Addressing equality, diversity and inclusion as part of CBT training: a course evaluation study

Abstract

Abstract

It is widely documented that Black, Asian and Minority Ethnic (BAME) communities experience poorer mental health, and have a poorer experience of mental health services. Therapists delivering cognitive behavioural therapy (CBT) in services such as NHS Talking Therapies Services for Anxiety and Depression, are working with increasingly diverse client groups, but treatment access and recovery rates remain below what they should be compared with the White British population. Previous research indicates that CBT therapists may not receive appropriate training that allows them to develop the skills required to work effectively transculturally. The present study therefore aimed to evaluate a CBT training programme within this context, from the perspective of previous course graduates. Thirty participants took part in an online survey with questions requiring both Likert-scaled answers and free-text responses. Descriptive statistics, inferential statistics and template analysis were used to analyse the data. The results of the survey were favourable overall for both White British and BAME respondents. Positive areas of practice highlighted included peer learning within a diverse cohort, building awareness of own biases, and reflective learning spaces. Areas of development included increased integration of teaching focused on adapting CBT models for minority groups, diversification of teaching staff, and reducing fear and avoidance of exploring issues related to equality, diversity and inclusion. Tentative implications for improving CBT training course delivery in this context have been offered.

Key learning aims

1. (1) To acknowledge the challenges faced by service users from BAME communities in accessing equitable mental health treatment, including cognitive behavioural psychotherapy.

2. (2) To consider explanations for why CBT therapists working in NHS services might find it difficult to work effectively transculturally.

3. (3) To establish ways in which CBT training programmes might help therapists to embark more successfully on the journey of developing cultural competence, during training and beyond.

Introduction

It is widely documented that Black, Asian and Minority Ethnic (BAME) ¹ communities experience poorer mental health, and have a poorer experience of mental health services, than those from White British backgrounds. Individuals from minoritised communities are often under-represented in primary care mental health services, yet over-represented in secondary care mental health services (Butt et al., 2015). Indeed, research has shown that those from a BAME background are four times more likely to be diagnosed with a severe mental illness and detained under the Mental Health Act than those from a white background (Barnett et al., 2019). Individuals in BAME communities are often labelled as ‘hard to reach’ (Butt et al., 2015; Memon et al., 2016), without fully understanding barriers to access and engagement with mental health services such as fear and discrimination (Memon et al., 2016). Given that those from minority ethnic and racial groups are more likely to experience psychological difficulties, for example due to their experiences of racism, discrimination and social inequalities (Allen et al., 2014; Wallace et al., 2016), it is imperative that timely and equitable access to mental health services is available.

The NHS Talking Therapies for Anxiety and Depression Service (formerly known as IAPT) was commissioned in 2008 to provide increased access to first-line psychotherapeutic interventions for individuals presenting with common mental health problems. Those practitioners and therapists employed by these services work with an increasingly diverse client population (Faheem, 2023); however, individuals from BAME communities continue to experience lower access, treatment and recovery rates compared with White British individuals (Baker and Kirk-Wade, 2023). The Office for National Statistics (ONS) (2022) reported that between April 2017 and March 2018, individuals from an Asian ethnic background were under-represented in NHS Talking Therapies Services, and treatment rates were also lower for individuals born outside of the UK and those whose first language was not English. Further research has found that racial and ethnic minority groups were less likely to self-refer to these services, or if they did, were less likely to (a) receive an assessment, and (b) receive treatment if assessed (Harwood et al., 2023). Encouragingly, the gap in recovery between White British and BAME individuals has narrowed in recent years. However, concerns remain that this improvement has not been sustained post-pandemic (National Audit Office, 2023). Indeed, authors have advocated for NHS Talking Therapies services to address barriers to accessing treatment for minoritised groups, and eradicate racial inequalities that exist within service provision (Beck et al., 2019; Cénat, 2020). Despite this, providing equitable access to treatment for all ethnic populations is not currently included as a key performance target by the NHS (Lawton et al., 2021).

