Introduction
Cognitive and behavioural psychotherapies (CBT) all share the ‘self-help’ paradigm at the centre of the working model which means that it is ultimately the client that help themselves to reduce clinical symptoms and improve wellbeing. The Socratic method allows the therapist to also be naive to client difficulties and develop a playful curiosity in seeking evidence or a shared understanding of the client's inner world (Overholser, Reference Overholser1994). This questioning and reflecting back helps the client to better understand themselves. Client learning within CBT sessions is based on the notion of ‘collaborative empiricism’ (Beck et al. Reference Beck, Rush, Shaw and Emery1979) and engagement as co-scientists (Kelly, Reference Kelly1955); therefore, discovery and the development of psychological self-management should be a shared process between the therapist and client. A good therapist, therefore, is one who can work well with clients to engage them in the process and to produce good client outcomes. However, the teaching and supporting of the development of self-help skills can be fraught with difficulty, particularly ‘resistance’ to the therapy process (Leahy, Reference Leahy2001). Homework tasks allow the client more space and time with the possibility that resistance will diminish. However, there is a risk that clients who are not engaged in the therapeutic process will also not undertake the homework activities.
Homework is a well-established technique in psychotherapy (Broder, Reference Broder2000) while also being an essential ingredient within CBT (Kazantzis et al. Reference Kazantzis, Macewan, Datillio, Kazantzis, Deane, Ronan and L’Abate2005) with empirical support for its use (Beck & Emery, Reference Beck and Emery1979; Kazantzis et al. Reference Kazantzis, Deane and Ronan2000, Reference Kazantzis, Whittington and Dattilio2010; Rees et al. Reference Rees, McEvoy and Nathan2005; Gaynor et al. Reference Gaynor, Lawrence and Nelson-Gray2006; Thase & Callan, Reference Thase and Callan2006; Haarhoff & Kazantzis, Reference Haarhoff and Kazantzis2007; Dozois, 2010; Garland & Chavira, Reference Garland and Chavira2010). Homework enhances the clients’ experiential learning (Rouf et al. Reference Rouf, Fennell, Westbrook, Cooper, Bennett-Levy, Bennett-Levy, Butler, Fennell, Hackmann, Mueller and Westbrook2004), provides an in-vivo learning experience (Freeman, Reference Freeman2007), and is claimed to be an essential agent for change in alleviating psychological distress (Bryant et al. Reference Bryant, Simons and Thase1999). It provides new insights for clients about their behaviours, thoughts and emotions and seeks to establish new patterns of responding to promote long-term change. Homework is important to help the client generalize from the session with the therapist to their life as a whole (Squires, Reference Squires2001). It also enables behavioural experiments to be undertaken in which the client tries out different ways of responding (Bennet-Levy et al. Reference Bennet-Levy, Butler, Fennell, Hackman, Mueller and Westbrook2004). Homework setting is a collaborative exercise in which the therapist uses problem formulation to help joint decisions about areas to work on and then explores this with the client. However, it can be difficult to get clients to comply with homework tasks and this may impact adversely on the efficacy of the therapy or increase the time for which therapy is needed in order to be effective at producing change.
It is not surprising that homework setting, design and utilization are key competencies for therapists who are applying CBT within UK practice (DoH, 2007). The use of homework features within therapist rating scales such as the Cognitive Therapy Rating Scale (CTS; Young & Beck, Reference Young and Beck1980) and the Cognitive Therapy Scale – Revised version (CTS-R; Blackburn et al. Reference Blackburn, James, Milne, Baker, Standart, Garland and Reichelt2001). Competency rating scales are increasingly used to rate therapist skills on the current national curriculum for the training of CBT therapists in England (DoH, 2008). However, there is a paucity of literature evaluating the in-session processes within CBT (Clark, Reference Clark2004) and on the rating of therapists by patients and how meaningful the therapy was to the patient.
The linking of experiences from one session to the next and taking account of attempts to carry out homework tasks is referred to as ‘bridging’. In her book, Judith Beck introduces a ‘Session Bridging Worksheet’ (SBW) as a focus on the ‘within session’ processes (Beck, Reference Beck1995). The SBW can be benchmarked to a series of evidence-based competencies from the cognitive therapy scales (CTS and CTS-R) and it provides the client with useful prompts to assist with therapy engagement and the therapist with a map of an evidence-based session structure (see Table 1).
Table 1. Relationship between SBW and CTS competencies
SBW, Session Bridging Worksheet; CTS-R, Cognitive Therapy Scale – Revised version.
Other items which the SBW may impact on indirectly could include:
• Pacing and keeping to time (CTS item 6, CTS-R item 4) due to the presence of a shared strategy from the outset.
• Guided discovery (CTS item 7, CTS-R item 9) as the patient may develop psychological insights from preparing for the session.
• Change strategy (CTS item 9, CTS-R item 11) as the patient is invited to identify what is important to them and set a strategy based on the previous session.
• Interpersonal effectiveness (CTS item 4, CTS-R item 5) as the patient may feel valued, validated and understood from being an active participant in the session.
Subjective ratings of adherence to CBT using the CTS and CTS-R represented in the literature are often undertaken by clinical supervisors or expert professionals within the literature. There is a lack of research on ratings by clients and how adherence to the CBT model is meaningful to them. The aim of this study was to explore the use of the SBW as a means of: providing a consistent session structure for all clients (or an ‘in session protocol’); evaluating its effect on homework adherence and the relationship between homework completion, client satisfaction and treatment outcome. Four research questions were explored:
• RQ1: Does the use of the SBW alter the client's perception of the ability of the therapist to set homework effectively? The structure used in the SBW should help clients see the relevance of the homework activity and its suitability for their individual circumstances.
• RQ2: Did the use of the SBW lead to the client understanding the homework task better, reduce barriers and lead to greater compliance and homework adherence?
• RQ3: Does the SBW reduce the amount of time that clients spend in therapy compared to treatment as usual (TAU)? If the SBW reduces clinical symptoms more effectively than TAU then coping levels should be reached in less time.
• RQ4: Does the SBW have an effect on clinical symptom improvement compared to treatment that does not use a SBW. It was believed that structured bridging would lead to beneficial effects compared to TAU.
Methods
The study adopted an independent groups A-B design (Thomas & Hersen, Reference Thomas and Hersen2011) for clinical symptom measurement which measured clinical symptoms at baseline (A) and then after the treatment intervention (B). Participants were randomly allocated either to the SBW condition or the TAU condition.
Participants
The participants represented a convenience sample of clients who attended a clinic that operates an open referral system. Participants were either assessed using a secure online version of the Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I; First et al. Reference First, Spitzer, Gibbon and Williams1997) or a full clinical assessment by a suitably qualified clinical psychologist as part of a personal injury civil claim. The SCID-I has good psychometric properties (Ventura et al. Reference Ventura, Liberman, Green, Shaner and Mintz1998) and has demonstrated effectiveness when used by a range of health professionals (Kashner et al. Reference Kashner, Rush, Suris, Biggs, Gajewski, Hooker and Altshuler2003; Rogers, Reference Rogers2003). All participants in this study met the criteria for a DSM-IV Axis I diagnostic category.
Once the referral was accepted the client signed a consent form for treatment. Participation in the trial required further informed client consent. At this stage, clients were randomly assigned to one of two conditions: TAU or treatment supplemented with the SBW. Participants were all offered standard treatment; however, random allocation of the bridging sheet was approached by a system of random number allocation and offering the SBW to those ‘numbers’ (as opposed to presenting participants), irrespective of clinical presentation and personal characteristics. The number allocation was made prior to the study to reduce selection bias. The targeted sample size in this study was N = 22 (11 per condition); however, some of the original 22 participants dropped out part way through the study. Following attrition, the final sample was 16 participants and the two conditions differed slightly in the number of participants (see Table 2). Participants were all of white British origin and mostly female (n = 10, 62.5%) with a mean age of 39.81 years (s.d. = 12.08).
Table 2. Participants’ characteristics in both conditions
SBW, Session Bridging Worksheet; TAU, treatment as usual.
The sources of referral were similar for both conditions with the majority of participants being referred through insurance companies (see Table 3).
Table 3. Referral route
SBW, Session Bridging Worksheet; TAU, treatment as usual.
CBT models were applied idiosyncratically depending on the case formulation for each client (see Table 4). The variation in the allocation of participants to conditions means that this cannot be considered to be a small-scale randomized control trial.
