An Introduction to Cognitive Behaviour Therapy, 2 edition


Alternative Methods of Delivery


Traditionally, ‘classical’ psychotherapy comprises weekly sessions of a 50-minute ‘therapeutic hour’ of face-to-face contact between therapist and client. CBT has often followed this style but has also investigated whether other modes of therapy may have advantages over traditional models. These alternative approaches are usually motivated by the desire to achieve one or more of the following goals.

Making therapy more cost-effective

This goal arises from a perennial problem in most publicly funded health-care systems, namely that the resources available for psychological therapy are insufficient to meet demand, with the inevitable result that long waiting lists for treatment are common.

It is fairly easy to derive a common-sense ‘equation’ for waiting time, as follows:

This is not a true mathematical equation, but, nevertheless, it helps us to remember that in principle we can reduce waiting times by:


  • reducing the number of people seeking therapy (e.g. by limiting referrals or by improving the general psychological well-being of a population so that fewer referrals arise);
  • increasing the amount of therapist time available (e.g. by providing more therapists or by increasing the proportion of their time therapists spend doing therapy);
  • decreasing the average number of therapist hours taken by each client (e.g. by seeing clients for shorter times or seeing more clients at one time in groups).

Some of the approaches we shall consider reflect the last of these variables: by reducing the amount of therapist contact per patient, they anticipate increasing the throughput of clients and hence offer more or faster treatment.

Improving the accessibility and/or convenience of therapy

Finding an hour a week during the working day (plus possibly lengthy travel time) is not easy for many clients. They may have jobs that do not easily allow them to take time off or that lead to loss of pay if they do so; they may have children or others to care for; or they may live in places that make getting to therapy difficult and expensive. These difficulties make it hard for many people to access therapy, and we need to think of ways to overcome them. A classic paper summarising the arguments for greater flexibility about modes of delivery was written by Lovell and Richards (2000), where they coined the acronym ‘MAPLE’, standing for Multiple Access Points and Levels of Entry. In essence, they argued that CBT should offer clients the mode of treatment that best combines effectiveness, accessibility and economy for that individual.

Improving the effectiveness of therapy

Some ‘non-standard’ ways of delivering therapy have as their main aim the harnessing of extra resources, which the clinician believes will increase the effectiveness of therapy. Thus, practitioners of group or pairs therapy (see below) believe that these approaches are not just economical or convenient but also that they allow the clinician to tackle problems in ways that are not available in conventional one-to-one therapy.

Table 16.1    Main goals of different delivery methods

IAPT Low Intensity interventions

As well as discussing some of the specific strategies under the above headings, we will also briefly cover the programmatic approach that has become known as Low Intensity (LI) CBT within the UK’s Improving Access to Psychological Therapies (IAPT) project.

Modes of delivery for CBT

We consider five main alternatives to traditional therapy in this chapter: self-help, large groups, conventional groups, couple therapy and pair therapy (see Table 16.1).


Here we refer to a range of approaches where clients use media to teach themselves CBT therapeutic strategies, with therapist contact being either entirely absent or much reduced, compared to traditional therapy. Thus we include under ‘self-help’ the following approaches:


  • Bibliotherapy, i.e. client use of CBT books to carry out their own therapy. Although CBT books are frequently used by therapists as an adjunct, we shall focus here on the use of bibliotherapy as a more or less complete substitutefor traditional therapy, where an important aim is to reduce the amount of therapist contact time. Bibliotherapy can be used by clients without any actual contact with a therapist, in what we might call pure self-help (either because some clinician recommends they use a book instead of therapy or because they just pick up a book in a shop); or they may still see a therapist, but for a reduced amount of time (assisted self-help). Although we shall concentrate here on books that are specifically about therapy, we recognise that novels or other books that do not directly offer therapeutic advice might nevertheless assist in the process of therapy.
  • Computerised CBT (CCBT), i.e. the use of computer programs, delivered either by CD- or DVD-ROM or via the Internet, aimed at teaching clients how to use CBT. Such programs often use a multimedia approach, including, for example, video clips, written text, user-completed questionnaires or diaries and so on.
  • Recently developed approaches to self-help include the use of so-called book prescriptionschemes. In this approach, developed by Frude (2005), the public libraries in a locality stock a list of self-help books that people can borrow on extended loan by getting a ‘book prescription’ from a primary-care health worker. Another recent approach is the assisted self-help clinic in primary care, in which clients have brief appointments with a mental-health worker, who guides and supports their use of CBT bibliotherapy materials (Lovell, Richards & Bower, 2003).

