NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Collaborating Centre for Primary Care (UK). Chronic Fatigue Syndrome/Myalgic Encephalomyelitis (or Encephalopathy): Diagnosis and Management of Chronic Fatigue Syndrome/Myalgic Encephalomyelitis (or Encephalopathy) in Adults and Children [Internet]. London: Royal College of General Practitioners (UK); 2007 Aug. (NICE Clinical Guidelines, No. 53.)

Cover of Chronic Fatigue Syndrome/Myalgic Encephalomyelitis (or Encephalopathy)

Chronic Fatigue Syndrome/Myalgic Encephalomyelitis (or Encephalopathy): Diagnosis and Management of Chronic Fatigue Syndrome/Myalgic Encephalomyelitis (or Encephalopathy) in Adults and Children [Internet].

Show details


2.1. Introduction

This chapter describes the methods used to generate the GDG’s recommendations for the diagnosis and management of CFS/ME in adults and children.

The methods were based on those of the National Institute for Health and Clinical Excellence (NICE) in: National Institute for Health and Clinical Excellence (April 2006) ‘The guidelines manual’. London: National Institute for Health and Clinical Excellence. Available from: The Guideline Development Process – an overview for stakeholders, the public and the NHS describes how organisations can become involved in the development of a guideline.

Consensus development methods were used in addition to the usual guideline development processes, and these are also detailed below.

2.2. The developers

The National Collaborating Centre for Primary Care (NCC-PC) is based at the Royal College of General Practitioners and has an academic partner, the Clinical Governance Research and Development Unit (CGRDU), based in the Department of Health Sciences at the University of Leicester. Its other partner organisations are the Royal Pharmaceutical Society of Great Britain and the Community Practitioners’ and Health Visitors’ Association. The Collaborating Centre was set up in 2000, to undertake commissions from NICE to develop clinical guidelines for the National Health Service (NHS) in England and Wales.

This guideline was developed by the NCC-PC and supported by an evidence review carried out by the University of York. The NCC-PC Methods team consisted of a project lead, project manager/researcher and health economist. The evidence review was commissioned from the Centre for Reviews and Dissemination at the University of York (see Appendix 1), and was an update review based on a previous systematic review on the diagnosis, treatment and management of CFS/ME in adults and children.5 The NCC-PC health economist undertook a review and subsequent modelling of the available health economics evidence, and details can be seen in the appropriate chapters.

2.3. The Guideline Development Group (GDG)

The Guideline Development Group (GDG) was deliberately convened to have a sufficiently large and broad membership to reflect the wider expertise amongst the various specialties to which people with CFS/ME may be referred. It chiefly comprised consumer representatives and healthcare professionals with daily, clinical experience of treating CFS/ME, rather than purely academic expertise. Nominations for GDG members were invited from various stakeholder organisations and members were selected to ensure appropriate representation. Nominations were also received for co-opted experts. Each nominee was expected to serve as an individual expert in their own right and not as a representative of their nominating organisation, although they were encouraged to keep their nominating organisation informed of the process. Co-optees contributed to aspects of the guideline development by attending up to two meetings at the invitation of the group members and reviewing papers as appropriate, but they were not full members of the GDG.

The GDG met on 18 occasions, at approximately 4–6-weekly intervals over 24 months, to review the evidence identified, to comment on its quality and completeness, and to develop recommendations for practice, based on the available evidence and using formal consensus techniques.

2.4. Developing key clinical questions

The first stage in the development of the guideline was to narrow the scope into a series of key clinical questions. The key clinical questions formed the basis for subsequent evidence reviews and helped the GDG to develop recommendations. The key clinical questions were developed by the GDG with assistance from the Methods team, including the team from York University who did the searching and reviewing. The project team then refined the questions into specific research questions to aid literature searching, appraisal and synthesis. The GDG reviewed, commented on and refined the protocol (see Appendix 1) that directed the searching and reviewing.

The following key clinical questions were addressed,

  • Question 1 (two parts): What are the existing case definitions for CFS/ME in adults and children? What evidence exists to substantiate or validate these case definitions?
  • Question 2: Are there any substantiated or validated evaluations to support the diagnosis of CFS/ME in adults and children?
    (Subquestion: In people presenting with early suspected CFS/ME (before 6 months) what are the risk factors/prognostic flags that might be linked with progression to CFS/ME?)
  • Question 3: Does the evidence show that any particular intervention or combination of interventions is effective in treatment, management or rehabilitation of adults and children with a diagnosis of CFS/ME?
    (Subquestion: In people presenting with early suspected CFS/ME what interventions might be effective in preventing progression to CFS/ME?)
  • Question 4: What are the information needs of healthcare professionals, patients and carers?
  • Question 5: What are the support needs of healthcare professionals, patients and carers?

2.5. Layout of chapters 4–7

For details of chapters and the layout, please see Structure of the guideline documentation.

2.6. Identifying the evidence

2.6.1. Literature search and evidence reviews

The aim of the literature search was to identify relevant, published evidence to answer the key clinical questions, in order to produce an evidence review using a systematic and transparent approach. One search was carried out to cover all five key clinical questions. The search was broad and aimed to pick up all studies of CFS/ME and related synonyms. Databases searched included Medline, EMBASE, Psych Info, CENTRAL, Social Science Citation Index, Science Citation Index, Index to Scientific and Technical Proceedings, PASCAL, Inside Conferences, AMED and HEED. Details of all literature searches are available in the systematic review (see Appendix 1). GDG members suggested further references. Evidence submitted by stakeholder organisations that was relevant to the key clinical questions and was of at least the same level of evidence as that identified by the literature searches was also included. Searches were conducted in May/June 2005, with update searches being carried out in August 2006. The extraction tables for the original search can be found at the end of Appendix 1 and for the update searches in Appendix 2.

Patient stakeholder organisations were invited to submit evidence on the ‘patient experience’ and the GDG reviewed and discussed the summaries of these (see Chapter 3 for details). This information was mainly from membership surveys. The use of patient surveys in guideline development is increasingly seen as important as such surveys allow a more complete picture to be established concerning the effectiveness of, and satisfaction with, given aspects of patient care (for example, a therapeutic intervention). However, information gathered through patient surveys is generally considered as relatively low-level evidence, for several reasons. The most important potential types of bias associated with patient surveys are the following:

  • selection bias: the systematic inclusion or exclusion of certain patients during selection to participate in a survey;
  • non-responder bias: systematic differences between participants who respond and those who do not respond to a given survey;
  • social desirability bias: a tendency to answer questions in a way that a given community/society may regard as expected; and
  • confounder bias: when a relationship found between two variables in a given survey (for example, patient satisfaction with a therapeutic intervention) does not in fact reflect reality but rather is disturbed by the effect of one or more other variables (confounders, e.g. provider performance).15

In addition, there are other potential biases more intrinsic to a given survey itself (such as potential biases in a survey’s ‘usability’, including its format, instructions and understandability; see 16).

Randomised controlled trials (RCTs) are considered to be at the top of the hierarchy of evidence,17 with patient surveys found further down the hierarchy. RCTs attempt to minimise many of the biases associated with patient surveys. For example, they may attempt to deal with the potential problem of non-responder bias (i.e. lost data from participants) through performing an intention-to-treat analysis on a dataset18 or address the problem of confounder bias through attempting to equally distribute both known and unknown determinants of a given outcome (and therefore potential sources of bias) between groups through the randomisation of participants.17 Thus, when evidence from an RCT is available to answer a given clinical question it is generally given priority over and above other types of evidence, including patient surveys.

The GDG found the information submitted by patient stakeholder organisations helpful in understanding the patient view. However, it also recognised that surveys from self-selected respondents are subject to bias and that this information was therefore not necessarily representative of the wider population of people with CFS/ME.

2.6.2. Response to criticisms of the evidence review

A large number of the criticisms of the evidence review appear to have been due to a misunderstanding about its nature and purpose. The evidence review was commissioned as an update of the Centre for Reviews and Dissemination’s (CRD’s) original systematic review13, and reviewed only RCTs and controlled trials of interventions for the treatment/management of CFS/ME. This review was then only one of the resources available to the GDG during the guideline development process (see the NICE Guidelines Manual 2006 for details as used in this guideline: Other resources, including patient evidence and clinical expertise, were also considered by the GDG.

The aim of the review was to identify all relevant RCTs or controlled trials: it did not exclude any RCTs or controlled trials on the basis of their age, country of origin or validity. The validity of the trials was consistently highlighted throughout the review, with a discussion of the methodological flaws.

Much of the existing evidence is of poor quality, and the review was restricted to those study designs at the top of the evidence hierarchy, i.e. RCTs and controlled trials. Where RCTs or controlled trials are available, widening the inclusion criteria to include poorer study designs would not improve the quality of the evidence, but would introduce the problem of comparing and weighting data from different study designs, making the evidence even more difficult to interpret.

As noted above, RCT or controlled trial evidence is not the only information considered when developing clinical guidelines, which is why the GDG also considered the experiences of both patients and clinicians. Please see the detailed description of how the recommendations were developed later in Chapter 2.

The purpose of the review was not to determine, or report evidence relating to the possible underlying processes of CFS/ME; the review therefore summarised the evidence from the trials of interventions and did not comment on any association with possible underlying disease processes. Details of all the included trials, including reported adverse events, can be seen in Appendix 1.

2.6.3. Health economics

The Methods team health economist, liaising with other team members as appropriate, reviewed the literature to assess the economic evidence. As such evidence was limited, a broad evidence search was performed, designed to identify information about the costs or resources used in providing a service or intervention and/or the benefits that could be attributed to it. No criteria for study design were imposed a priori. In this way, any evidence that might be of use was more likely to be identified. Thus, papers were not restricted to RCTs or formal economic evaluations, but papers included were limited to those written in English and containing health economics information that could be generalised to England and Wales. Extraction was then undertaken on any formal economic evaluation identified, and the results were presented to the GDG. The extractions can be found in Appendix 2.

2.7. Reviewing and grading the evidence

The titles and abstracts of records retrieved by the searches, provided by the GDG or submitted by stakeholders were scanned for relevance to the key clinical questions. Those relevant were reviewed to identify the most appropriate evidence to help answer the key clinical questions and to ensure that the recommendations would be based on the best available evidence. This process involved selection of relevant studies; assessment of study quality; synthesis of the results; and grading of the evidence. The methods used are outlined below.

2.7.1. Review of the clinical evidence

Details of the review methods can be seen in the full review in Appendix 1. Studies were graded according to Chapter 7.2 of the NICE Guidelines Manual 2006 (available at

2.7.2. Review of the health economics evidence

Health economics evidence was reviewed by the health economist in the Methods team. NICE methods were used (see the Guidelines Manual 2006 at, and details of the reviews can be found in each of the relevant clinical chapters.

2.8. Developing evidence statements and recommendations using formal consensus methods

2.8.1. Background

Where there is a good evidence base of well-conducted experimental studies, this forms the basis for the development of clinical guideline recommendations. The evidence is synthesised into evidence statements which are clearly linked to the recommendations of the GDG.

As there is little good research evidence for some aspects of CFS/ME care at present, formal consensus methods were used to assist the GDG in making recommendations. The consensus methodology approach adopted was one of the methods reviewed in the 1998 Health Technology Assessment publication: ‘Consensus development methods, and their use in clinical guideline development’.19

2.8.2. RAND/UCLA Appropriateness Method (RAM)

A modified version of the RAND/UCLA Appropriateness Method (RAM) was used ( The Methods team adapted it for this guideline in consultation with Professor Rosalind Raine, who has researched this method in the UK and advised the Methods team and the GDG on its use. The RAM uses a highly structured list of clinical indications, and consideration is restricted to the basic measurement of appropriateness,20 where the concept of appropriateness refers to the relative weight of the benefits and harms of a medical or surgical intervention.