There are inevitably complex, entrenched and systemic injustices that impede individuals from BAME communities from receiving appropriate psychological therapy in services such as the NHS Talking Therapies Services. Whilst there are evidently problems that require solutions at a structural level, there is an increasing focus on individual therapists being accountable for undertaking anti-racist professional practice, and for delivering therapies such as cognitive behavioural therapy (CBT) in a culturally competent way. Indeed, the updated national curricula for CBT training programmes now more clearly stipulate the competencies that all therapists should develop during their training in order to achieve this (British Association for Behavioural and Cognitive Psychotherapies (BABCP), 2022a, 2022b; Health Education England, 2022). This is an important development, particularly as authors have variously documented the difficulties that therapists experience in meeting the needs of BAME individuals meaningfully as part of treatment. Many therapists lack the confidence to open up discussions with service users about their race, ethnicity or culture due to fears about making mistakes, causing offence, appearing racist or culturally incompetent and coming into conflict with the client (Beck, 2019; Faheem, 2023; Naz et al., 2019). White therapists may also experience difficult emotions such as anxiety, guilt and shame when addressing such issues (Naz et al., 2019). The emotional context of exploring service users’ cultural identities can lead to therapist avoidance behaviours, resulting in problem formulations that are not culturally informed, and which fail to contextualise the person’s distress in a way that validates their experiences of being of BAME heritage. This may be further exacerbated by the use in CBT of models which are heavily influenced by the majority culture (Hays and Iwamasa, 2006) and which understand psychological distress in a way that is separate to the person’s cultural context and experiences (Beck and Naz, 2019). As such, therapists can struggle in practice to work in a culturally sensitive way when using these models, or else can overlook the importance of doing so. Indeed, therapists have reported difficulties in this respect, for example finding it difficult to engage BAME service users in a treatment model that assumes autonomy and independence, when their own cultural context assumes familial expectations, community and collectivism (Faheem, 2023). Further difficulties reported by therapists when working with service users from BAME communities include the challenges of working effectively with interpreters, a lack of culturally adapted therapeutic resources and the absence of key therapeutic words or terms in their own language (Faheem, 2023). Outside of these more practical considerations, it has been strongly advocated that therapists have self-awareness of their own prejudices, bias and stereotypes (Beck and Naz, 2019) and be conscious of how these might present in the interpersonal transactions with the service user (Lawton et al., 2021).

In short, whilst cultural competency is a core skill in psychotherapy, many therapists struggle to work effectively with service users from BAME backgrounds for a range of different reasons; this contributes to poorer treatment outcomes for minoritised groups compared with those from a White British background, and may be in part due to inadequate training in this area (Faheem, 2023). The current evidence, although sparse, would support this position. Authors have suggested that therapists rarely receive training that is sufficient in terms of enabling them to work inclusively (Edge and Lemetyinen, 2019) or that encourages them to examine the role of their own cultural identity, biases and positions of power and privilege (Sue, 2001). Bassey and Melluish (2012) found that CBT therapists working in NHS Talking Therapies Services viewed their training as insufficient in addressing issues of equality, diversity and inclusion, with trainees advocating for more reflective learning activities, culturally competent supervisors, and increased opportunities to learn from peers and lived experience colleagues from BAME backgrounds (Bassey and Melluish, 2012). Some authors have suggested that training programmes often focus on helping trainees to build self-awareness and other-awareness, but give insufficient time to helping them to develop the skills required for competent cross-cultural practice post-training (Jones et al., 2016). Others have noted that cultural competence training often improves therapist knowledge but fails to change attitudes or develop skills, and that this calls for a more integrated approach to teaching and learning strategies (Benuto et al., 2018). Furthermore, such an approach should normalise discussions of difference and diversity within safe spaces where trainers model offering their own experiences, beliefs and limitations (Benuto et al., 2018; Calloway and Creed, 2022).

