• Welcome to Phoenix Rising!

    Created in 2008, Phoenix Rising is the largest and oldest forum dedicated to furthering the understanding of and finding treatments for complex chronic illnesses such as chronic fatigue syndrome (ME/CFS), fibromyalgia (FM), long COVID, postural orthostatic tachycardia syndrome (POTS), mast cell activation syndrome (MCAS), and allied diseases.

    To become a member, simply click the Register button at the top right.

PACE Trial - Control & Comparison Groups. Research and Discussion

oceanblue

Guest
Messages
1,383
Location
UK
The PACE thread has become a bit unwieldly so I'm starting a sub-thread just to look at the issue of Control and Comparison groups used in the study i.e SMC and APT. The PACE protocol describes the study as 'controlled' trial yet the final paper talks about 'comparison' of different treatments and carefully avoids ever using the phrase 'control group'. It formally compares GET and CBT with both SMC and APT. Some of the latest posts on the PACE thread will be relevant to this discussion too. I'll be posting on at least 3 relevant studies, as I have the energy. Hope no one objects to this approach.
 

oceanblue

Guest
Messages
1,383
Location
UK
Appropriate Control groups for RCTs of Psychological Therapies [eg PACE]. 2009

This is an unusually readable and thoughtful paper, and many of its points are highly relevant to the PACE Trial.

The Selection and Design of Control conditions for Randomized Controlled Trials of Psychological Interventions

Psychotherapy and Psychosomatics, Mohr 2009
[You can download the free full text from Academia.edu but you need to register first (free). This is the link to the download page (thanks to Alex3619)]

Abstract highlights

The randomized controlled trial (RCT) provides critical support for evidence-based practice using psychological interventions.

The control condition is the principal method of removing the influence of unwanted variables in RCTs. There is little agreement or consistency in the design and construction of control conditions. Because control conditions have variable effects, the results of RCTs can depend as much on control condition selection as on the experimental intervention. The aim of this paper is to present a framework for the selection and design of control conditions for these trials....

(full abstract)
 

oceanblue

Guest
Messages
1,383
Location
UK
Key Points of 'Appropriate Control Groups' study

Different types of control conditions produce significantly different effects on the outcomes of RCTs [2]:
Why Control Conditions Are Needed

The use of control conditions and randomization are the principal methods ensuring that an RCT is internally valid. Internal validity refers to the extent that the outcome for a trial can be attributed to the experimental treatment and not to any alternative explanation such as the natural course of the target problem. [note: or indeed response bias]
They discuss 3 potential pitfalls - the most pertinent ones to PACE are clinician allegiance to particular therapies (eg CBT rather than Pacing) and unblinded patients in the 'no treatment' group expecting to do less well and affecting outcomes (eg SMC vs the other therapies):
Clinician Selection and Allegiance Effects

... Clinicians also tend to arrive with an allegiance to the treatment they are to deliver, or develop one during the trial. If clinicians provide a treatment that they are well trained in, believe in, or know to be the experimental treatment, they will likely be more skilled and enthusiastic in their delivery of that treatment, which in turn may influence the outcome [11] . Failure to maintain equivalence in therapist selection and allegiance may threaten internal validity.

11. The researchers own therapy allegiances: a wild card in comparisons of treatment efficacy:

This report examines a possible distortion in the results of comparative treatment studies due to the association of the researcher's treatment allegiances with outcomes of those treatments. In eight past reviews a trend appeared for significant associations between the researcher's allegiance and outcomes of treatments compared. In a new review of 29 studies of treatment comparisons, a similar trend appeared. Allegiance ratings were based not only on the usual reprint method, but also on two new methods: ratings by colleagues who knew the researcher well, and self-ratings by the researchers themselves. The two new allegiance methods Interco related only moderately, but each allegiance measure correlated significantly with outcomes of the treatments compared, and when combined, the three measures explained 69% of the variance in outcomes Such an association can distort comparative treatment results. Our report concludes with how the researcher's allegiance may become associated with treatment outcomes and how studies should deal with these associations.
Unblinded 'No treatment' patients may have lower expectations, affecting outcomes:
Control Condition Effects

Although control conditions are an essential method of managing threats to internal validity, they may also have unintended effects on the outcomes of RCTs. Knowledge that one is not receiving treatment affects outcomes. Patients who become unblinded and learn they are receiving placebo do more poorly than patients who do not know they are receiving placebo [12, 13] ...