2.9. Use of the consensus methodology in guideline development

2.9.1. Overview of the use of consensus methods

The GDG used formal consensus methods at different stages in guidance development where required by the strength of the available evidence. At each stage, the GDG members:

  1. rated the evidence statements privately and returned them to the Methods team for analysis and pooling
  2. received the pooled results and, to aid discussions, their individual ratings and their position compared to other raters
  3. discussed the statements on which there was not consensus. Wording was altered as necessary to improve clarity
  4. re-rated statements.

This process was omitted for areas of the guideline where the GDG considered there was greater certainty on best practice and recommendations in these areas were developed through informal consensus methods according to the standard NICE methodology.

2.9.2. Ratings and measure of agreement

For the GDG consensus evaluation, agreement was rated using a 9-point Likert scale (see below). If the rater did not have an understanding of the statement (for example, the rater genuinely did not know whether antidepressants were appropriate), they were instructed to tick ‘Don’t know’.

1 2 34 5 67 8 9‘Don’t know’

A researcher from Professor Raine’s team analysed the results to determine strength of consensus (indicated by the median) and the level of agreement within the group (indicated by the mean absolute deviation from the median – the MADM).21 The MADM measures variations about the median and does not give extra weight to extreme observations. This measure combines the group judgement for each item with the extent of agreement around each judgement. The operational definition of disagreement used was a new measure named the IPRAS (interpercentile range adjusted for symmetry).20 This represents a new approach to measuring disagreement which has now been tested in a variety of datasets. It is a continuous measure that can be applied to any size of panel and can be used to create either stricter or more relaxed definitions of agreement and disagreement. Another advantage in comparison to the classic definition is that it smoothes the rigid frontier between 3–4 and 6–7 (that is, between ‘disgree’ and ‘uncertain’, and between ‘uncertain’ and ‘agree’), and is a better measure of the degree of dispersion among ratings.

The overall group rating for a statement is categorised as AGREE (median rating > 6), DISAGREE (median rating < 4) or UNCERTAIN (median rating 4–6). A group rating UNCERTAIN* is used where the median rating indicates AGREE or DISAGREE but there is wide variation in the participants’ individual ratings (see Figure 1 below).

Figure 1. Explanation of results from wider survey questionnaire.

Figure 1

Explanation of results from wider survey questionnaire.

The results were reported in terms of the numbers and identities of the statements attracting strong, moderate and weak support, and those for which there was disagreement. Results were pooled and presented as an aggregate. Each rater’s results remained confidential at all times.

2.9.3. Stages of the process where formal consensus techniques were used

Evidence statements

The evidence statements were drafted and graded by the York team on the basis of the systematic review. The GDG members received the systematic review and evidence statements before a GDG meeting, rated, discussed and edited, if necessary during the meeting, and then re-rated the evidence statements as described in sections 2.11.1 and 2.11.2. The evidence statements with a positive consensus agreement were included in the guideline. Evidence statements on which there was a consensus of disagreement or about which the GDG was ‘uncertain’ were discarded.

Clinical scenarios and cues

Over several meetings, the GDG used the method described in Murphy and colleagues19 and Raine and colleagues22 to develop and refine a series of statements known as clinical scenarios. Clinical scenarios are statements of options for the use of interventions. They were developed on the basis of the evidence, current practice, and the expertise of the GDG and the expert co-optee advisors.

The GDG took this step in addition to the usual NICE process to ensure transparency and fairness in decision-making in areas about which there was uncertainty and possible disagreement. Scenarios were developed related to diagnosis, investigations and management of CFS/ME on which, in the view of the GDG, there was uncertainty. Scenarios were not developed for the information and support sections.

Cues are factors that may influence clinical decision-making. Murphy and colleagues19 describe them as ‘dimensions or indications that group members are asked to take into account when making their decisions’. For this guideline, the cues were determined by the classifications of symptom severity3 used in the CFS/ME Working Group report to the Chief Medical Officer4, by information submitted by patient organisations, and by input from the GDG and expert co-optee advisors.

The GDG agreed that that the two crucial factors (i.e. cues) influencing clinical decision-making were the severity of CFS/ME symptoms and the age of the patient (adult or child).

