2.1 Design

This study was qualitative and used interviews.

2.2 Participants

Participants (n = 13) had all been assessed as fulfilling the criteria for a diagnosis of CFS or ME by a doctor in the service. The service adopted a pragmatic approach to diagnosis, in which clinicians used more than one set of diagnostic criteria. The median age of the participants was 49 (range: 19–72). As shown in Table 1, six participants were in paid work, one in education, five were retired or unemployed and one was a home-maker. Participants reported attending between 6 and 10 sessions of specialist CBT at the service (median = 6). Three participants had previously undertaken additional episodes of specialist CBT for CFS/ME at other services. They reported that, as within the service, the goal had been positioned as better management of CFS/ME; therefore, data relating to these experiences were included in the analysis. Change scores were not noted for this study, but participants mainly reported lack of change on physical symptoms. The lack of questionnaire data is revisited in Section 4, 4.

2.3 Procedure

Participants were recruited through an NHS specialist community CFS/ME service. In contrast to the CBT models adopted in many CFS/ME services,¹¹ the model of CBT delivered in this service was not one with the primary aim of increasing activity to facilitate ‘reconditioning’ and a reduction in fatigue. Instead, the primary goal was increased quality of life by better adjustment to and management of CFS/ME. The therapists did not follow a manual or use a specific model of CBT for CFS/ME, but rather drew on a range of CBT techniques and concepts, both specific to CFS/ME (e.g., pacing) and more general (e.g., lowering standards, assertiveness, self-soothing), tailoring the therapy to each individual. A reduction in fatigue was an anticipated secondary benefit for the intervention. CBT was delivered by two CBT therapists in the service who had been qualified for 7 and 11 years, respectively.

Following ethical approval, the service contacted all service users who had completed CBT within the previous 2 years about the study by letter. Participants were given an information sheet containing information about confidentiality, risk and the right to withdraw from the study at any time. Individual semi-structured interviews lasted between 52 and 93 min. Six interviews were conducted in person in a research space adjacent to the service and seven were conducted over Skype. All interviews were conducted by the first author, a female trainee clinical psychologist aged between 30 and 40 years with previous experience of conducting research interviews at the doctoral level. The interviewer was not associated with the service and was not known to the participants. Interviews were audio-recorded and transcribed. Analysis was primarily conducted by the first author.

2.4 Data analysis

Analysis began after the first interview and continued concurrently with data collection. It was informed by the approach described by Strauss and Corbin.¹³ Constant comparison and memoing were used throughout to identify and record similarities and differences between data from different participants and to aid category development. Theoretical sampling¹⁴ was achieved by adapting interview questions according to emerging categories within the data.

Analysis began with detailed line-by-line open coding aimed at identifying concepts within the data. It then progressed to axial coding, exploring the properties and dimensions of categories. Data were then analysed for context, keeping in mind the concept of conditional/consequential matrices.¹³ Relationships between categories were explored by creating maps of each participant's experience. The later stages of analysis involved selective coding and broader theoretical integration. In the final stages, constant comparison highlighted that new data reflected existing major categories in the model and that these categories were dense in terms of properties and well integrated, suggesting theoretical saturation.¹³ Dimensional variation was more limited due to the absence of participants who reported poorer outcomes from the intervention.

2.5 Quality assurance

Elliott et al's.¹⁵ quality guidelines were used to increase the credibility of the research. A research diary was maintained, a bracketing interview was conducted and a positioning statement was created. Two line-by-line coded transcripts, memos around emerging categories and integrative maps were shared with the second author, who offered comments, challenges and elaborations, which were discussed and incorporated into the analysis. The second author, like the first, had no connection with the service. Finally, respondent validation was undertaken to ascertain the credibility of the analysis to participants.¹⁶ A proposed draft of the theory was sent to participants and discussed via phone or email.

3.1 Categories

4.1 Acceptance

In line with studies of other psychosocial interventions for CFS/ME,^19-21 including CBT,¹² several participants reported greater ‘acceptance’ of having a chronic illness. They acknowledged that there was no immediate way of curing or controlling their symptoms, while retaining hope of gradual longer-term improvement. Acceptance seemed to enable participants to further adjust their daily life, such as reprioritising their needs and putting less pressure on themselves. This echoes Pinxsterhuis et al.,²² who found that outside of interventions, the use of adaptive coping strategies in CFS/ME, such as pacing, was facilitated by participants accepting the reality of their condition and finding ways of rebuilding their identities.

Making adjustments in turn gave participants the experience that they could cope with a life with CFS/ME. This then appeared to promote further acceptance of the condition. This suggests a virtuous circle of acceptance, better coping and further acceptance. In their study of CBT, Brooks et al.¹² suggested that acceptance may have been driven by improvements in fatigue, which resulted in the condition ‘interfering’ less in people's lives. The cycle highlighted in this study suggests that acceptance can be driven by a perceived improvement in functioning, even if participants do not perceive improved fatigue.

The study extends existing knowledge by identifying factors that seemed to facilitate this. For some, acceptance resulted from reduced guilt and anxiety after receiving information and/or validation that CFS/ME was not their ‘fault’. For others, acceptance resulted from letting go of high self-expectations, which was described like ‘grieving’ a former identity. Rebuilt identities could then accommodate a representation of CFS/ME, allowing acceptance.