Assignment to no treatment may strengthen participants beliefs that they will not improve, thereby reducing the chance of spontaneous improvement [16] . Some control conditions may also be less credible than the experimental treatments with which they are paired. Treatment credibility has been shown to influence outcomes, particularly for brief interventions [17] . In short, control conditions may create threats to internal validity by influencing outcome expectancies and/or altering health-promoting behaviors.
Specific and nob-Specific controls

The authors then look at control conditions created by researchers (as opposed to waiting list, or 'treatment as usual'). These include 'Specific' controls and which aim to control for specific aspects of the active treament; in some ways this includes the SMC group of PACE since the 'active treatment' is SMC plus CBT etc while the 'control' is SMC alone.
When the control and experimental conditions are comparable in terms of treatment format, implementation, and credibility, the research design permits strong causal inferences.
Of course, in the case of SMC a few sessions with the doctor are not comparable in length, format or credibility with any of the 6 months of therapy so causal inferences are not appropriate.

'Non-specific' controls are particularly relevant to PACE, though are notable through their absence:
Luminaries such as Meehl [21] and Paul [22] argued that RCTs for psychological interventions required some form of control condition analogous to the placebo controls used in medical research. Nonspecific control conditions commonly hold constant some or all of the following: (1) attention or amount of treatment contact, (2) human interaction variables, including clinician warmth and empathy, or social support and interaction among group participants, and/or (3) provision of a treatment rationale that controls for participant outcome expectations [23, 24] .
Since PACE didn't control for attention, human interaction (13 hours of therapy + 3 session of SMC on average, vs 5 sessions of SMC in the SMC alone group) or treatment rationale (SMC was rated as 'logical' by 49% of patients vs 70-80% for participants) these could all contribute to the differences between SMC and CBT/GET.

Two final recommendations made by the authors that are relevant to the PACE trial:
.... To minimize clinician selection and allegiance effects, clinicians should be selected with the same level of care and, if possible, allegiance across treatment arms. When possible, clinicians should be blinded to whether they are delivering the experimental or the control treatment. Negative effects of a control condition on outcomes should at least be measured and controlled for statistically. When these threats to internal validity can not be controlled for, this should be clearly articulated in any report.
This best practice was not carried out in the PACE trial and the lack of such practice was not discussed in the PACE study. Since the Mohr study wasn't published until 2009, it would have been to late to change the experimental design, but the lack of discussion of having no non-specific control is something of a glaring omission.
Conflicts of Interest: While not specifically the focus of this paper, the influence of investigator allegiance has been well documented [11] . RCTs should include plans to manage such conflicts of interest, for example by including co-investigators of different allegiances in planning and monitoring.
The PACE study group was largely composed of enthusiastic proponents of CBT & GET - including at least one statistician who had publicly commented that CFS has a behavioural explanation.
 
Messages
13,774
The PACE thread has become a bit unwieldly so I'm starting a sub-thread just to

Quitter!!

I'm not even sure if PACE is meant to be an RCT. Sometimes they call it one, especially in the early days, but more recently they don't seem to. It's possible that they realised that calling the control group a 'control group' instead of Specialist Medical Care would lead to results which seemed to argue against the funding of any treatment.
 

*GG*

senior member
Messages
6,389
Location
Concord, NH
The PACE thread has become a bit unwieldly so I'm starting a sub-thread just to look at the issue of Control and Comparison groups used in the study i.e SMC and APT. The PACE protocol describes the study as 'controlled' trial yet the final paper talks about 'comparison' of different treatments and carefully avoids ever using the phrase 'control group'. It formally compares GET and CBT with both SMC and APT. Some of the latest posts on the PACE thread will be relevant to this discussion too. I'll be posting on at least 3 relevant studies, as I have the energy. Hope no one objects to this approach.

So what does SMC stand for?

GG
 

Sean

Senior Member
Messages
7,378
SMC = SSMC = Standard Specialist Medical Care

Though how much that differs from standard medical care is unclear.
 

Enid

Senior Member
Messages
3,309
Location
UK
Thanks ocean - all your hard work. I'm no scientist but reading some of this seems utter common sense (obviously) when exposed.
 

oceanblue

Guest
Messages
1,383
Location
UK
Quitter!!

I'm not even sure if PACE is meant to be an RCT.
Today this is just a splinter thread but one day we will take over the world.

The PACE paper called it a "parallel-group randomised trial", which Wikipedia lists as the best type of RCT, though PACE's ommission of the word 'controlled' still puzzles me slightly; more on that later.
 

biophile

Places I'd rather be.
Messages
8,977
Good find and analysis, oceanblue. Provides additional clues to the types of biases that could have occurred in the PACE Trial and explain the lack of objective improvements to match the subjective improvements. It has been nearly 14 months since the first PACE paper was published in the Lancet and we haven't seen any additional papers yet as promised.
 
