Cognitive behavioural therapy has a pivotal role in the management of anxiety. Primary care should consider ways to deliver this intervention, says Dr Alan Cohen


Anxiety disorders are common. An Office of National Statistics (ONS) survey in 2000 found 164 cases per 1000 of neurotic disorder, which represents about 1 in 6 of all adults.

The survey found that the most prevalent neurotic disorder among the population as a whole was mixed anxiety and depressive disorder (88 cases per 1000). The authors report that this is a ‘catch-all’ category, in that it included individuals with significant neurotic psychopathology which could not be coded as any of the other five neurotic disorders.1

The second most common disorder found was generalised anxiety disorder (44 adults per 1000). The survey reported that the remaining disorders (depressive episode, phobias, obsessive compulsive disorder and panic) were less prevalent, ranging from 7 cases to 26 cases per 1000.1

These figures fail to represent the significant impact that anxiety disorders have on the sufferer and his or her family, and on primary care services. Some 30% of general practice consultations have a significant mental health component. Moreover, 90% of people with a mental health disorder are managed entirely in primary care.

Anxiety and depression are common and frequently co-exist in the same individual, they generate a high level of interest among healthcare professionals and the public, and it has been perceived (probably erroneously) that they are not well managed. Any assistance in managing these conditions is therefore welcome.

In December 2004, NICE published Clinical Guideline 22 on the management of generalised anxiety disorder and panic disorder.

At the same time, NICE published Clinical Guideline 23 on the management of depression, and the Medicines and Healthcare products Regulatory Agency (MHRA) released a review of the safety of antidepressant drug treatments. As well as recommending psychological treatments and setting standards of care, both guidelines are intended to help health professionals implement the MHRA’s advice.

Developing the guideline

The remit from NICE was to produce an evidence-based guideline for both primary and secondary care services that describes best practice in the management of generalised anxiety disorder and panic disorder in adults over the age of 18 years. In line with the NICE guideline development process, a guideline development group met at regular intervals for approximately 18 months to review the evidence and develop the recommendations.

The group, made up of primary and secondary care health professionals as well as patient representatives and carers, was managed by the National Collaborating Centre for Primary Care at the RCGP and supported by research scientists from Sheffield University. Two drafts of the guideline were produced and the comments of more than 100 stakeholders were incorporated.

Throughout this process, the guideline group was in close contact with the group developing the guideline on depression, reflecting the fact that individuals often suffer from both anxiety and depression. The wording of some sections of both guidelines is the same.

Both guideline development groups were aware of the MHRA’s review of antidepressant therapies, and it was essential that the advice offered in the three publications was consistent.

The recommendations

The group considered the recommendations by first describing the presentation and assessment process, then describing the treatment of the condition, and finally setting out the reasons or criteria for referral.

There are therefore three main groups of recommendations, which reflect the three steps in the care pathway of individuals with generalised anxiety disorder or panic disorder.

The grading of each recommendation was based on the standard NICE guideline scheme (Figure 1, below).

Figure 1: Grading scheme and hierarchy of evidence
Reproduced from Anxiety:management of anxiety (panic disorder, with or without agoraphobia, and generalised anxiety disorder) in adults in primary, secondary and community care by kind permission of NICE

The recommendations on general principles of care (Box 1, below) are very similar to those made by the depression guideline, because the management of these conditions is so closely related. The guideline includes an algorithm showing which guideline should be followed, according to the patient’s symptoms (Figure 2, below).

Box 1: Recommendations on general principles of care
Shared decision-making and information provision
  • Shared decision-making between the individual and healthcare professionals should take place during diagnosis and all phases of care (D)
  • To facilitate shared decision-making:
    • provide evidence-based information about treatments (D)
    • provide information on the nature, course and treatment of panic disorder or generalised anxiety disorder, including the use and likely side-effect profile of medication (D)
    • discuss concerns about taking medication, such as fears of addiction (D)
    • consider patient preference and experience and outcome of previous treatments (D)
    • offer information about self-help groups and support groups for patients, families and carers (D)
    • encourage participation in self-help and support groups (D)
  • Use everyday, jargon-free language, and explain any technical terms (D)
  • Where appropriate, provide written material in the language of the patient, and
    seek interpreters for people whose first language is not English (D)


Figure 2: Algorithm to help decide what guideline to follow
Reproduced from Anxiety: management of anxiety (panic disorder, with or without agoraphobia, and generalised anxiety disorder) in adults in primary, secondary and community care by kind permission of NICE

The recommendations represent good care, and it is interesting that there is little evidence to support what seems obvious – that providing the patient with the information to make an informed choice about interventions will result in a superior outcome.

Recognition and diagnosis

Unlike depression, for which there are simple screening tools, no specific questionnaires have been shown to be effective in primary care for the general screening of populations for anxiety disorders, although there are questionnaires that are useful in diagnosis and follow up.

The guideline makes clear that for patients assessed as having mixed anxiety and depression, the depression guideline should be followed. Interestingly, there is little research in this area, as mixed anxiety and depression is a common but ill-defined condition.

The clinical experience of both development groups was that individuals undergo episodes of anxiety and depression as part of a single illness. Treating the depression element was seen as most likely to produce lasting effects, as anxiety is frequently symptomatic of an underlying depressive disorder. The strength of this recommendation was unanimously agreed by both groups as ‘D’, because there is little or no evidence to support this statement (Box 2, below).