The evidence on these approaches is at least modestly hopeful, suggesting that both bibliotherapy and CCBT can give outcomes in primary-care settings which are superior to treatment as usual, and one recent review of guided self-help concluded that it may be as effective as face-to-face therapy (Cuijpers, Donker, van Straten, Li & Andersson, 2010). However, the evidence is limited and the quality of studies is often not high, so further evaluation is needed (Bower, Richards & Lovell, 2001; Lewis et al., 2003; Richardson & Richards, 2006). For example, although an early uncontrolled pilot study on assisted self-help clinics was very promising (Lovell et al., 2003), more recent controlled trials have not shown the same advantages (Richards et al., 2003). In addition, it should be noted that most research findings to date come from primary care, so there is less evidence to support the use of such approaches with more severe or complex problems in secondary or tertiary care.

Despite these uncertainties, self-help approaches continue to be developed and recommended as one stage in a stepped-care programme (e.g. NICE, 2004a). As well as their benefits in terms of cost-effectiveness and accessibility to a wide range of people who might not come to conventional therapy, self-help approaches may have other advantages. They help clients avoid extensive involvement in psychiatric systems – perhaps minimising stigma and dependence; they promote self-efficacy; and they provide a form of help that is permanently available to the client for future revision. There are, of course, also some potential negative effects. Apart from the possibility that they do not work, some people have suggested that ‘failed’ self-help attempts may ‘inoculate’ clients against CBT: they may conclude that CBT is useless and then miss out on what might have been an effective treatment (we know of no evidence about whether this theoretical risk is significant in actual clinical practice).

Our view is that self-help approaches are well worth trying, particularly in primary care, but, whenever possible, their efficacy should be evaluated. Clinical experience suggests that the main guidelines for using such approaches as total or partial substitutes for conventional therapy are:


  • Clients need to be literate and comfortable with reading (or using computers for CCBT) and not have physical or mental disabilities that prevent reading.
  • Self-help should be used as the first step in a CBT approach (not for clients who have already had CBT, except perhaps as a ‘top-up’ for a client who merely wants to be reminded of CBT strategies).
  • Clients need to be willing to give self-help a try: it is probably wise always to check clients’ thoughts about self-help and to help them think through any significant doubts.
  • Self-help may be more likely to succeed with relatively mild and circumscribed problems, rather than complex and enduring problems (but may still be of some help in some aspects of the latter).
  • At least some therapist contact – i.e. ‘assisted’ or ‘guided’ self-help – seems to increase the chances of success (see for example Gellatly, Bower, Hennessy, Richards, Gilbody & Lovell, 2007). This contact might be very limited: for example, Lovell’s self-help clinic used 15-minute appointments, and the average total therapist contact time in a course of ‘therapy’ was just over one hour. Such limited contact is usually focused on suggesting appropriate literature, supporting and encouraging clients’ attempts at self-help and helping them to problem-solve when difficulties arise.
  • In bibliotherapy, there is insufficient evidence to compare the relative efficacy of different books, but the book prescription schemes described above can guide you towards books with some consensus of support from clinicians (see, for example, the list available on the internet from the Devon Book Prescription Scheme, 2004). For CCBT, the National Institute for Health and Clinical Excellence recommends ‘Beating the Blues’ for depression and ‘FearFighter’ for panic and phobias (NICE, 2006).

See Williams (2001) for further discussion of some of these points.

Large groups

Another approach to ‘economical CBT’ is White’s stress-control programme for anxiety (White, Keenan & Brooks, 1992; White, 2000). White’s approach is delivered to groups of 20–50 clients, who also receive a written version of the course content which they can work on during and after the course.

Although calling this approach ‘large group’ conveys one of its distinctive features – the sheer number of people involved – it may in other ways be misleading since it is not group therapy in the usual sense, but is educational, more akin to an evening class. The course consists of six two-hour sessions, usually held in the evenings, in a primary-care or non-health-care setting, and clients are encouraged to bring their partners if they wish. Outcome studies suggest that the programme can result in good outcomes for anxiety disorders, and that improvements are well maintained in follow-up (White et al., 1992; White, 1998). White (2000) gives a comprehensive account of the approach, including practical advice on how to set up and run classes.

