects of outcomes on mediators in future papers. Some correspondents suggest that objective outcomes are more important than self-report measures. In view of the subjective nature of fatigue and the overriding importance of the patients own judgement of their functioning, we believe that self-rated outcomes remain the most important ones. Although it is entirely understandable that some patients with chronic fatigue syndrome are cautious about engaging in activity and exercise, increasing those activities gradually offers both improved functioning and reduced fatigue in some patients. Our findings also suggest that better targeting of fearful beliefs in treatment may result in better outcomes.
PDW has done voluntary and paid consultancy work for the UK Government and a reinsurance company. TC has received royalties from Sheldon Press and Constable and Robinson. MS has done voluntary and paid consultancy work for the UK Government, has done consultancy work for an insurance company, and has received royalties from Oxford University Press. KAG and ARP declare no competing interests.
References
1White, PD, Goldsmith, KA, Johnson, AL..., and on behalf of the PACE trial management group. 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: 823–836
2Chalder, T, Goldsmith, KA, White, PD, Sharpe, M, and Pickles, A. Rehabilitative therapies for chronic fatigue syndrome: a secondary mediation analysis of the PACE trial. Lancet Psychiatry. 2015; 2: 141–152
3Nijs, J, Roussel, N, Van Oosterwijck, J et al. Fear of movement and avoidance behaviour toward physical activity in chronic-fatigue syndrome and fibromyalgia: state of the art and implications for clinical practice. Clin Rheumatol. 2013; 32: 1121–1129
4Lancet. Patients' power and PACE. Lancet. 2011; 377: 1808
5White, PD, Chalder, T, and Sharpe, M. The planning, implementation and publication of a complex intervention trial for chronic fatigue syndrome: the PACE trial. BJPsych Bull. 2015; 39: 24–27
http://www.thelancet.com/journals/lanpsy/article/PIIS2215-0366(15)00114-5/fulltext
Even by the standards we've come to expect from PACE, the author's reply is weak.
eg:
Old McGrath points out that the did not show mediation, but only an association. In their reply they start by saying "we noted that fear avoidance beliefs mediated both GET and CBT" and then go on to "McGrath reiterates the limitations of this mediator analysis, which we have already acknowledged in the Article." So do they accept that their data does not show fear avoidance beliefs mediated both GET and CBT or not? If you go on ignoring the limitations acknowledged in your article, expect people to go on pointing this out to you.
Will come back to slag it off a bit more soon, but thought I should let others have some fun too.
Last edited: