PACE 2011 summary

*STILL UNDER CONSTRUCTION: the statistical analysis is hard to make simple*

I was asked by z to provide a simpler version of the 2011 PACE paper. This is a simplified version of what the paper says in order to help you read it. It is not what I think of the trial. It is most certainly not an endorsement of the trial. Bad science makes me angry. There’s a summary at the top and then it  follows the layout of the actual paper. It’s pretty long but you can just read whichever section(s) you’re interested in. If you want to read the original paper it’s here.

Summary:

The PACE trial looked at patients who met the Oxford criteria for CFS. They were put in 4 treatment groups:

  • CBT + medical care
  • GET + medical care
  • adaptive pacing therapy + medical care
  • just medical care.

The medical care provided was not what you would just get with a GP, but more specific to living with CFS.

Patients received treatment and were assessed after a year for how they were functioning.

They scored themselves on how they felt they were doing with fatigue, how they felt they were doing physically and how they felt their health was overall.

The researchers concluded:

CBT and GET plus medical care are effective treatments for CFS. They have a moderate effect.

Adaptive pacing was no different to just medical care.

These findings apply to patients with CFS whose main symptom is fatigue.

All four treatments are safe.

Since the treatments are only moderately effective, more research into better treatments is needed. Behavioural treatments being effective does not mean the illness is psychological.

Limitations of trial:

  • Patients who could not attend hospital could not be part of the trial.
  • Results only apply to patients who were referred from their GP to other care.
  • The medical care provided in the trial to all groups is not the same as the usual care that patient’s get with a GP.
  • The most common reason for a patient not entering the trial was not actually having CFS
  • Medical care was not as closely monitored as the other therapies.
  • Those with just medical care received more medical care than those also receiving therapy.
  • Masking was not possible.
  • The ratings used to work out if the treatment had worked were self-rated by the participants and were subjective.

 

PACE 2011 paper

Contents:
Summary
Introduction
Selection
Treatment
Outcomes
Statistical Analysis
Results
Discussion

Summary:

Four groups for one year:
1. Cognitive Behavioural Therapy (CBT) plus medical care – 161 people
2. Graded Exercise Therapy (GET) plus medical care – 160 people
3. Adaptive Pacing Therapy plus medical care – 160 people
4. Just medical care – 160 people

Each group was rated for fatigue. The scores of the groups were compared.

CBT group had 3.4 points less fatigue than medical care only.
GET group had 3.2 points less fatigue than medical care only.
APT group had 0.7 points less fatigue than medical care only.

Conclusion: Adding CBT or GET to medical care moderately improves fatigue.

Back to contents

Introduction:

Chronic Fatigue Syndrome (CFS) is being researched. Some people say this is the same as Myalgic Encephalomyelitis and others don’t. 0.2% – 2.6% of the population suffers from CFS. When untreated people are unlikely to get much better.

Therapies can make it more likely patients will get better. Some research says CBT and GET help patients. Patient groups say that they are harmful. Patient groups recommend health care and pacing.

This trial compares CBT, GET and adaptive pacing therapy on top of medical care. These three are compared to just having specialist medical care. The expected outcome was that CBT and GET would work better than pacing, and pacing would work better than just medical care.

Back to contents

How patients were selected:

Oxford criteria:

  • Main symptom is fatigue
  • Significant disability
  • No psychiatric diagnosis

OR

  • 6 or more out of 11 points on Chalder Fatigue questionnaire
  • less than 60 points on physical function subscale. This was later changed to less than 65.

Patients who were excluded:

  • under 18’s
  • at risk of self-harm
  • unable to attend hospital appointments
  • couldn’t read or speak English
  • medical needs that “made participation inappropriate”
  • had already tried PACE

Patients were also assessted via International Criteria for CFS and London Criteria for ME

Randomisation and Masking

[Masking means making sure people who are assessing the trial do not know which patients received which treatment.] This was impractical for therapists so they were not masked. Doctors and therapists giving treatment knew which patients received which treatment. The statistician and those who were not involved in treatment were masked.