Despite the clear importance of ensuring that therapists working in NHS Talking Therapies Services are equipped via training to meet the needs of BAME service users, very little research has been conducted to investigate whether this is the case. Given the additional prominence that cultural competence now has in the national curricula for such training courses, it seems timely that programmes evaluate their own teaching and learning strategies in this context. As such, the present study sets out to perform an initial, exploratory evaluation of the integration of issues relating to equality, diversity and inclusion in the Postgraduate Diploma in CBT at the author’s employing university, using the lived experience perspectives of course graduates working in the NHS.

Method

Design

The study utilised a survey design, capturing participants’ quantitative (Likert-scaled) and qualitative (free-text) responses.

Participants

Eighty-nine participants were invited to take part in the survey, all of whom were graduates of the Postgraduate Diploma in Cognitive Behavioural Therapy. Thirty participants agreed to take part, giving a response rate of 34%. In order to avoid enforced categorisation, all participants were asked to self-describe their ethnic background. Twenty participants identified as being from a White British background; of the remaining 10 participants, one described themselves as Black British, three as British Asian, one as Asian/Indian, one as British Indian, one as Asian, one as ‘mixed’ and one as ‘White other’; one participant chose not to disclose their ethnic background. The ethnic background of participants broadly matched that of the Talking Therapies workforce and general population (Health Education England, 2023), with two exceptions. Comparably the proportion of White British respondents was lower (67% compared with 80% in the workforce and 83% in the general population), and the proportion of Asian respondents was higher (20% compared with 9% in the workforce and the general population). All participants were given the opportunity to input any other protected characteristics that they felt were personally relevant and comfortable to share. From these answers, it was evident that the sample was diverse, including participants who identified as cis-gender, disabled, neurodivergent, gay, heterosexual, religious, atheist, having caring responsibilities, married, and dealing with physical or mental health issues.

Measures

In order to evaluate how well participants perceived that the training course had integrated issues pertaining to equality, diversity and inclusion, a comprehensive survey was developed based on the Kirkpatrick model for evaluating training programmes (Kirkpatrick, 1996). This model suggests that evaluation should take place at four distinct levels:

1. (1) Reaction: how did participants feel about the training programme?

2. (2) Learning: to what extent did participants improve knowledge, skills and change attitudes as a result of the training?

3. (3) Behaviour: to what extent did participants change their behaviour back in the workplace as a result of the training?

4. (4) Results: what organisational benefits resulted from the training?

Questions were therefore designed to capture feedback at each level, with each level consisting of a set of questions requiring answers on a 5-point Likert scale ranging from 1 (‘strongly disagree’) to 5 (‘strongly agree’). Each level also included three questions giving the opportunity for respondents to add additional free-text answers to further support their views. Whilst the survey was focused on eliciting feedback in the context of race, ethnicity and culture, one of the free-text boxes was provided for feedback relating to any other protected characteristics that the participants felt were important to raise. A copy of the full survey schedule can be found in the supplementary material.

Procedure

The study received ethical approval from the author’s employing university (ref. P142395). A purposive sample was recruited into the study by emailing all graduates of the PgD Cognitive Behavioural Therapy from the previous three academic years and inviting them to take part. Participants were emailed a copy of the Participant Information Sheet and a link to the survey which could be completed online via Jisc Online Surveys. Informed consent was obtained as part of the online survey link in order that participants could remain fully anonymous as part of their involvement in the study. Furthermore, it should be noted that standard demographic questions around age and gender were deliberately omitted in order that participants felt that they could control the personal information shared during the survey and further keep their identity confidential. This was intended to encourage honest feedback and reduce any social desirability bias that may otherwise be present.

Data analysis

Data obtained from the survey were analysed in the following stages:

1. (1) Descriptive statistics were obtained for the fixed-response Likert-scaled answers in order to provide a summary of responses for the whole sample, as well as for the majority and the minority ethnic group. (The participant who chose not to disclose their ethnic background was included in any whole sample analyses, but excluded from any analyses comparing the two ethnic groups.)