Possible advantages of this format include its obvious capacity to provide help to large numbers of people in a way which is very economical of time, both for clinicians and for clients. Its approach to anxiety problems, conceptualised as ‘stress’ that can be managed using teachable skills, may also appeal to populations who would be less likely to access conventional psychological therapies – White originally developed the approach partly to appeal to such groups. Apart from the size of the large group, which means that no one stands out unless they want to, one of the course’s guidelines is that members are discouraged from discussing their particular problems in any detail, a rule that some clients find very reassuring. There may also be non-specific and destigmatising benefits from the sheer size of the class: ‘I can’t be that weird if 40 other people have the same kind of problem!’ On the other side, of course there are clients who will not respond to such a relatively un-personal approach and who might find it difficult to cope with such a large number of people – although bringing a partner along may counteract this.

Low Intensity interventions

As noted in Chapter 1, a major new development of the past few years has been the UK government’s Improving Access to Psychological Therapies (IAPT) project, which aims to improve the availability of psychological therapies for common mental health problems in NHS primary care settings. The first stage of IAPT contains two different CBT approaches: High Intensity (HI) treatment (which is essentially what most of this book is about); and Low Intensity (LI) treatment, which we cover briefly in this section. We do not have enough space to cover LI in depth, so if you are interested we would advise consulting the sources in the ‘Further reading’ section of this chapter.

The LI approach contains several of the features already described above. It is an explicitly ‘high volume, low intensity’ approach to CBT, whose goal is to increase the accessibility of CBT by providing it in ways that minimise both the restrictions on clients and the use of scarce and expensive professional resources. The main characteristics of IAPT-style LI interventions are:


  • It uses a stepped-caremodel, in which a client making contact with services is triaged and allocated to the form of treatment that is least restrictive, whilst still being effective. Clients are routinely monitored on outcome measures, and can be ‘stepped up’ to more intensive forms of therapy if necessary.
  • It uses collaborative careto ensure different professional contributions are co-ordinated and to maintain contact assertively with clients.
  • It uses many of the ways of delivering economicaltherapy that have been described above. Richards (2010) describes LI clinically as ‘characterised by fewer sessions; more emphasis on self-management; the structured and central use of written material as a core strategy, rather than merely an adjunct to therapy; [and] variation in administration methods, such as delivery via the telephone or computer …’.

Typical LI interventions include guided self-help in CBT strategies such as working with thoughts or exposure to anxiety-provoking situations (often using brief telephone contacts rather than full-length face-to-face sessions); a brief form of behavioural activation (see Chapter 17); medication support; and computerised CBT. See the IAPT website (IAPT, 2010) for downloadable resources and the curriculum for the training of LI workers. In the first academic publication on IAPT outcomes, Clark et al. (2009) provide some data suggesting that LI is capable of achieving both good outcomes for clients and impressively high capacities for services treating those clients, with a mean of 2.6 hours of treatment contact.

Two other points about LI work are worth remembering: (a) it should not be thought of as a kind of ‘diluted’ version of conventional CBT – rather both the therapy and the therapists are different in important ways to standard CBT; (b) because of the different therapeutic approach and the high volume of clients (with caseloads of 50 per worker not uncommon), the approach to supervision has to be different to that described in Chapter 19 of this book. For instance it is recommended (Richards, 2010) that cases are monitored and automatically flagged up for supervision by computer systems. For both these reasons, it is not necessarily the case that conventional CBT therapists make good LI therapists or supervisors.