Patients were sorted into different groups by computer program. This added randomisation but also made sure each group had a roughly even split of types of patient. This took into account which centre they were at, other CFS criteria, ME criteria and depression criteria. The trial team was not involved in this.

Most of the 6 centres had one therapist.

How the trial was carried out:

Therapists were trained by a therapist with experience in CFS.
Individual therapy was provided once a month.
Group therapy was provided once every three months.
Audio was recorded.

Two independent clinicians listened to recordings of the 10th session of 62 of the patients. They did not know which treatment the patient was allocated and recorded if the name of the treatment came up during the session. They rated the recordings on:

  • Therapy type
  • Sticking to the therapy manual
  • Relationship between therapist and participant.

Doctors were trained in specialist medical care. Competence was assessed and sticking to the medical care manual was monitored for most.

For each type of therapy, 10 sessions were needed to be able to analyse if it worked. For specialist medical care, 3 sessions were needed.

Sessions attended, leaving treatment, having other treatment and leaving follow-up were recorded.

Back to contents

Treatments:

Treatments were kept the same in different places by providing manuals for doctors, therapists and participants.

Results were collected:

  • At the start
  • In the middle of therapy (12 weeks)
  • After therapy (24 weeks)
  • After 1 year

Results were self-rated by the participant, while face-to-face in the clinic.

Specialist medical care:

Participants were offered at least 3 sessions, more if it seemed they needed it.
The manual followed current good medical practice.

Treatment:

  • Given a leaflet explaining CFS and specialist medical care.
  • CFS was explained.
  • Generic advice given:
  • Avoid extrme activity
  • Avoid extreme rest
  • Self-help advice
  • Drugs for insomnia, pain or mood if needed.

Therapies:

  • Up to 14 sessions over the first 23 weeks:
    • Once a week for the first four weeks.
    • Then once every two weeks.
  • Then an additional “booster” session at 36 weeks.

These were delivered face to face, sometimes by telephone.

Adaptive Pacing Therapy (APT):

Based on “envelope theory” of CFS: CFS is seen as an organic disease which can’t be reversed by changing behaviour or thought.
CFS causes a reduced and finite amoount of energy – the “energy envelope”.

Goal of therapy: Adapt to the energy amount available.

Treatment:

  • Helping to plan activities to reduce or avoid fatigue.
  • Help achieve prioritised activities.
  • Help provide conditions for natural recovery.
  • Diary used to find links between activity and fatigue.
  • Working out early warning signs of too much activity.
  • Limiting stress and demands.
  • Planned rest and relaxation.
  • Alternating different activity types.
  • Not doing things which require more than 70% of perceived energy envelope.
  • Doing more activities was encouraged as long as symptoms did not get worse.

A manual was created to describe this treatment as one did not exist.
APT was provided by occupational therapists.

Cognitive Behavioural Therapy (CBT):

Based on “fear avoidance theory” of CFS: CFS is seen as reversible. CFS caused by avoiding activity due to fear of activity.

Goal of therapy: change behaviour and thoughts which cause symptoms. Symptoms should then get better.

Treatment:

  • Challenge unhelpful thoughts by carrying out activity.
    Eg: challenging fears of activity or fears of symptoms.
  • Make baseline of rest and regular sleep pattern.
  • Slow increases in physical and mental activity.
  • Help with social and emotional problems.

The manual used was based on previous trials.
CBT was provided by psychologists and nurse therapists.

Graded Exercise Therapy (GET):

Based on “deconditioning and exercise intolerance theories”: CFS continues because activity is avoided and the body is not used to activity. Because the body is not used to it, activity seems even harder, so is avoided.
Goal of therapy: slowly increase physical activity and so get body used to it. Fatigue and disability should then get better.

Treatment:

  • Make a baseline of physical activity which can be managed.
  • Increase time spent in physical activity.
  • Aim for 30 minutes of light exercise 5 times a week.
  • Then increase intensity of exercise.

The manual used was based on previous trials.
GET was provided by physiotherapists or exercise physiologists.