2. (2) Reported satisfaction for individual questions were reviewed, and responses recategorised for analysis into two groups: 1, strongly disagree/disagree/neutral; and 2, agree/strongly agree. Due to the small sample size, Fisher’s exact tests were then performed to ascertain any significant differences in responses between the two ethnic groups.

3. (3) For each level of the Kirkpatrick model, a mean Likert scale score was obtained for the whole sample. A repeated measures ANOVA was then used to elucidate any significant differences between reported satisfaction at each level.

4. (4) For each level of the Kirkpatrick model, a mean Likert scale score was also obtained for both the majority and the minority ethnic group. Independent t-tests were then performed to explore any significant differences in reported satisfaction between the two groups at each level of the survey.

5. (5) Finally, template analysis was utilised by the author to analyse qualitative feedback provided in the free-text answer boxes designed to capture examples of positive practice, areas of improvement required and important issues outside of race and culture. This approach to thematic analysis was chosen due to its contextual flexibility and adaptability to this type of research design (Brooks et al., 2015; Kent, 2000). The six-stage process outlined by Brooks et al. (2015) was utilised, allowing themes in the data to be identified based on an iterative process of data coding and template refinement.

Results

Descriptive statistics

Descriptive statistics for the Likert scaled questions are summarised in Tables 1–4, illustrating how participants responded to each question within each of the four levels of the survey in line with the Kirkpatrick model. Overall, it can be seen that responses were consistently favourable, and that this is the case for the sample as a whole, as well as for both the White British and BAME group. There were, however, some questions which attracted slightly fewer positive responses. At Level 1 (‘Reaction’), only 55.6% of the BAME group agreed with the notion that equality, diversity and inclusion was felt to be a core value of the training programme. At Level 2 (‘Learning’), responses indicated that more teaching from both peers and Experts by Experience from diverse backgrounds would be welcomed, with 73.3% and 83.3% of the sample agreeing with this position, respectively. At Level 3 (‘Behaviour’), the integration of cultural factors into the client’s formulation shows some ambivalence, with 31% of the sample offering a neutral response. Finally, at Level 4 (‘Organisational Benefit’), only 46.7% of respondents agreed that there were coherent links between how the university and individual placements addressed issues of equality, diversity and inclusion.

Table 1. Kirkpatrick Level 1: Reactions – how did participants feel about the training programme? (n=30*)

* White British: n=20; BAME group: n=9; undisclosed ethnicity: n=1.

Table 2. Kirkpatrick Level 2: Learning – to what extent did participants improve knowledge, skills and change attitudes as a result of the training? (n=30*)

* White British: n=20; BAME group: n=9; undisclosed ethnicity: n=1.

Table 3. Kirkpatrick Level 3: Behaviour – to what extent did participants change their behaviour back in the workplace as a result of the training? (n=30*)

* White British: n=20; BAME group: n=9; undisclosed ethnicity: n=1.

Table 4. Kirkpatrick Level 4: Results – what organisational benefits resulted from the training? (n=30*)

* White British: n=20; BAME group: n=9; undisclosed ethnicity: n=1.

Inferential statistics

Individual question responses

It is clear from inspecting the descriptive statistics that both the majority and minority ethnic groups gave similar levels of agreement for most questions. There were two questions where this was not the case, and differences were tested using Fisher’s exact analyses. Firstly, 80% of the White British group agreed or strongly agreed that equality, diversity and inclusion appeared to be a core value of this training programme, compared with only 55.6% of the minority ethnic group, although this difference was not statistically significant (p=0.209). Secondly, only 35% of the White British agreed or strongly agreed that they had had the opportunity to share their learning with colleagues in the wider team, compared with 66.6% of the minority ethnic group, although again this difference was not significant (p=0.206).