Messages
13,774
The PACE paper called it a "parallel-group randomised trial", which Wikipedia lists as the best type of RCT, though PACE's ommission of the word 'controlled' still puzzles me slightly; more on that later.

Thanks OB. I missed that, and just noticed that they started omitting all mention of 'control' at some point. I could just be unsued to the terminology.

Thanks for summarising all of your other reading too.
 

Don Quichotte

Don Quichotte
Messages
97
SMC = SSMC = Standard Specialist Medical Care

Though how much that differs from standard medical care is unclear.

if you look at the supplementary material you will see that it is mostly standard care by psychiatrists and if I remember correctly many of them in training.
 

alex3619

Senior Member
Messages
13,810
Location
Logan, Queensland, Australia
Hi Don Quichotte, I have not had time to go into the standard care in detail yet but it occurs to me that this might be substandard care by our measure, and so they are comparing a poor treatment protocol versus substandard care. This might be worth looking into fiurther. Bye, Alex
 

oceanblue

Guest
Messages
1,383
Location
UK
From the PACE paper:

Specialist medical care (SMC)

SMC was provided by doctors with specialist experience in chronic fatigue syndrome (webappendix p 1). All participants were given a leaflet explaining the illness and the nature of this treatment. The manual was consistent with good medical practice, as presently recommended. Treatment consisted of an explanation of chronic fatigue syndrome, generic advice, such as to avoid extremes of activity and rest, specific advice on self-help, according to the particular approach chosen by the participant (if receiving SMC alone), and symptomatic pharmacotherapy (especially for insomnia, pain, and mood).

So probably better than many get from their GPs, given that most GPs will have very few (if any) ME patients, while at least these doctors will have experience of treating CFS patients. Nothing spectacular but seems a reasonable basic package.

edit: it's also worth noting that the SMC-only group improved from baseline more than the CBT/GET groups improved relative to SMC, suggesting that SMC could have been the most promising therapy trialled in the study :). This is made more compelling since the SMC group were probably less prone to response/expectation bias as they were in the 'non-treatment' group, were unimpressed wtih the logic of the treatment and were the least satisfied with treatment of all the groups.

edit2: it's also worth noting that all particpants received SMC e.g. it was SMC alone vs GET plus SMC.
 

Don Quichotte

Don Quichotte
Messages
97
Basically, what this study proves is that patients need the attention of their physicians, need hope and need to feel that they are receiving treatment which they and their physicians think and believe can alter the course of their illness or at least alleviate their symptoms and improve their quality of life. This is known from the time of Hippocrates.
 
Messages
1,446
.
Don Quichotte wrote: "Basically, what this study proves is that patients need the attention of their physicians, need hope and need to feel that they are receiving treatment which they and their physicians think and believe can alter the course of their illness or at least alleviate their symptoms and improve their quality of life. This is known from the time of Hippocrates."


.

Hey, Don - possibly millions of words have been written about the PACE Trial since it was first announced in 2004.

.
It does not appear that the PACE Trial has "proved" anything whatsoever.

I would direct you to Tom Kindlon's extensive analysis and deconstruction of the Trial.
 
Messages
1,446
.

It may not be generally known that in Britain CBT is a deeply political issue, proposed as a Government-endorsed official Therapy for All, universal panacea.



Member of the House of Lords, and London School Of Economics Professor Richard Layard (Government Happiness Tzar) announced and promoted a policy called "Therapy for All".

.
Richard Layard announced the training of 10,000 CBT therapists to address common anxiety, depression amongst the general population and the unemployed circa 2005/6. His promotion of CBT as universal therapy came across as proposing a universal panacea for all of societys ills. Richard Layard gained a lot of press coverage, and government attention for CBT at the time, but the issue of where the 10,000 therapists would come from, and who would pay for their training, was never addressed, and the project quietly sank eventually. Certain persons (ie me) had accurately predicted that the proposal to promote CBT as official Government policy would result in Therapy Turf Wars, and that proponents of other therapy modalities would retaliate.


Lord Layard has since resorted to promoting Happiness as universal panacea, via numerous international conferences and reports.