Back to contents

Trial Outcomes:

Primary outcomes were worked out from:

  • Chalder Fatigue Questionnaire.
    Fatigue scores from 0 to 33. A low number is less fatigue.
    Each question was rated 0-3 by the patient. Before analysis the scores were adjusted to 0-1. This made the fatigue scores from 0 to 11.
  • Physical function subscale: 0-100, high number is better function.

Safety outcomes were worked out from:

Two doctors and a psychiatrist who all specialise in CFS reviewed all bad events/reactions on their own. They did not know which patient had which treatment when looking at events. The things they reviewed were:

  • Bad reactions to trial treatments
  • Getting much worse
  • Dropping out
  • Any medical change
  • Any reported disease or disorder

They then were told which patient had which treatment and used that to work out if these events were caused by the treatment.

A “serious deterioration in health” was recorded if:

  • The physical function decreased by 20.
  • The participant left after 8 weeks because they felt worse.
  • A very bad reaction to treatment.

Secondary outcomes were worked out from:

A scale recording the change in health from the original rating: “Clinical global impression scale rating”). This was self-reported by the patient. It was rated with 7 options. These 7 options were later converted into 3 options (negative change, not much change, positive change).

  • Disability based on a scale of work and social adjustment
  • Distance in metres walked over 6 minutes
  • Score for distorbed sleep
  • Score for hospital anxiety and depression
  • Number of CFS symptoms
  • Individual symptoms from the international criteria
    • post-exertional malaise
    • poor concentration
    • poor memory

Before treatment, the participants rated whether they thought the treatment would work. This was rated with 5 points but later made into 4.
After one year, participants rated how happy they were with treatment received. This was rated with 7 points but later made into 3 (satisfied, neutral, dissatisfied)

Back to contents

Statistical analysis:

[Much of this is difficult to make readable because its statistical analysis. Sorry. If you want to look into this I really recommend reading the paper. This should at least give you a rough idea of what they did]

At least 150 patients per group meant that enough patients could be compared for each type of therapy, with room for 10% of patients to drop out.

At first it was planned to measure improvement by meeting a certain score or achieving 50% change. This was later changed to continuous scores to help analysis.

20 patients missed one or two items on the rating scale. The average of their ratings was used for the missing value.

Continuous variables were summarised with the mean or median score. They converted the continuous variable into categories and used the frequencies and proportions.

The ratings from the participants were compared to the ratings from the independent analysers to see how similar the ratings were. Therapy received, relationship with therapist and sticking with the therapy were compared using Kruskal-Wallis tests (a test to see if the differences have any meaning or just happened randomly)

Categorical variables were compared using Fisher’s exact test (a test to see if the results have any meaning or just happened randomly).

A meaningful difference in result was calculated as:

  • 2 points difference in fatigue score
  • 8 points difference in physical functioning score

They analysed:

  • How many participants had improved by a meaningful difference.
  • How many participants scored in the “normal range” after a year. The normal range was:
  • 18 or less fatigue (worked out from adults who went to the GP: mean + 1 standard deviation)
  • 60 or more for physical function (worked out from working age adults: mean – 1 standard deviation)

[ Section still to be completed ]

Results

Back to contents

Participants

3158 patients were looked at to enter the trial. 641 were recruited. Reasons for not being recruited:

  • Not meeting Oxford criteria for CFS (most common)
  • Too high physical function
  • Currently in a CFS episode
  • Could not carry out the trial treatment
  • Trial treatment not suitable
  • Psychiatric issue
  • Fatigue score too low
  • Too young
  • Could not read or speak English well enough
  • Did not provide trial consent

The four groups had a similar mix of patients. The one exception was those just receiving medical care had had CFS for less long on average.

si1

Patient demographics

Treatment received

Those just receiving medical care received more sessions of medical care than other groups. All the groups had at least 85% of participants get enough sessions to be “adequate treatment”.

Patients had high expectations of APT and GET, lower of CBT and just medical care.

Most patients receiving therapy plus medical care were satisfied with their treatment.

Half of patients just receiving medical care were satisfied with their treatment.