Mean responses across survey levels

The mean scores for Likert scaled questions for the total sample slightly decrease at each of the four levels of the survey (see Table 5). A repeated measures ANOVA determined that mean scores differed statistically significantly between survey levels (F _3,87=3.022, p=0.034). Post-hoc analysis with a Bonferroni adjustment revealed a borderline difference between levels 1 and 4 of the survey, with lower scores at the latter level (0.322 [95% CI, –0.018, 0.663], p=0.072). Mean scores at each survey level were calculated for both the majority and minority ethnic group. Normality of the data were referenced using histograms with no problems detected; as such, independent t-tests were conducted which revealed no significant differences between the groups at each level of the survey (see Table 5).

Table 5. Mean scores for Kirkpatrick survey levels

* White British: n=20; BAME group: n=9; undisclosed ethnicity: n=1.

Qualitative analysis

Results of the template analysis for both groups are detailed below in terms of over-arching themes relating to areas of positive practice identified, and areas of course development. Four participants did not offer any free-text responses, and therefore the data are based on feedback from 25 participants only (17 White British and 8 BAME respondents). A visual diagram illustrating these themes is presented in Fig. 1.

Figure 1. Themes identified via template analysis from qualitative feedback responses.

Themes in areas of positive practice identified

Peer learning

Opportunities for peer learning were valued by the respondents in both the majority and minority ethnic group, with it being clear that having a diverse cohort was viewed as being key to this. One participant noted, ‘we had a diverse cohort and had interesting discussions around this’, whilst another reported that ‘students brought their own diversity difficulties which allowed me to explore my own too’. The sharing of experiences was viewed positively, with one respondent stating, ‘as a black person on the course I felt that I was able to at times within group discussions offer my insight regarding diversity’. Feedback here was consistent with the Likert-scaled answers, which suggested that participants would have valued more opportunity to learn from peers from cultural backgrounds different from their own, although it was noted that not all participants would have been happy to share their own experiences in this context.

Awareness of biases

Participants from both groups reported that they had become more aware of their own biases as part of the training programme. Some participants noted that they had become aware of their own prejudices: ‘Shamefully I have understood my own racism that I was unaware of; this leads to an increased awareness of my subconscious racism and bias’, and ‘I learnt about my [own] bias and racism through the discussions that took place’. Some participants also made links between this and their clinical practice, for example: ‘[I have] more awareness of my own biases, experiences and process and how this may impact my work with clients’. Feedback as part of this theme was congruent with the highly ranked scores as part of Level 2 of the Kirkpatrick model, which assesses ‘Learning’ and considers how participants may have changed attitudes as part of the training.

Reflective learning and assessment activities

A strong theme across the sample was the value placed on being supported to engage in reflective learning opportunities and assessment activities. Participants noted how self-practice/self-reflection (SP/SR) groups and reflective discussion groups (RDGs) had been particularly important: ‘reflective groups and SPSR sessions directly raised and addressed issues of diversity’ and ‘Reflective Discussion Groups and SPSR smaller groups explore[d] how race and culture can impact on work’. Other participants also valued the reflective aspect of clinical supervision, noting, for example, ‘during supervision I was actively encouraged to consider how my ethnic background may have been impacting the therapeutic relationship’. Other participants added the importance of assessment activities, for example stating that they were ‘encouraged to reflect on diversity in every assignment’. Again, this qualitative feedback was congruent with the Likert-scaled answers obtained at Level 2 of the survey.

Themes in areas of development identified

Adapting CBT for specific groups

Although some participants noted feeling more able to incorporate a client’s culture into the treatment process, a theme across the sample was the need for increased teaching aimed at adapting CBT for specific minority groups. One participant summarised this by stating: ‘whilst it was implicit that these things could be discussed and indeed were in reflective groups, outside of that, the academic side in terms of teaching protocols were generic and no explanation of how … it might need adapting when considering EDI’. Whilst a number of participants wanted to know more about ‘how CBT can be culturally adapted’, others also noted that it would be useful to have training on adapting CBT when using interpreters, when working with the LGBTQ+ community, and when working with clients who identify as neurodivergent. Qualitative feedback here reflected the slight reduction in mean Likert-scaled scores at Levels 3 and 4 of the survey, which focus on behaviour change and organisational benefit in practice.