4.2 Limitations

This study did not attempt to represent the experience of all people with CFS/ME who have undertaken CBT. The sample size in this study was typical for qualitative studies, including those in the area of CFS/ME interventions, for example, Picariello et al.,¹⁷ but nonetheless relatively small, and it will be important for future research to explore the extent to which these findings are generalisable to a wider range of people with CFS/ME. Most participants seemed to have benefited from CBT; it may have been the case that service users who did not participate in the research had a less positive experience of the therapy. It would have been desirable to use purposive sampling to capture a breadth of experience, as those with poorer outcomes from CBT on standardised measures may have provided more data around barriers to therapeutic change; however, this was not possible due to limitations in recruitment. Ideally, it would have been helpful to have known the self-reported change for each participant on such measures. However, posttherapy data were incomplete. A wider range of ethnic variation amongst participants would also have been desirable.

Although a strength of this study is the various ‘credibility checks’ that were put in place, participants' responses both during the interviews and respondent validation may have been influenced by a social desirability bias.

Lastly, in terms of replicability, the therapists at the service were not able to provide a therapy protocol. They described general issues often addressed in the therapy (e.g., lowering high standards, pacing), but were guided by clinical judgement to individualise therapy for each client. For this reason, it may be important in future studies to incorporate in-depth interviews with therapists and video-assisted analysis of therapy to more clearly understand therapeutic processes and link these with client-reported experiences and outcomes, such as those described in this study.

4.3 Clinical implications

This study suggests explanations as to how CBT with the primary aim of improving quality of life rather than increasing activity and reducing fatigue can be experienced by service users as beneficial and result in reduced distress about life with CFS/ME. And that this can be achieved in a time-limited and cost-effective way (median six sessions). The results suggest two key areas on which this type of intervention should focus: First, by addressing barriers to service users accommodating CFS/ME in their daily lives and accepting the nature of the condition, for example, providing education and validation and supporting service users to lower their expectations of themselves and better prioritise their own needs. This should involve exploring service users' identities, roles and values and how these may be reconciled with CFS/ME. Second, interventions should provide cognitive and behavioural tools for managing anxiety about CFS/ME and its symptoms (e.g., challenging catastrophising, using grounding techniques) and managing the demands of daily living (e.g., problem solving, addressing worry by separating what can and cannot be controlled). Interventions should also teach pacing where appropriate.

It is important to consider this in the context of the NICE guidance.^{1, 2} The CBT model in this study did not focus on increases in activity despite the 2007 guidance recommending ‘Establishing a stable and maintainable activity level (baseline) followed by a gradual and mutually agreed increase in activity’.¹ The model in our study is, however, in line with the updated 2020 draft NICE guidance that urges against this, except in specific circumstances.² Some of the components of the CBT model in this study could be said to be included in the NICE guidance for example, ‘Addressing complex adjustment to diagnosis and acceptance of current functional limitations’. However, several of the key components of the model are not specified in the NICE guidance and require therapists to draw on a range of CBT concepts and techniques broader than those specific to CFS/ME (e.g., challenging high expectations of self and teaching tools for managing life with reduced resources). Participants in the study valued the ‘holistic’ (participant Susan) therapeutic approach, which allowed them to address a wide range of issues in their lives preventing them from accommodating and adapting to CFS/ME. Some components of the model go even broader than CBT and are more features of supportive therapies such as counselling (offering hope and a space to ‘lighten the load’). It is useful to note that for common mental health difficulties, it has been found that therapeutic alliance and therapist characteristics are key predictors of outcomes across client groups, for example see Budd and Hughes.²³ Both of these may be important factors in this study, in which experienced therapists were adapting treatment to each participant. It is important to note that all participants in this study found it appropriate that the goal of therapy was positioned as better management of CFS/ME and quality of life rather than physical recovery. This model of CBT may not be suitable for service users who view the goal as recovery of pre-CFS/ME activity levels.

4.4 Directions for future research

Research exploring psychological outcomes and processes in CBT for CFS/ME, aimed at better adapting to the illness rather than with the primary aim of increasing activity, is in its infancy. It may be appropriate to explore aspects of the model developed here through quantitative research: for instance, measuring changes in distress, acceptance and quality of life at intervals during this model of therapy. Mediation analyses might include self-efficacy, sense of control, self-compassion, perfectionism, ‘goal discrepancy’ and acceptance or denial of having a chronic illness or condition. It may also be useful to identify the active components of therapy using component analysis.²⁴

Several theories implicate the body's stress hormonal systems as the basis of CFS/ME; it has been suggested for example that the condition arises from the same physiological processes as those that occur in ‘burnout’.²⁵ It could therefore be hypothesised that the improvement in well-being found in this study could lead to a reduction in fatigue in the long term. This requires further research. Multiple case research might examine how apparent improvements in daily living may lead to less experience of fatigue and activity restriction, and the points at which small successes and setbacks occur during the course of CBT, as well as outside of therapy.

Further research might also consider whether the current model could be extended to service users with more severe fatigue, for example, those who are homebound.