Box 2: Key recommendations and grade
Consultation skills
  • A high standard of consultation skills is needed so that a structured approach
    can be taken to the diagnosis and management plan (D)
  • Ask about relevant information such as personal history, any self-medication,
    and cultural or other individual characteristics that may be important
    considerations in subsequent care (D)
  • Be alert to comorbidity, which is common (particularly anxiety with depression and anxiety with substance abuse) (D)
  • Identify the main problem(s) through discussion with the patient (D)
  • Clarify the sequence of the problems to determine the priorities of the comorbidities – drawing up a timeline to show when different problems developed can help with this (D)
  • If the patient has depression or anxiety with depression, follow the NICE guideline on management of depression (D)
Presentation with a panic attack

If a patient presents with a panic attack, he or she should: (D)

  • Be asked if they are already receiving treatment for panic disorder
  • Undergo the minimum investigations necessary to exclude acute physical problems
  • Not usually be admitted to a medical or psychiatric bed
  • Be referred to primary care for subsequent care, even if assessment has been undertaken in A&E
  • Be given appropriate written information about panic attacks, and why they are being referred to primary care
  • Be offered appropriate written information about sources of support, including local and national voluntary and self-help groups

The most important recommendations in this section are that primary care is the most appropriate place to manage most patients with anxiety disorders and that all healthcare professionals involved in diagnosis and management should have a high standard of consultation skills.

The evidence base for these recommendations is weak; there are no meta-analyses or randomised controlled trials to demonstrate that primary care is better than secondary care, or that a clinician with excellent consultation skills is more effective than one with poor consultation skills. It was nevertheless an area on which the group members were unanimous.

Defining excellent consultation skills presented some difficulty, and it was felt that the criteria of the RCGP membership examination would be appropriate. The intention with this recommendation was to ensure that clinicians who are not doctors, who now have increasing contact with patients, would have the same consultation skills as doctors and that their training in this area should be equivalent to that of GPs.

The recommendation does not require every GP to have the MRCGP qualification, but is intended to ensure that every clinician who has first contact with an individual presenting with an anxiety disorder has excellent consultation skills.

Treatment options

For both generalised anxiety disorder and panic disorder three treatment options are recommended:

  • psychological therapy (cognitive behaviour therapy; CBT)
  • medication
  • self-help (bibliotherapy; using written material to help individuals understand their psychological problems and learn how to overcome them by changing their behaviour).

The group was unable to decide which was more effective, or more cost effective, as the evidence comparing different interventions was poor. (The group recommended that further comparative studies should be carried out.) However, the group did find that CBT was most effective in the long term, followed by medication and then bibliotherapy. There was no other evidence on which to base the selection of intervention for the individual other than patient choice.


Evidence shows that the use of benzodiazepines in patients with panic disorder produces a less good outcome in the long term, and they should be avoided.

The use of benzodiazepines in generalised anxiety disorder should not exceed 4 weeks (in line with the British National Formulary). SSRIs are effective in both panic disorder and generalised anxiety disorder. If alternatives are required, venlafaxine (at a maximum dose of 75 mg per day) may be considered for generalised anxiety disorder, providing that the MHRA prescribing advice is followed; clomipramine or imipramine may be considered as alternatives in panic disorder.

Assessing the evidence

The evidence reviewed for this section was extensive, but the development group encountered several confounding factors.

Many papers studied different forms of psychological interventions to identify the most effective. Cognitive behavioural therapy is easy to research because of its strict structure and controlled manner of implementation. Other forms of talking therapy are less easy to quantify, which makes it difficult to assess their effectiveness.

Other papers were published by eminent secondary care researchers who had used their outpatient referrals for detailed studies. As some of the later papers acknowledged, this population is relatively atypical because most cases are managed in primary care.

One of the research recommendations that the group made was to encourage more naturalistic or primary care research to assess how effective these interventions were outside secondary care.

There were many other questions that the group would have liked to answer for which there were no papers, for example which patients respond better to cognitive behavioural therapy than medication, and does the age, sex, or race of the patient indicate which intervention is more likely to be successful?

Referral to specialist services

Referral to a specialist mental health team is recommended when the patient has had any two interventions and is still significantly disabled by anxiety.

While few people have criticised the recommendations, many have pointed out that the cognitive behavioural therapy services are not routinely available. The value of the NICE guideline is that it provides the evidence for commissioners to change the way that they commission a service.

Already, many PCTs provide their own counselling services. It would be appropriate to consider discussing with counsellors how they will be implementing these guidelines, and what changes in service provision will be made.

The guidelines do not impose a timescale on implementation, and without doubt there will be a significant training issue. However, the evidence base for the recommendations relating to psychological therapy is strong (A or B). PCT commissioners should therefore be expected to consider how cognitive behavioural therapy can be delivered in primary care over the next few years, so that an effective evidence-based intervention can be delivered safely to all those who would benefit.


The guideline’s audit section attracted no comments during the wider consultation process. However, this section contains a draft quality and outcomes framework which PMS practices would find useful to assess their progress in implementing the guideline.

Further reading

Those who are interested in the source material on which these recommendations are based can download the full guideline, Anxiety: management of anxiety (panic disorder, with or without agoraphobia, and generalised anxiety disorder) in adults in primary, secondary and community care, from the NICE website ( Here you can find the evidence,the group’s opinion on each paper and information on how it has been used to develop recommendations.


  1. Singleton N, Bumpstead R, O’Brien M. Psychiatric morbidity among adults living in private households, 2000. London: Office for National Statistics, 2002

Guidelines in Practice, February 2005, Volume 8(2)
© 2005 MGP Ltd
further information | subscribe