Table 4. CBT model used
SBW, Session Bridging Worksheet; TAU, treatment as usual.
Ethical considerations
This study was involved a naturalistic evaluation based on routinely delivered treatments provided within a private outpatient clinic. All of the clients received the standard treatment offered by the clinic, but some of the clients also used a SBW as part of their treatment protocol. Ethical clearance was obtained from the University of Manchester Research and Ethics Committee prior to commencement. In addition professional codes were also followed to ensure ethical practice. The treating therapist was a suitably qualified and accredited CBT therapist, accredited with the British Association for Behavioural and Cognitive Psychotherapies (BABCP). The therapist had been qualified in CBT practice for 9 years and was also experienced in the supervision of other therapists. All cases were anonymized and conformed to statutory data protection standards (The Data Protection Act, 1998), and the British Psychological Society code of ethics (BPS, 2010).
Measures
Clinical symptom change was evaluated by the use of the Beck Depression Inventory – Revised (BDI-II; Beck et al. Reference Beck, Steer and Brown1996), the Beck Anxiety Inventory (BAI; Beck & Steer, Reference Beck and Steer1987) and the Beck Hopelessness Scale (BHS; Beck & Steer, Reference Beck and Steer1993). All three measures were used with all participants.
The amount of time in therapy was the number of sessions attended. All participants received therapy adapted to their individual needs and treatment sessions were open ended with no cut-off points.
The ‘Homework Questionnaire’ (HQ; Startup & Edmonds, Reference Startup and Edmonds1994) has four domains and was considered to have good face validity in terms of assessing how well the client thought that the therapist had set the homework tasks (Fig. 1). It was judged for its quick completion and focus on the areas necessary for the evaluation.
Fig. 1. Domains of the Homework Questionnaire.
Participants were invited to rate each domain on a visual analogue scale of 0–7 (see Fig. 2). Item 3 on the HQ, entitled ‘problems expected’ is reverse-scored.
Fig. 2. Example of the scale used.
Since the procedure involved half of the participants using a SBW during the sessions, neither the therapist nor participant were blind to the treatment condition. This could potentially lead to biases (such as the client wanting to please the therapist), so participants were asked to complete the questionnaire and place it in sealed white envelopes away from the therapist and participants were asked for an honest response. Envelopes were sealed by the participant and then labelled by the research therapist with a client number and stored for later analysis. Participants were also reassured that the envelope would only be opened after the trial and following their treatment completion. Participants could request to view the unopened envelopes at any time during their treatment to be reassured that the therapist remained unbiased to their responses.
It is argued that the aim of CBT homework is not to complete tasks in their entirety (Beck et al. Reference Beck, Rush, Shaw and Emery1979; Beck, Reference Beck1995; Garland & Chavira, Reference Garland and Chavira2010), but to learn from the homework, which should be presented as a ‘no-lose’ proposition. Therefore the extent to which homework was completed was converted into a continuous scale and represented as percentages for evaluation. Homework adherence was not recorded specifically for the study and was part of routine clinical practice. Therefore in accordance with professional codes of conduct, all clinical documentation was accurate and contemporaneous. Cut-off points were created in order to provide an operational definition of homework adherence. Ratings between 80% and 100% were rated as full adherence, as this would allow for some variance of a ‘public and private commitment’ to the homework assignment (Kazantzis et al. Reference Kazantzis, Macewan, Datillio, Kazantzis, Deane, Ronan and L’Abate2005). Partial adherence was between 10% and 80%. No adherence was defined as existing between 0% and 10% as a 10% adherence level would be unlikely to provide benefit and demonstrates low motivation towards the task. The cut-off points were applied prior to the collection of clinical documentation narrative and reviewed following the study.
Results
The first question was concerned with how the client perceived the therapist's ability to set homework effectively (see Table 5).
Table 5. Clients’ perception of homework setting
SBW, Session Bridging Worksheet; TAU, treatment as usual.
The SBW seems to have improved the clarity of the homework explanation and improved the anticipation of potential problems in completing the homework (small effects for both). There seems to have been no effect for the clients’ understanding of the rationale and no real difference in the amount of involvement. However, the mean scores indicate that rationale for homework is well emphasized and clients do feel well involved in both conditions.
The second question leads on from this and is concerned with whether the use of the SBW increases homework compliance. High levels of homework adherence were observed in both groups. However adherence was higher for the SBW group (
$\bar{x}$ = 92%, σ = 10.23), with slightly less adherence levels in the TAU group ($\bar x$ = 86%, σ = 20.56). The effect size was small (Cohen's d = 0.4).
The third question was concerned with the amount of time that a client needed in therapy. On average, participants in the SBW group spent eight sessions ($\bar x$ = 8.44, s.d. = 4.79) while the TAU group needed an average of 10 sessions ($\bar x$ = 10.28, s.d. = 2.69). A small effect size was found (Cohen's d = 0.4). This suggests that using the SBW reduces the amount of sessions needed; however, we need to be cautious as there were differences in clinical symptoms between the two groups.
The fourth question asked whether using a SBW would have a positive effect on clinical symptoms (see Table 6).
Table 6. Effect on clinical outcomes
SBW, Session Bridging Worksheet; TAU, treatment as usual.
Values in parentheses are standard deviations.
Irrespective of which condition the participant was allocated, progress in all three measures was evident. In terms of raw scores and magnitude of change, the use of the SBW seems to have had no impact on the clinical outcome. However, the SBW group reached the cut-off levels post-treatment for minimal depression and minimal to mild anxiety; while the TAU group were in the higher categories of mild to moderate depression and mild anxiety. Hopelessness scores for both groups were in the mild level post-treatment.
Within-group effect sizes demonstrated a very large effect observed for the SBW intervention on depression outcomes (d = 1.2), which was observed as a more appreciable effect than the TAU group (d = 0.9). Although a large effect was observed for anxiety outcomes within the SBW group (d = 1.1), a much larger effect was observed for the TAU group (d = 1.5). A smaller, but moderate effect was observed for hopelessness outcomes with a more appreciable effect for the TAU group (d = 0.7), compared to the SBW group (d = 0.5). Overall, the results of using the SBW on clinical outcomes are inconclusive. The TAU group started with more severe clinical symptoms than the SBW group in terms of depression, anxiety, and hopelessness and is a product of random allocation to conditions. This leads to the greatest potential for change in the TAU group and this is supported by the larger change between pre- and post-measure raw scores for this group.
Discussion
Reassuringly the clinical outcome measures suggest that clients made progress, irrespective of whether they were in the SBW or TAU arm of the study. Within-group treatment effects for depression and anxiety are consistent with other larger scale studies.
From studies reviewed in our own literature review (Persons et al. Reference Persons, Burns and Perloff1988; Al-Kubaisy et al. Reference Al-Kubaisy, Marks, Logsdail, Marks, Lovell, Sungur and Araya1992; Startup & Edmonds, Reference Startup and Edmonds1994; Edelman & Chambless, Reference Edelman and Chambless1995; Leung & Heimberg, Reference Leung and Heimberg1996; Bryant et al. Reference Bryant, Simons and Thase1999; Addis & Jacobson, Reference Addis and Jacobson2000; Schmidt & Woolaway-Bickel, Reference Schmidt and Woolaway-Bickel2000; Dunn et al. Reference Dunn, Morrison and Bentall2002; Woody & Adessky, Reference Woody and Adessky2002; Taft et al. Reference Taft, Murphy, King, Musser and DeDeyn2003; Rees et al. Reference Rees, McEvoy and Nathan2005; Gaynor et al. Reference Gaynor, Lawrence and Nelson-Gray2006; Granholm et al. Reference Granholm, Auslander, Gottlieb, McQuaid and McClure2006; Westra et al. Reference Westra, Dozois and Marcus2007; Neimeyer et al. Reference Neimeyer, Kazantzis, Kassler, Baker and Fletcher2008; Ryum et al. Reference Ryum, Stiles, Svartberg and McCullough2010), the mean optimal treatment dose across all studies was 14.67 sessions. It should be noted that the average number of sessions for both groups is not only below that in the literature we reviewed, but also below that reported in the wider literature on optimal dose response of CBT treatments (Hansen et al. Reference Hansen, Lambert and Forman2002). On face value, the average amount of time in therapy for the SBW clients was two sessions less than the TAU clients. This seems potentially to be a positive result in financial terms by saving around £200 per client. The clinical recovery in therapy realized for the SBW group at a mean number of eight sessions is reflective of the optimal number of sessions reported for sudden gains in CBT (Tang et al. Reference Tang, DeRubeis, Beberman and Pham2005, Reference Tang, DeRubeis, Hollon, Amsterdam and Shelton2007). However, this finding needs to be treated with caution as the clinical symptoms at pre-treatment were higher for the TAU group and it may have been that this group required longer to bring their levels of anxiety or depression down to manageable levels and therefore their treatment plan is justifiably more expensive.