Greater diversity of teaching staff

A number of responses pointed to the suggestion that ‘the teaching staff could be more diverse’. One participant commented that ‘having more input from other cultural backgrounds would have been interesting, especially as at the time I was working with a diverse client base’. Others noted the value of having lived experience speakers from more diverse backgrounds, stating ‘bring in guest speakers who are non-white and have experienced poor treatment … This could give a real emotional impact on trainees to understand what it is like being targeted for their “characteristics”’ and ‘it could be helpful to have individuals with lived experience share concerns about equality and diversity’. Again, feedback here was largely congruent with relevant Likert-scaled survey responses.

Better integration

In both groups, it was suggested that issues relating to equality, diversity and inclusion required better integration across the training programme, and that this should be evident ‘from the start’. Responses suggested that rather than having separate reflective spaces or teaching, these issues should be incorporated ‘not as an add-on but as part of the norm in [the] approach to teaching’, should be ‘more integrated within the course materials’. One participant additionally noted that ‘it sometimes risks fuelling further segregation by having a separate module for this’. Better integration of learning with clinical placements was also advocated. One participant suggested ‘if clinical placements are made aware of what we learn about diversity, etc. they might help us build upon this in our services’. Another stated that the university should be ‘aware of the usual demographics of placement areas and provide additional teaching to ensure trainees are more equipped to deal with some of the clients they are likely to meet’. Outside of the training programme, better integration with wider university support was also suggested by one participant, who noted that there should be ‘access to forums where students with difference can link with peers across the university’.

Fear and avoidance (self and staff)

One theme in the responses from the White British group pointed to difficulties with fear and avoidance when exploring issues related to difference, and the need for the course staff to create safe spaces for open discussion. One participant reported that ‘it was difficult at times to speak about honest feelings through worry of causing offence’. Another participant suggested that the course should further consider ‘how difficult this work is. This might help all participants to open up. Think about how all participants feel when confronting their bias’. Themes around fear and avoidance were also present in the feedback from participants in the BAME group, although this focused more on the course team. One participant offered, ‘I feel people [lecturers] are scared to encourage students to do this as they feel they don’t want to cause any trouble’, whilst another encouraged ‘as lecturers, please do not be scared to approach the unapproachable, be vulnerable as we all fit into the EDI/protective characteristic box’.

Discussion

Key findings

The Likert-scaled survey responses in this study were favourable overall, with no observed significant differences between the majority and minority ethnic groups. These numerical responses were congruent with the qualitative feedback provided, but also allowed helpful areas of course development to be elucidated. There was a borderline difference between indicators of satisfaction at Levels 1 and 4 of the survey. This suggested that whilst participants’ reaction (i.e. how they felt about the integration of equality, diversity and inclusion issues) to the training was largely positive, they were less confident that this had translated into clear benefit at the organisational (i.e. NHS Talking Therapies Service) level. Although there were no significant differences between the majority and minority ethnic groups at each survey level, this difference does appear to have been driven by the White British group; this may be explained in part by issues around fear and avoidance, which are discussed further below.