Development of the questionnaire

The methodology used by Raine and colleagues rates appropriateness of treatment for each combination of clinical scenario and cue. In the example below, the CUES are in capitals and the clinical scenarios in italics:

  • For a CHILD with MILD CFS/ME, treatment A is appropriate
  • For a CHILD with MODERATE CFS/ME, treatment A is appropriate
  • For a CHILD with SEVERE CFS/ME, treatment A is appropriate
  • For an ADULT with MILD CFS/ME, treatment A is appropriate
  • For an ADULT with MODERATE CFS/ME, treatment A is appropriate
  • For an ADULT with SEVERE CFS/ME, treatment A is appropriate

A questionnaire was developed on this basis, which the GDG rated and discussed. The first version of the questionnaire was very lengthy; because each clinical scenario had six possibilities, there were over 700 statements to rate. The full questionnaire is in Appendix 3.

The GDG changed some ambiguous statements. They re-rated these and the statements that were rated as uncertain. Details of results and changes to statements are given by topic in Chapters 5 and 6.

Questionnaire to a wider group

The methodology recommended by Raine and colleagues advised the use of a one-round modified Delphi process (postal) involving a wider group (rather than the GDG alone) to inform the GDG.

This wider group questionnaire contained all the statements (clinical scenarios) that the GDG had rated as ‘uncertain’ on the second rating round. In addition, a 20% random sample of questions from each section on which the GDG had reached consensus (either agree or disagree) was also included in order to give an indication of the range of the scenarios considered and the GDG’s responses to date.

The Methods team contacted all stakeholder organisations registered with NICE in May 2005 and asked them to nominate 5 to 50 people with knowledge or experience of CFS/ME to complete the questionnaire. In order to ensure a representative sample of healthcare professionals with experience of the condition, nominees from the CFS/ME Clinical Centres were also solicited. People who had been nominated to join the GDG, but had not been selected were also invited to participate. Patients and carers were nominated by a stakeholder organisation, who was asked to obtain their agreement before the Methods team contacted them directly. Therefore ethical approval through an Ethics Committee was not sought. The standard letters sent to participants are in Appendix 4.

Wider group participants had to agree:

  • to read the evidence review, including the evidence and evidence statements, and take part on this basis
  • that the GDG would have final authority on the content of the guideline
  • to complete the work and return the questionnaire with their ratings within the allotted deadlines.

Participants chose whether to receive the documents by post or email.

Participants were first sent the systematic review from York University (see Appendix 1) and the agreed evidence statements so that they understood the evidence base that the GDG had reviewed before completing the questionnaire.

Four weeks later, participants were sent the questionnaire (scenarios chosen for inclusion as above), and the GDG’s combined consensus rating on these scenarios.

Participants were given a further 4 weeks to complete the questionnaire and return it to the NCC-PC. Postal questionnaire responses were input by professional data entry services.

Questionnaires were sent to 399 participants. Of these, 219 completed the questionnaire (giving a response rate of 55%). The categories and numbers of respondents can be seen in the table below.

Healthcare professional (including professionals from the following groups - GPs, dietitians, immunologists, neurologists, nurses, occupational health physicians, occupational therapists, physiotherapists, psychiatrists, psychologists, infectious disease specialists)63
Not given8

The questionnaire was analysed as described in section 2.9.2.

The main output (see Figure 1) categorised participants as patients/carers or health professionals, and combined all participants in the ‘All’ category. The output shows the Guideline Development Group’s overall rating. The distribution and percentage of ratings (including ‘don’t knows’) are shown for each subgroup, along with the subgroup’s overall rating. The median rating for a subgroup is indicated by a box around the value.

In addition, the Methods team had median ratings broken down by separate healthcare profession categories if the GDG wanted a specialist rating on a particular question.

Examples follow of positive and negative comments received from wider survey participants. The first two are representative of many similar comments.