Only one therapy session was not correctly identified by the independent assessors. But the therapy type was revealed in 40% of those assessed. The independent assessment was similar to patient reported results. All therapists assessed stuck well to the manuals.

si2

Treatment details

Primary outcomes

Adjusted scores showed:

  • Less fatigue and better physical function in CBT and GET.
  • No difference between adaptive pacing and just receiving medical care.

There was no effect from attending the same therapist. There was no difference in how effective the treatment seemed based on whether the participants met different criteria:

  • international criteria for CFS
  • London criteria for ME
  • depressive disorder.

si3

gr2.jpg

Improvements:

The percentage of patients who had improved fatigue scores by at least 2 points and improved physical function scores of at least 8:

  • Adaptive pacing: 42%
  • CBT: 59%
  • GET: 61%
  • Medical care only: 45%

The percentage of patients who were within normal ranges [as defined in this trial] of fatigue and physical functioning after a year:

  • Adaptive pacing: 16%
  • CBT: 30%
  • GET: 28%
  • Medical care only: 15%

Safety outcomes

Rates did not differ between treatment groups.

Non-serious bad events were common. CBT patients reported least.

Serious events, serious deterioration and serious bad reactions were uncommon. GET patients reported most.

si4

Safety outcomes

Secondary outcomes

Clinical global impression ratings

Patients self-reported their personal impression of their health. More patients from CBT and GET reported “much better” or “very much better”. Less than 9% of every group rated “much worse” or “very much worse”.

si5

Self-rated health

CBT and GET had better scores for:

  • work and social adjustment
  • sleep disturbance
  • depression
  • post-exertional malaise

Adaptive pacing had better scores for anxiety.

CBT had fewer CFS symptoms reported.

GET had better scores for how far patients could walk in 6 minutes.

Concentration and memory did not differ between groups.

Adaptive pacing groups and just receiving medical care caused the same results.

si6.gif

Secondary outcomes

Back to contents

Discussion

GET and CBT reduced fatigue and improved physical function the best. Adaptive pacing was no different to just medical care.

CBT and GET plus medical care are effective treatments for CFS. They have a moderate effect.

Adaptive Pacing therapy plus medical care is no more effective than just receiving medical care.

These findings apply to patients with CFS whose main symptom is fatigue.

CBT and GET were better than adaptive pacing in almost every measure.

GET and CBT reducing post-exertional malaise more than just medical care is important as patients wish to avoid this symptom.

Antidepressants were used across all groups so differences are unlikely to be from these.

CBT was best for depressed patients. GET was best for improving walking distance in 6 minutes.

There were no safety differences between groups. GET having more serious bad events was not considered related to treatment. All four treatments are safe.

There was more improvement from medical care than expected. There was less improvement from adaptive pacing than expected. Patient satisfaction was no different to other groups. The basic difference between pacing and CBT/GET is that it says to adapt to the illness. These results do not show pacing to be a good treatment for CFS.

Since the treatments are only moderately effective, more research into better treatments is needed. Behavioural treatments being effective does not mean the illness is psychological.

Strengths of trial:

  • low dropouts
  • patients stuck with treatment
  • clinicians were competent and followed the manuals
  • high patient satisfaction
  • good patient-therapist relationship
  • works for patients with CFS or ME who have fatigue as the main symptom

Limitations of trial:

  • Patients who could not attend hospital could not be part of the trial.
  • Results only apply to patients who were referred from their GP to other care.
  • The medical care provided in the trial to all groups is not the same as the usual care that patient’s get with a GP.
  • The most common reason for a patient not entering the trial was not actually having CFS
  • Medical care was not as closely monitored as the other therapies.
  • Those with just medical care received more medical care than those also receiving therapy.
  • Masking was not possible.
  • The ratings used to work out if the treatment had worked were self-rated by the participants and were subjective.

This summary is based entirely on the paper “Comparison of adaptive pacing therapy, cognitive behaviour therapy, graded exercise therapy, and specialist medical care for chronic fatigue syndrome (PACE): a randomised trial” by P.D. White et al, published in The Lancet 2011, accessed November 2015.

Tables and graphs are taken directly from the article.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s