The Independent reports on Richard Layards plans to train 10,000 CBT therapists. Re Lord Richard Layard [Member of the House of Lords, LSE Economist and Government appointed Happiness Tzar] 2007:

Last year, he published a report calling for a network of 250 treatment centres to be
established staffed by 10,000 new therapists. These would provide
"talking therapies" such as cognitive behaviour therapy (CBT) - a
treatment proven to help relieve low-level depression and anxiety
which enables patients to overcome negative thinking.



Lord Richard Layard to the House of Lords, 2009:
we have an excellent plan for providing face-to-face CBTobviously,
computerised CBT will be there as wellto all adults who need it.. In a debate on the future of CBT in this country we must focus mainly on CBT provided by live therapists to patients in the new way that the Government will make possible in this country. This is the most radical improvement in psychological therapy services ever undertaken.





http://news.independent.co.uk/uk/health_medical/article2516748.ece

'Happiness tsar' warns of therapy funding shortage

By Sophie Goodchild, Chief Reporter

The "Independent": 06 May 2007

Radical plans to set up free "therapy for all" centres across the
country could fail without proper funding, the Government's
"happiness tsar" has warned.

Professor Richard Layard, the Labour peer and No 10 adviser, said he
is concerned that patients suffering from anxiety and depression will
not benefit unless cash is set aside for training up therapists.

In an interview with The Independent on Sunday, Lord Layard said:
"There should be a proper plan for dealing with this problem and not
a bit of a fudge of a bit of money that ends up with the creation of
dumbed-down workforce."

The eminent economist has said he believes mental illness is the
single greatest threat to a happy life. Last year, he published a
report calling for a network of 250 treatment centres to be
established staffed by 10,000 new therapists. These would provide
"talking therapies" such as cognitive behaviour therapy (CBT) - a
treatment proven to help relieve low-level depression and anxiety
which enables patients to overcome negative thinking.

Patricia Hewitt, the Secretary of State for Health, has said she
wants to increase the use of counselling and announced the opening of
two talking therapy pilot centres in Doncaster and East London.

This Thursday at a conference in London, Ms Hewitt is expected to
further endorse the long-term benefits of therapy, as opposed to the
quick fix results from anti-depressants, for those suffering from
mild depression and anxiety. However, the Government has yet to
commit to a comprehensive programme for talking therapies on the NHS
on the scale suggested by Lord Layard. He says that unless the money
is set aside now, then neither the patients nor the economy will
benefit.

"I've never said CBT is a magic bullet," said Professor Layard,
professor emeritus of economics at the London School of Economics.
"But there is the danger that if people are not properly trained, the
patients will not benefit."

An estimated one million people suffer from clinical depression and
four million from clinical anxiety in Britainy. But only one in 10
gets to see a therapist and often only after a long wait.

This newspaper revealed earlier this year that around a third of NHS
trusts are struggling with a backlog of patients desperate for
talking therapies. More than 90 per cent of trusts have waiting lists
of longer than a year for CBT. Wakefield West PCT in Yorkshire has a
waiting list of 78 weeks.





~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~





Lord Richard Layard, LSE Economist, etc, to the House of Lords 2009:

http://www.theyworkforyou.com/lords/?id=2009-03-18a.306.0


.....I should like finally to say a word on children. The programme that I have
been discussing relates to adults, but of course many of those who suffer as
adults also suffered as children. Child mental illness is even more tragic than
for adults. It is also the source of so many of our social problems. Ten per
cent of all children would be diagnosed as suffering from mental illness of all
kinds, and 5 per cent from anxiety disorders for which the prime treatment is
CBT. Of these children only a quarter are currently receiving specialist help or
have seen a specialist in the past year. That is just not good enough. Although
we have child and adolescent mental health services which in many cases are
excellent, their capacity is just too small.

There are many children in real need who get turned away or do not get referred
because the waiting list is too long, and not all the services are delivered in
accordance with the NICE guidelines.
What we now need is a strategy for expanding and upgrading CAMHS as well as
adult services. A number of us have suggested a five-year plan which would train
200 extra child therapists every year and be adequately funded to pay local
services for providing the on-the-job training within the NICE guidelines. I
think that that would be a powerful formula. It would cost no more than 35
million by the final year of the next spending round. I very much hope that the
Minister can undertake that these proposals will be seriously considered for
very high priority in the department's spending bid.

To conclude, we have an excellent plan for providing face-to-face CBTobviously,
computerised CBT will be there as wellto all adults who need it, but it still
needs to be refinanced for the second half. We need to do something similar for
children. I have every hope that the Government will do this because they have
shown their willingness to bite this bullet which had been neglected for so many
decades. This Government have been outstanding so far in their approach to this,
after decades of neglect. I really hope that they will complete the job.

.