  • ‘Good luck, I posted mine back to you yesterday!’
  • ‘Many thanks for the ME/CFS material. I am doing my ‘homework’ and reading the attachments!’
  • ‘I’m really sorry, but I’ve been unable to do this piece of work. I’m currently having a lengthy course of chemotherapy, and its effects have been more debilitating than I anticipated. I wish you well with the rest of the process and look forward to seeing the finished guideline.’
  • ‘I truly believe that a lot of people without the condition would have a problem getting to grips with the information and questionnaire!!! I, for one will not be able to help you by returning the questionnaire.
    When I agreed to be sent the questionnaire I assumed it would be a simple task of answering questions, that would go some way to helping the medical profession reach a worthwhile conclusion. I did not think for one minute it would need over 450 pages of accompanying notes!!!’
  • ‘How I, or anyone else with M.E. or even recovered could possibly read, digest and understand the NICE document enough to be able to answer the Questionnaire, is beyond my comprehension.
    I surely cannot be the only person who has had this problem, or am I the only honest one around?
    I would like this letter to go on record as I feel it is very important for Non-Sufferers to know how difficult a task this was for an M.E. Patient. Just writing this letter has been hard enough!’
  • ‘I would like to say firstly how refreshing it is to receive such a rigorous, balanced and fair summary of existing research in the excellent (if bulky!) Evidence Review. The team at the University of York have done an extremely thoughtful and professional job and I would like them to know that their work is much appreciated. As I’m sure you’re well aware, ME/CFS sufferers have the added misfortune of having contracted a “controversial” illness, so it’s good to see an objective and thorough examination of existing research. Often in the past this research has been poorly analysed, with unreliable findings being “over-spun” in the National Press (for example research on NADH was described as a “cure” in several major newspapers in 1999 on the basis of one very small piece of not very well constructed research).
    The second point I would like to make is about the most severely affected ME/CFS sufferers. As a “moderate to severe” sufferer I found this questionnaire very challenging to complete in my current state (I should also say that I have some past experience of this kind of process as I worked in Market Research before becoming ill). I can understand why the wording has to be so complex, and why a Likert scale was used for example, but my concern is that once again within this process there is a very real danger that the voices of those most severely affected may not be heard loudly enough. This is particularly concerning, in my view, as much of the research evidence and most existing patient services already exclude those in greatest need by default (attending major hospitals as outpatients is virtually impossible for severely affected patients). I also know from personal experience that it is in treating the most severely affected that the NHS in general and GPs in particular face the greatest challenges. I sincerely hope that this concern is in some way addressed within this process.’

The GDG met again to review the results of the survey and discuss areas of difference with the wider group to aid the drafting of recommendations.

Overall, there was good agreement between the GDG and the wider group in most areas. The GDG discussed all statements rated as ‘uncertain’, recognising that, as with other surveys, the results were subject to bias. The full results of the wider survey are given in Appendix 3 with the ratings presented by topic in each relevant chapter.

2.10. Developing recommendations

The Methods team drafted recommendations based on (i) the clinical scenarios on which the GDG had a consensus of agreement both among themselves and with the wider survey, (ii) analysis of the GDG discussion, and (iii) the evidence from the systematic review. The items on which the GDG had consensus of disagreement were reviewed individually by the GDG to decide whether they required a negative recommendation or no recommendation. These were presented by the Methods team to the GDG as per the normal NICE guideline development process.

The GDG received the drafted recommendations before a GDG meeting and were asked to rate their agreement with them on a numerical scale following discussion. They were then reviewed a second time and re-rated. The GDG added some additional general recommendations where they identified gaps in the specific recommendations. A final rating of recommendations was done after revisions made in response to the stakeholder comments.

Recommendations with a positive consensus rating were included in the guideline.

See Appendix 3 for details of the recommendation development and the ratings.

2.11. External review

The guidance has been developed in accordance with the NICE guideline development process, using additional consensus development. This included allowing registered stakeholders the opportunity to comment on the scope and the draft guidance. In addition, the final draft was reviewed by an independent Guideline Review Panel (GRP) established by NICE.

The comments made by stakeholders, peer reviewers and the GRP were collated and presented for consideration by the GDG. All comments were considered systematically by the GDG and the Methods team recorded the agreed responses. Responses can be seen on the NICE website

Copyright © 2007, Royal College of General Practitioners.
Bookshelf ID: NBK53570


  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (787K)
  • Disable Glossary Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...