Conventional groups

Another way of reducing costs while maintaining a more clinical relationship with one’s clients is to develop a CBT group, by generalising the CBT approach used with individuals to a small-group format but without mimicking the principles of psychodynamic groups. There are various established group protocol programmes that have been developed for specific disorders (Bieling, McCabe & Antony, 2006). The CBT structure of agenda-setting; monitoring of affect, thoughts and behaviour; re-evaluation of dysfunctional beliefs; homework tasks; and BEs has been maintained in group settings (Freeman, 1983). Initially the focus was on groups for depressed clients (e.g. Hollon & Shaw, 1979), but this has gradually been extended to a wide range of other disorders (see Ryder, 2010 for a review). Apart from economic considerations, there are other advantages of working in this way, including:


  • economy of therapist time (but see discussion below);
  • normalising group members’ experiences, as the symptoms and problems of others are shared;
  • clients often spot in others what was not obvious in themselves – e.g. increased ability to recognise links between thoughts and feelings or others’ cognitive distortions (Rush & Watkins, 1981);
  • group support for doing difficult tasks – e.g. behavioural experiments where courage is demanded;
  • development of a culture of homework completion, etc.;
  • potential for group members acting as co-therapists for each other, facilitating skills acquisition (Hope & Heimberg, 1993), for example, in tracking ‘hot thoughts’;
  • capacity for doing BEs within the group, particularly (but not exclusively) for social anxiety.

However, the advantages need to be offset against a number of possible disadvantages (Tucker & Oie, 2007), including:


  • reduced ability to tailor the sessions to the idiosyncratic beliefs/behaviours of each client;
  • possible reluctance to disclose shameful beliefs;
  • risk of one or a few individuals monopolising the sessions or not really engaging;
  • different improvement rates among the group may be discouraging for some;
  • drop-outs may have a dispiriting impact on the group;
  • potential for unhelpful culture to develop – e.g. off-target discussion or non-compliance with homework.

Nevertheless, the potential saving in therapist time has proved very tempting, and a number of different kinds of groups have been developed.

Format of CBT groups

Groups have been developed for different purposes (e.g. in-patient vs. out-patient), and Morrison (2001) has differentiated them as follows:


  • Open-ended: clients can join for a number of sessions at any point. Such groups may have a strongly educational tone. They necessarily focus on broad issues, for example, links between affect and cognition, with less opportunity to consider individual issues.
  • Open, rotating theme(e.g. Freeman, Schrodt, Gilson & Ludgate, 1993): there is a prearranged programme so not all sessions may be appropriate for any individual client. They are often at a higher frequency than usual – e.g. three times a week.
  • Programmed: highly didactic and the least interactive – similar to the large group format described above.
  • Closed: everyone joins the group at the same time and goes through the whole programme, so all are at a similar level of skill with CBT.

Membership of groups

This is largely dependent on the function of the group. If the group is designed to deal with problems like panic disorder or borderline personality disorder, then there would need to be a screening process. If, on the other hand, the group is intended to increase skills in the management of problems across diagnoses, as may be the case with an open-ended in-patient group, then it would be more likely that a wide range of clients would be included, independent of diagnosis. The relevant questions are what the aim of the group is and who would be most likely to benefit. As the success of a group depends partly on its membership, the issue of who to take in is crucial, and Ryder (2010) has useful suggestions for inclusion criteria (for example, socially able to benefit from the group) and exclusion criteria (for example, actively suicidal) which could be considered if you were planning to set up a group.

Therapist input

The general view (e.g. Freeman et al., 1993) is that it is easier to run a group with more than one therapist, partly because of the twin tasks of providing the technical input (e.g. teaching how to use a DTR) at the same time as attending to interactions between group members. Hollon and Shaw (1979) suggest that six group members is about the maximum an individual therapist can handle unless a co-therapist is available. Ryder (2010) indicates that a greater number can be included as long as the number of therapists is increased, so that group process can be attended to (Yalom, 1995), in addition to teaching technical skills. The roles of the therapists should be differentiated and agreed between them so that there is neither duplication nor gaps.


Open groups can continue for an indefinite period, but closed groups tend to run for between 12 and 20 sessions, usually on a weekly basis in out-patient settings, more frequently with in-patients. They generally last for one and a half or two hours, which allows sufficient time for group discussion in addition to the didactic and technical elements.

Group rules

It is helpful for group members to know what rules should be followed, for example about confidentiality, attendance, punctuality, respect for other members, dealing with crises, and so on.

What outcome can be expected from group CBT?

