An Introduction to Cognitive Behaviour Therapy, 2 edition


The Course of Therapy

This chapter will present an overview of the course of therapy and the tasks and problems likely to arise at different stages.

Overall pattern of sessions

For most straightforward problems of the kind described in this book, a course of therapy typically takes from six to 15 one-hour sessions. However, there are no hard and fast rules either about the length of each session or about the number of sessions. For example, sessions may be shorter towards the end of treatment, when the client has become responsible for much of the treatment; on the other hand, if treatment demands lengthy in-session behavioural experiments, then those sessions may last considerably longer than 60 minutes. Similarly, the number of sessions may be extended if the problems are more complex, or shortened if the problem is highly amenable to treatment. Sessions are usually weekly to begin with, and become gradually more spaced as treatment progresses, with a couple of follow-up sessions after the end of formal treatment.

In the first two or three sessions, you will usually be focused on assessing the client’s problems, with the aim of deriving a formulation to share with him. Running parallel with this, you will be attempting to educate him about CBT and his expected role as an active, skilled collaborator in the therapeutic endeavour. Most of the active work on the target problems will be within sessions 2 to 12, and the final couple of sessions will be concerned with drawing up a blueprint for your client to take forward after discharge.

There are some features that appear throughout the course of treatment, and these include:


  • agenda-setting
  • self-monitoring
  • dealing with setbacks
  • updating the formulation.


The setting of a mutually agreed agenda at the beginning of each session is a key feature of CBT. As it is a relatively brief therapy, it is important to ensure that time is used effectively, and agenda-setting contributes to this goal by:


  • allowing you and your client to prioritise the issues to be addressed in any given session;
  • promoting the structure that is characteristic of CBT;
  • helping you both to maintain a focus on relevant problems;
  • helping to engage the client as an active participant in the therapeutic process.

To encourage the development of a collaborative relationship, it is helpful to address agenda-setting in the first session or two, by saying something like:

‘It is important that the treatment sessions seem relevant and helpful to you, and given that we only have a limited time in each session, we usually find it helpful to decide at the beginning of the session what we will aim to cover. I usually have ideas about what I would like to include, but you will often want to discuss things that have happened in the week, or thoughts that have occurred to you, and so on. It is really important that we make time for those, and so it would be helpful if you would take a few minutes before each session just thinking over what you would like to include. We can then agree an agenda between us. Does that sound sensible? Would you be willing to have a go at that?’

Follow this up by asking at the beginning of each session what the client wants to include on the agenda, and then, following that, suggest items that you want to include yourself (if you start with your own items, the client may be less likely to come up with his). This process may take five minutes at the start of each session – this is time well spent but you need to take account of it when working out how much time is available for other items.

The items you will usually include on the agenda are:


  • Brief review of the events of the past week.This does not need to be extensive, and should only briefly identify matters that are identified as major agenda items. Clients may be unfamiliar with such a brief review, and go into too much detail. In that case you can model what is helpful by gently interrupting and summarising the main points, for example:

‘So what you seem to be saying is that for most of the week, your anxiety level was rather higher than it has been, and the major factor seems to have been your father’s wedding plans. However, you have managed to go to work every day, and that has felt positive. You only need to give me an overall outline at this point, but have I got the general picture correct? Would it be helpful to put the wedding on our agenda?’


  • Review of the last session.This may include any problems with what was discussed, any expansion on points made, etc. Many therapists give clients a recording of the session to listen to as part of their homework, and new perspectives may have been gained as a result of that. If you have asked the client to keep notes about the session in a therapy notebook, he can also review that over the subsequent week. This process may raise issues that could be included on the agenda. For example, look out for whether the client seems to be giving honest feedback, or saying what he thinks you would like to hear. In the latter case, you need to decide whether drawing attention to it at this point would weaken or strengthen the therapeutic relationship. If your client cannot remember what went on in the previous session, this should also be tackled as a problem, and ways of overcoming it need to be identified.
  • Assessment of current mood.This can be formally assessed using a standardised measure such as a BDI (Beck, Ward, Mendelson, Mock & Erbaugh, 1961) or BAI (Beck, Epstein, Brown & Steer, 1988), or more informally via questioning: has the client’s mood changed since last session? Does any aspect of it need to be included as an agenda item? It may be helpful to ask the client to rate his mood, using the same scale from week to week. For example, ‘How would you say your mood has been over the past week from 0 to 10, where 0 is as low as it could be, and 10 is absolutely fine, no problems at all?’ or for an anxious person, it could make more sense to have a scale where 0 is no problem at all, and 10 is as bad as you can imagine, highly anxious all of the time (see Chapter 5 for more discussion of how to devise measures). Make sure that you write down in your notes what scale you are using!
  • Review of homework, which may heavily overlap with the major topics for the day (while this is often referred to as homework, some clients have bad associations with that word, for example from school experiences. Alternative terms include assignment, behavioural experiment, tasks for next week, project, or a description of the specific task, like ‘survey’).
  • The session’s main topics for discussion.These may include symptoms (such as low mood, anxiety, sleeplessness), or current external problems (such as issues at work or relationship difficulties). You probably have plans to work on particular CBT skills, such as learning how to identify NATs, or the role of safety behaviours in the maintenance of a problem, and it is often possible to cover these at the same time as the client’s symptoms or problems are being addressed.
  • Homework assignment.This should arise from the main topics discussed, and may already have been negotiated as part of that. However, it is worth bearing in mind that setting up homework can take 10 minutes.
  • Feedbackon how the client has experienced the session. You need to make sure you have left time on the agenda for this. For example, you could say:

‘It would be very helpful if you would give me some feedback on how things have gone today. It may be difficult at first to tell me if things have been disappointing or if I have said something that has upset you, but as we try to work together on dealing with your problems, it is really important that you feel able to say whether things are helpful or not. What would you say are the take-home messages from today? … Is there anything else that has been helpful? … Is there anything I have said that is going to play on your mind, or has been unhelpful? … Any other comments on today?’

It will be apparent that, allowing for agenda-setting, agreeing homework and getting feedback, there is not much more than 35–40 minutes for the major topics for the day. This means that usually no more than two topics can be included, unless it is planned to allocate five minutes or so to an additional one.

In order to decide which issues to prioritise during agenda-setting, the following factors can be considered:


  • Issues of risk to the client or others, including children
  • Urgent problems, e.g. possible job loss, imminent exams
  • Level of distress
  • Centrality to the formulation
  • Potential for change
  • Relevance to a skill that needs to be learned
  • Whether the problem could be tackled with someone else outside therapy.

In the early stages, it is not usually helpful to tackle highly distressing and complex problems, as it is unlikely that the client yet has the skills to deal with them effectively. Similarly, issues directly related to rigidly held or core beliefs should be avoided. This should, however, be balanced against the importance of dealing with issues seen as important by the client.

Once the agenda is set, you should aim to follow it, and be explicit about any deviations from it. For example, if the client moves to a different topic, and perhaps becomes upset about it, you should not assume that the client would choose to prioritise the new topic. Instead, discuss the dilemma, by saying, for example:

‘This seems to be very upsetting for you, which makes me think it might be an important issue. Would you like us to spend some of our time thinking about this, or would you rather that we focus instead on …, as we agreed at the beginning of the session?’

This allows the client to make a choice, sometimes with surprising results. Similarly, if discussion brings up a topic related to risk, then you probably need to prioritise it over other items on the agenda, but again, be explicit with your client that you intend to move away from the agenda.

The handling of the agenda needs to be managed sensitively, with respect and understanding of the client’s position. The client may sometimes wish simply to ventilate feelings about a difficult situation, possibly without any expectation of problem resolution. This can be an entirely reasonable target for one or even two sessions, although it would probably need further exploration if the client wanted to take up a good proportion of each session in this manner.

In order to remain with the agenda, it is helpful if you either make, or request from the client, frequent summaries of the major points related to a topic or problem. A summary should cover the main points of a discussion in one or two sentences, and include, for example, important negative automatic thoughts, in the client’s own words. This helps therapist and client to remain on the same wavelength, and it also serves as a useful break between topics on the agenda. In the first five or six sessions, it is helpful to summarise about every 10 minutes, and to ask the client whether he has understood accurately. For example:

‘You seem to be saying … and … Have I got that right? Have I missed anything?’
‘Could you put in your own words what you see as the main points of what we were discussing?’

Common difficulties in agenda-setting

Some common difficulties you might encounter in setting an agenda for the session are:


  • Setting a vague agenda, where the topic is only described in broad outline rather than being operationalised in detail. For example, if the client says that he wants to talk about his relationship with his family, you need to ask him to identify what aspects of the relationship he wants to discuss; or if a client wants to talk about her weight, you need to establish what aspect of her weight is concerning her, and to clarify if ‘weight’ is the key issue or if it is something that contributes to it, such as binge-eating.
  • Putting too many items on the agenda. You will not usually manage more than two major items, and you and your client can be left feeling frustrated if you have to carry over items to the next session. If you and your client prioritise thoughtfully you will easily decide which are the one or two items you need to focus on.
  • Not prioritising which items to consider. Tackle the major ones first, using the factors described above to guide your prioritising.
  • Beginning to deal with issues as soon as they are mentioned rather than attempting to complete the agenda setting. Many clients need practice in setting agendas, and may launch immediately into a detailed discussion of the first item mentioned. Gently interrupt and remind them that it is important to agree explicitly about which issues to address, for example, you might say,

‘It looks as though this is a major topic for today. Can we just decide what else we want to include for this session, to make sure that we allocate enough time for everything?’


  • Not having genuine input from the client. This runs the risk that he will introduce his preferred items ‘off-agenda’ later in the session, or not at all.
  • Misunderstanding the meaning of an issue for a client. Continue to ask questions, use summaries to clarify meaning and make sure you have got it right, and ask your client for feedback.
  • Tackling issues not on the agenda without discussing it first. The agenda can be flexible, but changes need to be overt and collaborative.
  • Skipping from one topic to another from session to session, without achieving closure on any of them. Make sure that there is a broad strategy that will make progress across sessions.

It is helpful if you regularly review agenda-setting so that difficulties like these can be identified and dealt with. Although it may initially feel uncomfortable as therapist, especially if you are used to a less structured approach, it is worth experimenting to test out whether the consequences that you fear actually occur.

We will now go on to look at the stages in a course of therapy, beginning with the features of the early stages.

The early stages

Goal setting

Another aspect of CBT that serves to maintain its efficiency as a time-limited therapy is the agreement to work towards mutually agreed goals. This helps to structure therapy sessions, and to maintain their focus. Goals are established as a joint effort, and this further emphasises the collaborative nature of CBT: the goals for therapy are those that are relevant for the client, with input from the therapist.

Goal-setting implies the possibility of change, and this helps to engender hope and reduce the client’s helplessness in the face of what may seem insurmountable problems. It also raises the prospect of an end to treatment, and so helps you negotiate in an open and explicit way when discharge may be approaching.

How to set goals?

Goals should be ‘SMART’, i.e. they should be:


  • Specific
  • Measurable
  • Achievable
  • Realistic
  • and have a Time frame (i.e. a date for completion).

Setting out goals in specific detail can help clients feel more in control, because a global problem reduced to its component parts may feel more manageable. You can begin with general questions like:

‘How would you like things to be at the end of treatment?’
‘How would you know if treatment had been successful?’
‘At the end of treatment, what would you like to be different?’

For example, the first discussion with a woman who was feeling that her life was controlled by her health worries went as follows:


How would you know if treatment had been successful? What would be different?


I would stop checking myself for lumps … I wouldn’t be thinking about cancer all the time, and boring the family with it.


Is there anything else that would change?


I suppose the main thing is I wouldn’t get panicky every time cancer was mentioned.

This client’s response demonstrates a common problem: she described how she would like not to be, rather than how she would like to be. This has been called the ‘dead man’s solution’, i.e. the goals could be achieved by a dead man – no more panicky feelings, no more checking lumps, no more talking to relatives about cancer. Ask your client to describe how he wants to be or what he wants to do, rather than where he wants to move from.

The so-called ‘miracle question’ is sometimes a good way of getting at this:

‘Suppose whilst you are asleep tonight a miracle happens, and all your problems disappear, just like that. But you don’t know that it’s happened, because you are asleep. When you get up the next morning and go through your day, how will you come to realise that the miracle has happened? What would you notice was different about you or about other people? What would others see that would tell them that the miracle had happened?’

For the woman above, the goals eventually agreed were that she would carry out a breast check on a monthly basis; that she would discuss topics other than symptoms with her husband 95% of the time; that she would visit relatives in hospital if they were admitted; and that she would respond calmly if she developed symptoms. In order to measure her progress towards these goals, she was asked questions like, ‘Can we break this down into smaller steps?’ or ‘What would be the first sign that you were making progress?

Part of your role is to make sure that goals are realistic. Clients may have unrealistically extreme goals, such as a socially anxious person who wanted to find a life partner by the end of therapy; or the goals may be too limited, such as a client with obsessional disorder who wants to reduce his hand-washing to four hours a day. Occasionally it may be difficult for a client and therapist to agree on goals. For example, a client with an eating disorder may want help to lose weight; or a client with an obsessive-compulsive disorder may want help to make his rituals more thorough. In these cases delicate negotiation is required, but the process allows you and the client to be explicit about what he can and cannot hope to achieve through therapy.

For example, a woman with obsessional concerns about official forms identified that as one goal she wanted to learn how to be certain that she had filled in her tax form correctly. The session continued:


That’s interesting. You say that you would like to know how to be certain that you had made no errors in your tax form, so that your anxiety would fade. That sounds to me very much to be coming from the perspective of someone who still has an obsessional problem. Can you put yourself in the shoes of someone who no longer has OCD. What do you think things would be like then?


Perhaps I could ask the accountant to check it with me?


And if you had no worries about form filling, what effect would that have on your needing to check?


I suppose I could be aiming not to be doing any checking at all, just filling in the form.


Does that sound a reasonable goal to have? To fill in forms without checking them?


I think everyone probably checks things like that once.


OK. I guess that may be true. So what do you think would be a reasonable goal?


I think I should aim at being OK if I fill the form in, and then only check it once, with-out going over it in my head or anything.