At the start of the study there were sufficient participants for us to consider null hypotheses testing and to look for statistical significance. However, as is often the case with therapeutic work, some clients dropped out from the therapy or withdrew consent for involvement in the study. This meant that at the end of the study there were insufficient participants to use inferential statistics to look for statistical significance. The results need to be seen as giving a more tentative indication of what is going on and we have turned to looking at effect sizes. We can also consider the practical clinical significance of the findings.
Small positive effects from using the SBW were evident in the clients’ perception of how well the therapist explained homework tasks and considered potential barriers to the homework with the client. This has been accompanied with a small effect size in increasing the likelihood of the client engaging with the homework. In terms of practical significance, it can be argued that an overall increase of 0.09 on a 7-point scale for improving how well clients think that therapists explain homework is almost negligible (just over 1% increase), where clients were generally positive under both conditions and scored close to the scale ceiling. In contrast, there was more variability in client perceptions of the explanation of the rationale for using the SBW with a very small negative effect size (1% practical decrease for the treatment group). These two results seem contradictory and taken together suggest that the use of the SBW does not affect how clients perceive therapist explanations of the homework.
The more important difference seems to be in the way that clients perceived how well the therapist anticipated potential problems with homework. The effect size is small, but in terms of practical significance to clinical work accounts for 6% difference. Our impression is that its improved perception of dealing with potential problems leads to greater homework compliance. This part of the study suggests that the use of the SBW helps the therapist to address potential difficulties in completing the homework. This leads to greater adherence with the homework tasks and provides greater opportunities for generalization beyond the session. Greater inter-session involvement from the clients could potentially reduce the amount of time needed in sessions. This can be seen as part of the on-going therapeutic process (see Fig. 3.).
Fig. 3. Session Bridging Worksheet (SBW) as a process in the therapeutic work.
Caution should be exercised in this interpretation as there is more variability in the treatment group as indicated by the greater standard deviation for problems expected across the treatment group compared to TAU. The use of the SBW did not seem to make a difference to clinical outcomes when raw score change was examined. In terms of degree of difficulty, all participants in the study started by having a clinical diagnosis that could be addressed through the use of CBT. By chance, the participants allocated to the SBW group had less severe difficulties at the start, and by the end were in ranges that suggested their difficulties were mild while the TAU group remained with mild-moderate difficulties. These results are unclear and seem contradictory. This partly reflects the nature of practitioner research and the way that the participants were allocated to each arm of the trial.
The main limitations of this study have been the scope and size of the study. We are unable to say with any certainty that it makes a difference to the time needed in therapy and recognize the inconclusiveness regarding impact on clinical outcomes. A larger study would have allowed for attrition and for the use of inferential statistics to take account of some of the random variation. Participants were randomly allocated to each strand of the study and this reduced the opportunity to control for important clinical variables, making the interpretation of the clinical outcomes more difficult. This is a small-scale study and a larger study controlling for pre-treatment differences is needed to re-examine this aspect of the study.
Despite these limitations the results are encouraging, at least in terms of how the SBW improves client perception of the skills of the therapist in dealing with potential problems related to homework and the degree of adherence and compliance to homework tasks. It is our view that the SBW makes a positive contribution to the therapeutic process.
Declaration of Interest
None.
Recommended follow-up reading
Beck JS (1995). Cognitive Therapy: Basics and Beyond. New York: The Guilford Press. Chapter 14.
Learning objectives
(1) Session Bridging Worksheets can support the development of homework activities and improve client adherence.
(2) The use of a Session Bridging Worksheet may reduce the amount of time spent in therapy.
(3) The use of a Session Bridging Worksheet is helpful in the therapeutic process.
Comments
No Comments have been published for this article.