Morrison (2001) looked at the outcome studies for different kinds of groups, across diagnoses and formats, and these are summarised succinctly in her paper. Overall, it was difficult to demonstrate advantages for group over individual treatment, largely because the studies were inadequate for the purpose: in many studies, the samples were too small (e.g. Rush & Watkins, 1981, for depression; Scholing & Emmelkamp, 1993, for social phobia); or the outcomes for individual treatments were lower than in other published studies concerning the same problem (e.g. Telch, Luxcas, Schmidt, Hanna, Jaimez & Lucas, 1993, for panic); or the group programme offered was not consistently CBT (e.g. Enright, 1991, with OCD). Nevertheless, Morrison concluded that outcome studies generally support the efficacy of CBT offered in groups, although it seems probable that clients with more serious disorders, those with serious depression, or OCD, do better with individual treatment.

Cost-effectiveness of group CBT

Much of the argument in favour of group CBT lies in its cost-effectiveness, but this may be more apparent than real, for the following reasons:


  • group sessions usually last one and a half or two hours rather than the single hour typical of individual therapy;
  • the screening process may be very time-consuming, with referrers taking a chance on clients likely to be unsuitable for the group;
  • there is often a lot of preparation of materials for groups – hand-outs, questionnaires/ratings, etc.;
  • time is taken for preparation of the group programme, probably with a co-therapist;
  • time is required for debriefing with a co-therapist after each session;
  • it may be more difficult for clients to take two hours plus travel time off work; Antonuccio, Thomas and Danton (1997) argued that this cost needs to be factored in when looking at comparative costs;
  • there may be less treatment gain for each individual in the group, and the gain per unit of therapist time may need to be considered.

By all means, go ahead and develop CBT groups, but it is important to evaluate the progress made by individuals in the group and to compare this with the progress made by similar clients seen individually in your own practice or in published research. As long as your clients make progress, you may consider that it is more equitable to offer more people a group, with less expected gain, than it is to offer a smaller number of clients individual therapy, even if you know the small number are likely to do better if seen individually. Morrison (2001) suggested that it may be useful to offer clients two or three individual sessions before moving them into a group, to identify idiosyncratic features for attention in the group and to socialise them to the CBT approach. You may then get the best of both worlds.


Working with couples is another way of increasing the effectiveness of therapy when it is apparent that their relationship is central to the problems presented by one or both clients. The CBT approach to therapy with couples assumes that the beliefs of each client about themselves, their relationship and relationships in general are crucial in understanding how they feel about their own relationship and each other and how they behave towards each other. These beliefs may have been learned early in life and may not be verbally articulated, so a major task is to help the couple identify those beliefs (Beck, 1988). It is important to pay equal attention to each partners’ expectations about relationships and how those expectations may distort their perceptions of their current relationship.

The general principles and characteristics of cognitive therapy apply to this kind of working, with an emphasis on structured sessions and inter-session assignments. The assessment includes a joint session, plus individual sessions with each of the partners, where ground rules are laid down about, for example, telephone calls outside the session and arguing within the session (see Dattilio & Padesky, 1990). Having developed a problem list and formulation, therapy is likely to focus on three broad areas:

1 Modifying unrealistic expectations

This is done following the principles and techniques described for individuals in earlier chapters.

A woman who felt hopeless about her marriage held the belief ‘Unless I am the centre of his life, our relationship means nothing’ and had automatic thoughts such as ‘We never do anything together’, whenever her partner engaged independently in an activity. Therapy involved looking at the evidence for each partner’s NATs and gradually worked towards jointly defining a belief that took account of their experience of current relationships – for example, ‘Our lives can interconnect and overlap in important areas and be separate in other areas; and our relationship can still be meaningful’.

Beck (1988) gives good examples of typical cognitive distortions and how to tackle them.

2 Modifying faulty attributions of blame

It is common for couples to be locked in a vicious cycle of mutual recrimination and blaming, with neither partner accepting responsibility for the difficulties in the relationship. It is a priority to help them identify and re-evaluate their beliefs about responsibility so that they can collaborate in working on their problems.

3 Communication training and problem-solving

Couples typically need help in developing new skills to help them reduce destructive interactions. Communication training emphasises good listening skills, clearly stating one’s needs and taking responsibility for one’s feelings and is well described in Burns (1999). It is important that couples learn how to deal with intense anger while they are learning to communicate effectively, and this can usefully be rehearsed in treatment sessions.

Once they can communicate more effectively, many couples need to learn to problem-solve to deal with areas of disagreement. Jacobson and Margolin (1979) set out general principles for problem-solving in couples, including:


  • specifically defining the problem;
  • focusing on solutions rather than blame;
  • learning to compromise.