During treatment, it may be necessary for her to submit a form without any checking at all, but her goal for successful treatment was in this case less rigorous but more realistic.

It is also important that the goals are achievable and involve change in things that are within the client’s control: in particular, he should focus on changing things about himself, rather than other people. For example, it may be reasonable to have job-seeking as a goal, but obtaining a particular job is not ultimately determined by the client, and therefore may not be an achievable goal. It is also worth considering whether the person has the resources – finance, skills, persistence, time – to achieve the goals.

When considering the time frame, the issue of which goals to tackle first can be approached by considering similar factors to those relevant for prioritising topics in a session. It is helpful to tackle initially a goal where rapid change is possible, in order to increase the client’s hope. Other factors to take into account include risk or urgency, importance to or level of distress for the client, and whether any particular goal logically needs to be approached before other ones can be tackled (for example, you would need to be able to travel to an interview before you could apply for a post, and so you would need to tackle anxiety about travelling before you tackled anxiety about interviews). For the therapist, other considerations are the centrality of a goal in the formulation, and the ethical acceptability of the goal (for example, a man wanted to reduce the distress he felt about intrusive images about beating his wife; he was encouraged to consider anger management instead).


There is clear evidence (Kazantzis, Deane & Ronan, 2002; Schmidt & Woolaway-Bickel, 2000) that clients who complete homework tasks show greater improvement than those who do not, and this is presumably partly because they have more opportunity to generalise what they have learned from sessions into everyday life. Most problems are based outside the clinic rather than in it, and the client can use homework to collect information, test out new patterns of thinking and behaving, and learn through direct experience. As the general style of CBT involves handing skills over to clients, it is important that they have opportunities to practise the skills in real life, whether this involves identifying negative automatic thoughts, working out how to reduce safety behaviours, or how to increase assertiveness in particular situations.

Because between-session assignments are so central to CBT, it follows that time must be allocated to setting them up, which may need five or 10 minutes at the end of a treatment session. However, homework will often follow on directly from the major topics on the agenda, and will have been devised earlier in the session as part of that discussion. For example, if the agenda has been concerned with the role of negative thoughts in triggering anxious feelings, an obvious homework task might be for the client to begin monitoring triggers and thoughts associated with anxiety in the following week. If there has been a discussion about the role of internal self-focus in a client with social anxiety, then homework may involve monitoring, or experimenting with an external focus and recording the effects on anxiety.

The range of possible homework is boundless, and relies on the ingenuity of you and your client in setting up suitable assignments. It can include reading relevant material; listening to treatment tapes; self-monitoring of feelings, thoughts or behaviours; carrying out behavioural experiments; practising new skills such as the use of thought records or assertive responses; doing a historical review of past experience; or activity scheduling. It is important that it makes sense to the client, and that it will be useful either for the subsequent treatment session, or for the achievement of a particular goal. For example, the elimination of a safety behaviour may feed directly into the next session, where the results may flesh out the formulation, and then lead on to the next manoeuvre. On the other hand, a client may be keeping a positive data log aimed at low self-esteem as a long-term assignment; discussion of it from session to session may be minimal, unless it was identified as a major topic for the agenda.

Clients often do not do their homework, and this can be for a number of reasons. The principles below will help to ensure that homework is carried out and is useful:


  • The homework should follow logicallyfrom what happened during the session. In fact it can be devised during the session if it fits well there – you do not have to wait until the end of the session to discuss assignments.
  • The assignment should be relevant, and be seen to be relevant, by the client. Check this out with questions such as, ‘Does this make sense? Can you summarise for me how you think this would be helpful?’ This is less of an issue later in treatment as the client takes an increasing role in setting up between-session assignments, but may be more problematic earlier on, when the therapist is likely to take the lead in suggesting suitable homework tasks.
  • It is also worth bearing in mind that your client has a life outside of therapy. Although it is important that they prioritise treatment, there are limits to what they can be expected to do, and they will be less likely to complete the homework if they feel overburdened. Check this out with them.
  • Homework should be planned in detail, spelling out what is to be done, when, where, with whom, etc. Pitfalls and difficulties must be identified and discussed, by asking carefully what could prevent the task not being carried through.

A woman was very concerned about being what she described as ‘a doormat’ in relation to her mother and sister. However, after a role play in the session, when an assertiveness task was being set up for a subsequent interaction, she mentioned that she was not likely to see either of the relatives in the next month or so! As a result, an alternative more immediate assertiveness task was planned, and a homework opportunity was not missed.

Difficulties may range from embarrassment at using a self-monitoring form at work, through to not having any money to carry out a behavioural experiment in a social situation. Be mindful of any underlying beliefs that may interfere with completing homework assignments. For example, a client with perfectionist beliefs may find an activity schedule difficult to complete because he might think that none of his activities was challenging enough to include; a client with low self-esteem may find it difficult to do any task where the outcome could be construed as falling short of the therapist’s ‘wishes’. In the early stages of treatment, anticipated problems should be dealt with in the here-and-now, rather than attempting to modify such underlying beliefs.


  • Make sure that homework cannot be ‘failed’, but rather seen as a source of helpful information, whatever the outcome. For example, if a client is attempting to reduce avoidance of particular situations, set up homework so that he can collect useful information about anxious thoughts and feelings even if he cannot reduce the avoidance.
  • Provide relevant resources, such as diary form or reading material, at least early in therapy.
  • The agreed assignments should be written down, by both you and the client. Although it may be quicker for you to write it down forthe client, it is helpful to establish his active role in therapy and this kind of involvement can be an early step in that direction.
  • Homework review should always be included in the agenda of the subsequent session. Partly this is because it should have been designed to be relevant for the session, but at a more general level, a client is highly unlikely to persist with homework assignments that you never follow up.

If the homework has been completed or nearly completed, then it should be reviewed in detail. For example, if the client has read a chapter of a book, what was helpful? What rang bells for him? Were there any sections that were difficult to understand? If he has completed an activity schedule, what was the pattern of pleasure and achievements? What did he learn? How can this be taken forward?

On the other hand, if the homework was not completed it is important that this is explored and the reasons determined. There may have been practical reasons (someone off sick at work so workload unexpectedly increased); the client may have forgotten; it may not have been discussed in sufficient detail, or not written down; it may be that the task was too difficult in some way. In all these cases, the task can be modified for a subsequent assignment, or perhaps carried out with assistance from you or someone else.

If underlying beliefs have interfered with the completion of the task then, as described above, this should be tackled pragmatically, at least early in treatment, rather than attempting premature belief change. For example, if it seems that the client has beliefs about control or autonomy that have been activated by a particular assignment, then the task could be modified so as to give him more control. This may not necessarily be spelled out unless the beliefs have been discussed in detail in the formulation, or unless the treatment has progressed to the point where such beliefs are the current focus.

For example, a man did not do any homework on two consecutive weeks. When this was looked at in more detail, he raised concerns about the relevance of the tasks, although he had not mentioned this when the homework was agreed. The therapist wondered whether autonomy may be an issue for him, but did not raise it at this early stage in therapy, particularly as it did not seem to be related to the problems he had presented. Instead, it was agreed that the client would play a bigger role in setting up homework tasks. This meant that often homework tasks were weightier than the therapist would have suggested, but by and large were completed.

The general point is that it is important to establish from the outset that homework is an integral part of therapy, and that it is difficult to proceed without the information and feedback that it provides. This is particularly true when the amount of treatment available is limited by resource constraints. Well-devised homework can mean that very limited treatment can result in enormous changes for the client, as the majority of the work is done outside sessions.

Problems in the early stages

‘Low motivation for change’

At the beginning of treatment it may appear that the client is not engaged with treatment, but it is helpful to understand the client’s apparent reluctance to engage, rather than use a trait description like ‘poor motivation’. This means that attempts should be made to analyse the problem in terms of thoughts, feelings, and behaviours, so that ideas about management can follow on. The following possibilities should be considered:


  • Ambivalence about change. Prochaska and DiClemente (1986) defined a spectrum of stages in a person’s preparedness to change: pre-contemplation (no intention of changing, possibly no awareness of a problem), contemplation (aware of a problem and considering change), preparation (beginning to make changes), action (successful cognitive and behavioural change) and maintenance (working to prevent relapse). It is worth considering which of these stages best describes your client. This needs to be kept under review, as motivation can shift as therapy proceeds: it can increase as your client accrues more successful experiences; or it can diminish as he finds that he has to work harder for results, or if he encounters difficulties in treatment – for example, if anxiety does not diminish as expected. A lack of self-efficacy can also be important, insofar as it affects what the client believes he can achieve; ways of dealing with this is discussed in Chapter 6.
  • Inaccurate expectations about the nature of the treatment approach.This is particularly likely to be a problem with clients who have had experience of different psychological treatment approaches, or who have a lack of faith in the therapeutic approach.
  • Lack of understanding or acceptance of the formulation.In our view, if a client and therapist have not agreed on a formulation or the cognitive-behavioural approach by around session 4 it is less likely that treatment will be effective. It is therefore worth spending time on clarifying the formulation, asking for feedback, listening effectively to the client’s concerns and trying to take account of them. If the client and therapist cannot come to a shared understanding, however simple the formulation, it is probably not worth pursuing treatment.
  • Hopelessness.Apart from the hopelessness that is common in depressed clients, hopelessness may also occur in clients who have had a history of unsuccessful psychological treatments. This can be approached using standard CBT techniques, including identifying and evaluating negative automatic thoughts and behavioural experiments focused on hopelessness.

‘I don’t have any thoughts’

It is difficult for clients to make sense of a formulation that focuses on cognitions if they are unaware of thoughts. It can be helpful for them to practise looking for thoughts, to look for images, or to try to identify what situations mean to them, even if they cannot easily identify automatic thoughts. Chapter 8 discusses ways of dealing with this problem, but as cognitions are central to this approach, it is important to deal with the issue rather than try to circumvent it.

Role of health beliefs

Clients may have an understanding of their problems that is different from a cognitive behavioural view, and it is helpful to find a way of working with this as an experiment, without attacking the alternative explanation. For example, clients with somatic symptoms often construe them in terms of physical illness. It may be useful to try to negotiate with such a client to try an alternative approach based on a CBT formulation, as an experiment for a specified time to see whether it works any better than the physical illness formulation (the so-called ‘Theory A/Theory B’ approach). Similarly, some clients with obsessional worries may explain their intrusions in terms of a religious framework, and a similar approach may be helpful. On the other hand, some clients may have differing beliefs about the roles and responsibilities of therapists and clients (for example, ‘It is your job to cure me’). Awareness of this can help the therapist devise assignments which could, for example, draw attention to the important contribution that the client can make to the process of change. A useful metaphor in this case is that of a road map – that the therapist’s knowledge can put you on the right page of the map, but that you need the detailed information which only the client has to direct you along the right roads on that map. You then need to set up experiments that demonstrate that this approach can be helpful.

Balance of pros and cons

We need to remember that therapy has costs as well as benefits for clients: costs of emotional strain, investing time and possibly money, and implications for other changes in the client’s life. Sometimes it is necessary to help clients think through the balance of costs and benefits. For example, reluctance to do a particular homework task should not be taken immediately as evidence of an unwillingness to change, but rather as indicating that the case for the task has not been made. We are often asking clients to make changes that require a great deal of bravery, and they are only likely to do that if, on balance, they see that the possible benefits outweigh the probable costs.

A client with a severe vomit phobia was finding it very difficult to give up safety behaviours (such as carrying mints in her bag, driving with her car window open, carrying a moist flannel in her bag, and sleeping with the light on so that she could find her way to the bathroom if necessary), even though she understood the rationale for doing so. Her cost–benefit analysis for dropping the safety behaviours differentiating short-term from long-term effects is shown in Figure 11.1.

Figure 11.1    Costs and benefits to clients of giving up safety behaviours

This analysis, set out in this way, allowed her to put her short-term fears in context, and to begin to drop her safety behaviours. Although this will often allow clients to move forward, you should also remember that occasionally clients may decide that, on balance, the costs of therapy outweigh the benefits for them right now, and therefore not continue in treatment.

Relapse management

You might wonder why this section is called relapse management rather than prevention, and why it comes early in the course of treatment. It is placed here because it is unusual for any client to progress smoothly without any hiccoughs, so it is worth preparing clients for setbacks from early on, perhaps using a metaphor based on acquiring other skills such as driving or typing. Relapse management is discussed in detail in Chapter 6.

Review points

Because CBT is time-limited, focused and structured, you need to carry out regular reviews throughout treatment. This helps to retain the focus of treatment, and to establish whether progress is sufficient to warrant continuing with treatment, or whether changes in the approach are required. The review should be related to the goals agreed at the beginning of treatment, and it is helpful if intermediate targets have been identified, as well as end-point goals. Any other measures that are being used, such as questionnaires or other self-monitoring, are also helpful for reviews.

It is helpful to agree at the outset that you will review progress after four or five sessions, in order to assess whether CBT is likely to be helpful. Although a decision not to continue with CBT may be dispiriting for someone who had high hopes of its efficacy, it is easier to deal with this at an early stage than after 20 sessions that have resulted in little change. After this initial review, further reviews should be carried out at five- or 10-session intervals.

The formulation developed in the first one or two sessions is tentative, so it is important to review it regularly to take account of any new information that becomes available as therapy progresses. This may be derived from homework assignments, behavioural experiments carried out in sessions and so on. Although the basic outline of the formulation may not change, the details of maintaining cycles are likely to be fleshed out during treatment, with implications for what interventions are likely to be helpful.

A man with agoraphobia was unclear about the content of his catastrophic thinking because he had for so long avoided situations that would trigger such thoughts. Once it was established that he had thoughts about not being helped by other people, this could be built into the formulation, and experiments set up to test them out.

It is especially important to review progress if little change is being made, or if an impasse has been reached. This may mean that the formulation is not helpful or has significant omissions. It is also worth looking at the therapeutic relationship to see whether there are problems interfering with the application of the formulation to the client’s problems. Such problems may include your own blind spots, which should be discussed with your supervisor. If no solution can be found, it may be decided that treatment should be discontinued at this point.

Later stages

As treatment progresses, the focus moves increasingly onto intervention rather than assessment, but the results of any intervention should always be related to the initial formulation to see whether it needs modifying. The client becomes increasingly independent in determining what items go onto the agenda, how long is spent on each item, and what homework is taken away; and as more CBT skills are learnt the client takes the lead in, for example, evaluating negative thoughts and in devising behavioural experiments to test out new perspectives.

You will probably spend most of the time in treatment sessions on the details of thoughts, feelings and behaviour in current situations, but as the end of treatment approaches you may spend some time on identifying and evaluating unhelpful assumptions or core beliefs, particularly if you think that the client may be at risk of relapse if such beliefs are not modified. It is, however, not always necessary to modify underlying beliefs directly. If the client has worked successfully at re-evaluating negative automatic thoughts, both in sessions and in vivo, then very often the re-evaluation ‘leaks upwards’ to the more general beliefs, particularly to the level of dysfunctional assumptions.

For example, a client had strong beliefs about never expressing anger. In a range of situations, she experimented with being more assertive, including in situations where people were behaving unreasonably. Her beliefs about expressing anger were modified, although they were not directly addressed.

The emphasis on the portability of skills means that it is important for the client to reflect on what is taking place in therapy, so it is helpful to ask questions such as ‘What were we doing there?’ ‘Can you identify the kind of crooked thinking you were showing there?’ ‘How could you use that in other situations?’ It is important that you attribute progress to the client’s efforts, particularly if he is dependent and hence likely to attribute change to your attention and skill rather than his own efforts. This is discussed in more detail in Chapter 6.

As treatment progresses, the frequency of sessions may be reduced, perhaps moving to two-week gaps between two or three sessions, followed by perhaps a three- or four-week break before treatment ends.

Ending therapy

It is relatively easy to work towards ending therapy if the treatment goals were well defined, and if there has been good progress towards them. Similarly, you can keep in mind the idea that treatment will be coming to an end through regular reviews of goals and progress, as this emphasises the short-term nature of the treatment process. The client should gradually have acquired confidence in his ability to apply a CBT approach to his problems, using the skills learned during therapy.

As the end of treatment approaches, it is helpful to further develop with your client a blueprint for dealing with any problems that may emerge in the future, based on the relapse management work you have done together (see Chapter 6). This could include:


  • What has been learned during therapy
  • What strategies have been most helpful
  • What situations in the future may be difficult, or possibly lead to a recurrence of the problem
  • Ways of responding to this, given what has been learned in treatment
  • How to handle a significant problem, including if necessary a telephone review with the therapist.

The emphasis should be on the idea that the client is equipped to deal with most problems that are likely to arise, even if, in some circumstances, it would be reasonable for him to ask for your help.

Rather than having an abrupt end to treatment, it may make sense to plan a booster session or two over the subsequent year. You can then review progress, reinforce the client’s success at dealing with problems, check out how he dealt with problems that had been anticipated at previous sessions, check on the re-emergence of unhelpful patterns of thinking or behaviour (for example safety behaviours), and work together to trouble-shoot if necessary.

Despite the gradual withdrawal from therapy, and the emphasis on skills acquisition, some clients continue to worry that they will not cope by themselves after the end of treatment. This can be approached in a standard cognitive behavioural way, by identifying worrying thoughts, and helping the client to deal with them. This could include setting up behavioural experiments to test out alternative perspectives. If a client has general beliefs about not being able to cope alone, the booster sessions over the year following the end of formal treatment can be used to test out the beliefs, perhaps via a positive data log.

A 59-year-old client who had been depressed as a result of a number of events, including his increasing difficulty with a rapidly changing job, had responded well to treatment, and had maintained his progress over a number of months. He nevertheless had thoughts like: ‘If I am faced with a real problem, I shall not be able to deal with it, and everything will collapse round my ears’. As a homework task, he thought about what he would say to a friend in a similar position. He reminded himself of a number of situations over the preceding months when he had successfully tackled difficult situations, including getting a new job, coping with his wife’s unexpected illness, and developing bad dreams as his medication was reduced. He discussed with his therapist the risk that he might focus excessively on times when he was struggling and, to counteract this, they agreed that he would for a couple of months keep a log of any examples of successful coping.

Some clients will not have benefited from treatment, and this can be especially difficult for them if they came to cognitive therapy having had little success with other treatment approaches. If the absence of progress was identified at an early review stage, it may be less dispiriting for the client to finish at that point, where the lack of progress could be attributed to a failure in cognitive therapy, rather than the client. For example, the therapist could say:

‘It seems that we have not managed to make much difference to your problems. Cognitive therapy has been found to be useful with a lot of people, but there are cases when it does not seem to relieve the feelings, however committed the client is to working in this way. Research is still needed into how to find new ways of changing beliefs or behaviour, so that more people can benefit from it, but I think at this point we have to say that cognitive therapy is not going to be helpful for you. Perhaps we should look at what has been helpful, so that you can take away some strategies for helping you feel better. For example, we found that you were good at breaking problems down into different elements, and that then you could tackle difficult situations more easily. Is that something you can take away and use in the future?’

While it may be difficult to end treatment with little in the way of gain, it is unfair to maintain false hopes for a client who is unlikely to benefit. If it seems that a different approach would be more useful, then this should be discussed with the client: for example, if there are significant marital problems then couple therapy, or possibly systemic therapy, could be suggested; or it may be worth considering medication if this has not been exhausted previously. However, happily, the outcome of cognitive behaviour therapy with the majority of Axis I disorders is good and for most clients a plan based on the blueprint will be a more positive note on which to end treatment.



As CBT is structured and focused, it is relatively straightforward to describe the probable typical course of therapy:


  • It is likely to last between six and 15 sessions, longer if there are Axis II problems.
  • The length of sessions varies, but the average is 60 minutes.
  • The first two or three sessions focus on assessment and deriving an initial formulation, at the same time as developing a working alliance, and encouraging the client to become an active and collaborative participant.
  • The next sessions are focused on trying to tackle the client’s problems.
  • The final sessions focus on drawing up a blueprint about what to do if further problems develop.

Throughout treatment, each session will include:


  • Agenda-setting
  • Self-monitoring
  • Dealing with setbacks
  • Updating the formulation
  • Summaries for the client and therapist.

Agenda-setting is important because it facilitates the structuring of the sessions. An agenda typically includes:


  • Review of events since previous session
  • Review of the previous session
  • Assessment of mood
  • Review of homework
  • Work on current topics (about 35 minutes)
  • Homework assignment
  • Feedback on session.

Difficulties with agenda-setting include:


  • Too vague
  • Not prioritising
  • Too full
  • No client input in setting it
  • Diverting from the agenda
  • Misunderstanding the client
  • Beginning the session while setting the agenda
  • Hopping around from topic to topic.

Goal-setting is important, and goals should be:


  • Specific
  • Measurable
  • Achievable
  • Reasonable
  • Time-framed.

Homework is highly valued because being engaged with homework is known to relate to success in treatment. Clients are more likely to do homework if it:


  • Emerges from the session
  • Is relevant
  • Is not too burdensome
  • Is planned in detail with difficulties discussed
  • Cannot be ‘failed’
  • Is written down
  • Is reviewed in detail.

‘Poor motivation’ is analysed within the CT model. It may be a result of:


  • Being too early along the ‘readiness to change’ dimension
  • Inaccurate expectations of treatment
  • Lack of understanding/acceptance of formulation
  • Hopelessness
  • Differing health beliefs
  • Cons of change outweighing pros.

The middle phase of treatment is spent on discussing the details of current situations, in terms of the four systems – emotions, cognitions, behaviours and physiology.

The latter phase is spent devising a blueprint for action in the event of future problems. Your hope by this stage is that your client is competent in CBT as you are.

Learning exercises

Review and reflection:

If you have experience with a different kind of therapy, and if CBT is more structured than you are used to:


  • How comfortable do you feel with imposing structure on treatment sessions?
  • Are there bridges that you feel you cannot cross – for example being explicit if you are going to talk about something that is not on the agenda?
  • If you review the relevant section of the chapter, does that make you feel any more comfortable? Comfortable enough to try it out?

Taking it forward:


  • Take any of the points of possible discomfort from the review exercise, and try to adopt the suggested format for three or four clients. Keep a thought record of your NATs if you feel uncomfortable with the way you are trying to work at any time – you will probably only be able to note down key words/phrases to remind yourself what went through your mind, but you will need to pay attention to the NATs so that you can write them down more fully at the end of the session. Once the session is over, take some time evaluating the NATs, either by yourself or with a supervisor/peer.
  • Think of something you would like to change in your own life, and see if you can do a goal-setting exercise for it, deriving SMART goals. Note down any points that you learn from the process of carrying out the exercise.
  • With your next five therapy sessions with different clients, rate (with a rating scale you design yourself, see Chapter 5) how much you involved the client in the selection and design of homework. Was this at an acceptable level, and if not in a particular case, in what ways could you improve things? Look at how many of the clients actually carried out the homework, and think about how this could be improved.

Further reading

Beck, A.T., Rush, A.J., Shaw, B.F., & Emery, G. (1979). Cognitive therapy for depression. New York: International Universities Press.

This classic text gives a description of the structure and process of both the early and subsequent treatment sessions, and gives a good ‘feel’ for how therapy proceeds.

Padesky, C.A., & Greenberger, D. (1995). Clinician’s guide to mind over mood. New York: Guilford Press.

This book provides a clear account of the treatment process, particularly if you also read the accompanying manual for clients.