The importance of peer learning within a diverse cohort and the appetite for further opportunities to learn from peers from different cultural backgrounds was clear in both the numerical and qualitative feedback; the call for more diverse course staff and Experts by Experiences was also evident. This supports previous research which has obtained similar findings (Bassey and Melluish, 2012). Similarly, the appreciation of reflective spaces to discuss issues of difference and diversity such as SP/SR groups, RDGs and clinical supervision was congruent with research that has highlighted the need for such spaces within CBT training courses (Bassey and Melluish, 2012). Participants noted that these spaces had allowed them to uncover their own biases, prejudices and/or positions of power and privilege, adding further support for their inclusion within training programmes in line with writings by previous authors (Beck and Naz, 2019, Lawton et al., 2021; Sue, 2001). Despite this, some participants in the White British group noted feelings of discomfort as part of these processes, and resultant avoidance behaviours prompted by fears of inadvertently causing offence to others. This would suggest a process within training that parallels that observed in clinical practice (Beck, 2019; Faheem, 2023; Naz et al., 2019), whereby transcultural clinical skills development is thwarted by experiential avoidance. The further suggestion by participants in the BAME group that course staff should not collude with such avoidance, is congruent with previous authors who have advocated for trainers to normalise such discussions by modelling courage, authenticity and cultural humility as part of the process (Benuto et al., 2018; Calloway and Creed, 2022). Outside of the reflective learning spaces, participants clearly felt that not enough teaching content was included that would enable them to successfully adapt the cognitive-behavioural models taught to more diverse groups. This is consistent with established research which has also found that training courses tend to focus more on building self-awareness, but less so on transcultural skills development (Jones et al., 2016). As noted by participants in this study and previous authors, this suggests that a more integrated teaching and learning strategy is needed, and one which allows trainees to simultaneously build self-awareness whilst also developing key skills in working with BAME communities (Benuto et al., 2018). This may be facilitated by the university developing a more integrated focus on inclusive working practices with clinical placements, in line with participant feedback captured by the survey. Such course improvements may go some way to closing the gap observed between reported satisfaction at levels 1 and 4 of the survey, and to ensuring that all trainees feel that equality, diversity and inclusion are core values of the training programme.

Implications for CBT training

The present study poses some clear implications for CBT training. Firstly, training courses would benefit from ensuring that they have an inclusive recruitment strategy for staff (including lived experience colleagues) and trainees. This would facilitate the bringing together of a richly diverse staff and student body in which mutual learning about issues pertaining to equality, diversity and inclusion can take place. Secondly, courses could consider how they create safe, reflective spaces where discussions about difference are normalised, vulnerability is modelled by trainers, and emerging self-awareness of biases is supportively encouraged. Thirdly, courses might consider the balance of reflective learning spaces with transcultural skills development in their teaching, learning and assessment strategy – ensuring both are fundamentally integrated into the course structure. Finally, courses could consider how they work more closely with clinical placements to ensure a shared focus on inclusive working practices; this could begin with setting up an Equality, Diversity and Inclusion working group which has membership drawn from course staff, trainees, clinical supervisors and placement representatives.

Limitations and implications for further research

Despite highlighting some potential ways to improve how CBT training courses develop trainee therapist skills in working with BAME service users, there are some notable limitations to the present study. The sample size for survey research was relatively small (and smaller still in terms of the ethnic groups represented), limiting the scope and conclusions that can be drawn from the statistical analyses performed; the views collated also applied only to the training course in question. Further research that captures the feedback of a much larger and more diverse sample of CBT therapists working in NHS Talking Therapies Services, and drawn from a range of courses across the country, would be valuable. It should be noted that the survey design utilised is additionally open to response bias; it may be that respondents in the present study had particular motivations for participating (e.g. strongly held personal views or interests) that may influence the themes noted in the analysis. Although the Kirkpatrick model was utilised to facilitate the development of the survey questionnaire, this was developed solely for the purposes of the present study, and has not been psychometrically assessed. Furthermore, whilst highlighting some important learning points, a much richer dataset could be acquired by conducting further interviews or focus groups with CBT course graduates. This could be helpfully augmented by capturing the views of CBT trainers, supervisors and clinical placement staff as well, rather than relying on the unidimensional perspective of the current study. Future research could also incorporate measures of cultural competence more specifically, for example by collating the views of BAME service users that have worked with CBT trainees, and by evaluating recorded session material or other assignments associated with these clinical interactions.

Conclusions

In conclusion, this study set out to evaluate how well this CBT training programme has addressed issues pertaining to equality, diversity and inclusion by asking course graduates to consider their emotional reaction, learning, behaviour change, and perceived service benefits in this context. The respondents have very graciously offered invaluable feedback which the course team will strive to implement as part of efforts to co-create an inclusive curriculum, and one which promotes inclusive therapeutic working practices. It is hoped that their contributions will help other, similar courses in the same endeavour.