Behavioural approaches to couple therapy (for example, Stuart, 1980) emphasise the exchange of positive behaviours, where each partner acted in specific ways to please the other. Within CBT, this strategy can be used to identify dysfunctional beliefs and incorporated into behavioural experiments.

Issues that need special attention within couple therapy include crises, such as recently divulged infidelity or newly developed violence within the relationship. Defusing a crisis would take priority at the early stages of treatment. Other problem areas are where one partner wants the relationship to end; where one partner has a secret (e.g. infidelity) he does not want to disclose; where one partner has another ongoing relationship; and where one partner has a significant psychiatric disorder. Problems such as these are addressed in Dattilio and Padesky (1990) and should be discussed with a supervisor, ideally one experienced in couple work.

Pair therapy

This describes therapy delivered simultaneously to two clients with similar difficulties. To our knowledge, pair therapy in CBT was first presented by Kennerley (1995), who described offering it to trauma clients who wanted to share their difficulties with others in a structured and therapeutic setting but who were unable to join a CBT therapy group. The main reasons for wanting to work with other clients were to destigmatise the experience of childhood abuse and to discover how others coped, objectives that would have been met in a therapy group. The predominant reasons for not joining a group were: being too socially anxious to participate in a group, having personality disorders that precluded them from the group therapy on offer, or facing a lengthy wait for the next group.

Pairs were matched according to similarities in their traumatic history and current difficulties, and then a single therapist took them through the same programme used in a group intervention (see Kennerley, Whitehead, Butler & Norris, 1998). Norris (1995) gives a detailed account of two women’s experience of pair therapy. Although this approach to managing problems related to childhood trauma has not been used in a controlled trial, the preliminary indications were that clients found it acceptable, gained the social benefits of sharing their problems without having to join a group and did as well in treatment as those in group therapy.




  • CBT can be delivered in a variety of ways beyond the traditional format of ‘one patient and one therapist for one hour’.
  • This chapter reviews variations including:

o    Self-help

o    Large groups

o    The IAPT Low Intensity approach

o    Small groups

o    Couples

o    Pairs.

  • All these variations may be useful in improving costs, accessibility or effectiveness compared to conventional one-to-one CBT.

Learning exercises

Review and reflection:


  • Are there any possible disadvantagesto any of these approaches? What are they and how much of a problem do they represent?
  • One often-cited objection to some of these approaches is that they might ‘inoculate’ patients against ‘proper’ CBT – in other words that if a Low Intensity approach or a group approach is not effective, patients might be resistant to trying full-blown CBT and hence miss out on a treatment that might help them. We know of no evidence to support or refute this idea: how likely does it seem to you?
  • Which of these approaches would fit well with your way of working and your client group?

Taking it forward:


  • Is there scope for trying any of these approaches in your own clinical practice, or in the service for which you work?
  • What would you need to do to make this happen? With whom do you need to discuss CBT approaches? How can you best make your case?
  • If the ‘inoculation’ worry seems significant to you, how could you go about checking out how much of a problem it really is in practice?
  • In general – how will you evaluate the impact of adopting a different CBT approach?

Further reading

Bennett-Levy, J., Richards, D., Farrand, P., Christensen, H., Griffiths, K., Kavanagh, D., Klein, B., Lau, M., Proudfoot, J., Ritterband, L. White, J., & Williams, C. (Eds.). (2010). The Oxford guide to Low Intensity CBT interventions. Oxford: Oxford University Press.

IAPT (2010). Web page: Training resources for Low Intensity therapy workers. Retrieved 27 January 2010 from

Richards, D. (2010). Low Intensity CBT. In M. Mueller, H. Kennerley, F. McManus, & D. Westbrook (Eds.). The Oxford guide to surviving as a CBT therapist. Oxford: Oxford University Press.

Ryder, J. (2010). CBT in groups. In M. Mueller, H. Kennerley, F. McManus, & D. Westbrook (Eds.). The Oxford guide to surviving as a CBT therapist. Oxford: Oxford University Press.

Useful sources of self-help materials for patients include:

The Constable Robinson ‘Overcoming’ series of self-help books by leading CBT therapists. See:

Oxford Cognitive Therapy Centre’s booklets for patients. See:

Chris Williams’ Five Areas approach, with downloadable resources: