Chapter overview
We have demonstrated the feasibility of:
co-producing a culturally adapted version of an extant, evidence-based model of FI
197 in partnership with African-Caribbean service users, their families and health-care professionals
implementing this new intervention (CaFI) in community and inpatient mental health services using a RCT methodology
delivering CaFI in the absence of families via FSMs.
We have also demonstrated CaFI’s acceptability to African-Caribbean service users, their families, FSMs, therapists and other health-care professionals. Thus, we have successfully answered all our research questions and achieved all the aims of our study.
In the context of historically adverse relationships between African-Caribbean people and mental health services, characterised by fear, mistrust and avoidance,6,22,206 these are significant achievements.
The discussion that follows focuses on the cultural adaptation and co-production processes, and the feasibility of implementing and evaluating CaFI in today’s NHS. We shall highlight strengths and limitations of the study, including the generalisability of findings. We conclude by considering the implications for clinical practice and future research of such an approach both within the African-Caribbean community and in the broader context of BME groups across the NHS.
Discussion of findings
Developing CaFI
We have provided evidence on how to successfully culturally adapt FI for African-Caribbean people diagnosed with schizophrenia and related disorders (research question 1) using co-production approaches. This is important because although a number of previous studies have developed cultural adaptation frameworks,69,136 none has been explicit about the process.
First, our systematic review207 identified key components of adapting psychological interventions for minority groups. We used these to develop our cultural adaptation framework comprising:
concepts
cultural norms and practices
culturally relevant content
communication and language
context
cultural competence of practitioners.
Second, in key stakeholder focus groups, these concepts were applied systematically to the content and delivery of the extant FI34 model that we sought to culturally adapt.
Third, the final model was agreed in a consensus conference comprising ‘experts’ by experience and profession. As in previous studies,61,136 respondents did not suggest changes to the core components of the intervention. Rather, they highlighted additional factors to improve the cultural specificity of its content, acceptability and the feasibility of its delivery. Experts also identified the training needs of CaFI therapists and FSMs together with suggestions for how these might be met.
In developing CaFI, the ‘psychoeducation’ component of the intervention underwent the greatest change from current FI. The most significant change in this regard was adopting an explicit ‘shared learning’ ethos of delivery (vs. ‘psychoeducation’). ‘Shared Learning’ goes beyond good clinical practice, acknowledging therapists’ need to acquire/develop culturally appropriate knowledge and understanding,159 as well as families’ learning about schizophrenia and psychoses. This approach was advocated by stakeholders explicitly to address the power imbalance in therapy and to reduce social distance and hierarchy between therapists and families. Other major adaptations included adding specific communication and advocacy sessions to help families understand their rights and to improve interaction with services. These adaptations were regarded as especially pertinent for fostering therapeutic engagement with African-Caribbean people, given consistently reported delayed engagement or non-engagement with services.30
The therapy manual and supporting materials were designed to raise therapists’ awareness of what it means to be a person of African-Caribbean origin living with a schizophrenia diagnosis in the UK. The manual therefore includes information about the location, customs and cultural diversity of the Caribbean islands (‘there’s no such thing as Caribbean culture’), as, during focus groups, health professionals identified this as a gap in their knowledge. The manual also supported exploration of issues that many African-Caribbean people associate with onset of psychoses, including experiences of everyday discrimination and institutional racism.23,208
Issues of race and racism were integral to many African-Caribbean service users’ and families’ illness models. According to them, and to health professionals in our study, racism often remained ‘the elephant in the room’ (an obvious problem or risk no one wants/feels able to discuss) during therapy and encounters with health-care professionals more generally. This is because African-Caribbean people fear that raising racism might be either regarded as evidence of paranoia or interpreted as attacking health-care staff. From staff’s perspective, they reported a lack of skill in addressing this sensitive and potentially personally challenging issue. Understanding the relationship between African-Caribbean people’s illness models and their experiences therefore provides an important context for deeper understanding of contextual factors and establishing therapeutic alliances with people who, understandably, may be anxious and/or appear avoidant in therapy.
Can new interventions in family intervention be delivered in today’s NHS?
An important element of any feasibility process is understanding if and how new interventions can be delivered in the services for which they are designed. Psychological therapies have become well evidenced and NICE recommended in recent years in mental health broadly and in schizophrenia and psychosis specifically.30,58 Notwithstanding the evidence of its efficacy and cost-effectiveness, FI provision within mental health services is inconsistent.161,162 This is partly because FI is perceived as being labour intensive and, therefore, expensive despite being contrary to the evidence, suggesting failure to take into account the financial, social and psychological cost of the burden of care experienced by families.209
In the CaFI study, we encountered challenges in recruiting suitably qualified and experienced FI therapists (Agenda for Change band 7) as recommended by NICE11 in the host trust. We subsequently recruited band 6 (advanced practitioner) non-psychologists ‘with experience of FI’. Having recruited CCs as therapists, we found that they were delivering CaFI alongside their routine roles and responsibilities without agreed protected time. This often meant prioritising clinical responsibilities over delivering the intervention, causing delays in delivering sessions and the ironic situation whereby our ‘hard-to-reach’ group of service users and families were being placed on a waiting list to receive therapy. As a result, we employed newly qualified band 7 clinical psychologists from other Greater Manchester trusts on a sessional basis to fill the gap in local provision. The fact that we were able to retain service users, families and FSMs despite these delays bears testament to the high levels of engagement and commitment to receiving the intervention that we had achieved.
Although exacerbated by the organisational and financial issues affecting our host trust, our findings highlight previous concerns about the ongoing lack of suitably qualified therapists to deliver FI in the NHS,161 despite NICE guidance.30 The fact that we were able to recruit staff from neighbouring trusts does indicate that this may a feature of service configuration versus actual numbers of available therapists. In the absence of clear information about the availability of FI therapists across the country, we advocate that future trials consider alternative approaches to lack of trained FI therapists in the NHS such as using clinical psychology and psychiatry trainees to deliver psychosocial interventions. This proposal would have the additional benefit of building NHS capacity by creating a broader pool of available FI therapists.
Our experience of delivering CaFI raises concerns about the implementation of new and, indeed, existing, interventions in the NHS. We acknowledge that the financial constraints and service pressures we encountered within our host NHS trust, resulting in its dissolution and merger with a neighbouring organisation, were extreme. Although NICE guidance30 is likely to influence future practice, current evidence suggests that barriers to implementing psychological interventions are widespread and apparently entrenched across the NHS.210 It is noteworthy that we encountered additional barriers such as missing data and inaccuracies in recording information, including ethnicity and assessment of relapse. Given that the NHS has a statutory Public Sector Equality Duty to provide services equitably,132 it is difficult to understand why these data were not routinely collected, especially given policies aimed at ‘DRE [Delivering Race Equality] in mental health’.206 The NHS equality and diversity mapping exercise (2017) should provide more up-to-date information about the organisational capacity to deliver against the various mandated equality and diversity standards across England. We also advocate process evaluation and inclusion of implementation science in future trials to evaluate the extent to which the challenges we faced are being addressed by NICE guidance and related commissioning. Additionally, as some of the data inaccuracies related to mislabelling African service users as Caribbean, recruitment into future trials including both African and Caribbeans might be less problematic in terms of data accuracy.
The implications of increasing financial constraints and service pressures for our study and future studies in this domain are various if such interventions are to be taken forward based on successful cost-effectiveness and clinical effectiveness evidence. Not only do such interventions require impetus and support from local commissioning, local services and manpower also need forward planning to ensure the availability of therapists of sufficient seniority to supervise more junior clinicians should such an intervention be taken forward with successful cost-effectiveness and clinical effectiveness evidence. In other words, new interventions must be commissioned and specifically planned with specialist training and clear career pathway development considered from the outset. This means that investment in FI cultural adaptation and in future full-scale trials needs to be considered in the longer-term plans of those commissioning and delivering local mental health services.
Implementing and evaluating CaFI
We achieved our primary outcome of recruiting to target. Of 74 eligible service users, 31 (42%) consented (exceeding our original expectations that 1 : 3 would participate). We also achieved a high retention rate. Twenty-four of the 26 recruited family units that commenced CaFI (92%) completed therapy despite significant organisational service barriers that resulted in service users and families being on a waiting list for some time. This is a major achievement in this group, whose engagement with services has been poor historically and who are under-represented in research.
Our success is at least partly attributable to novel ways of engaging with this community. These include hosting events in ‘Black-majority churches’ and community centres, publishing articles about Black mental health in local newspapers and instigating discussions about mental health via ‘phone-ins’ on local radio. Going into this community and asking them to identify their research priorities and how to address them enabled us to co-produce a solutions-focused research.
It is noteworthy that, despite decades of previous research identifying the lack of access to psychological therapy and family disruption in this community, CaFI marks the first attempt to develop a culturally appropriate intervention. Although the fact that the PI was a member of the African-Caribbean community might have facilitated access, this alone cannot account for the high levels of engagement (especially as the remainder of the research team and all but two of the therapists were White).
All service users, families and therapists found CaFI acceptable and would recommend it to others. Service users and family members were especially positive about the ‘shared learning’ ethos of delivery and opportunities to acquire/enhance communication and advocacy skills. They also commented positively on the therapists’ personal qualities and self-disclosure. Willingness to share their own personal experiences was important for trust building and normalisation of experiences. This is especially important for members of this ethnic group given previously reported lack of discussion about mental illness among community members due to misconceptions, stigma and cultural taboos about ‘talking your business’.35,36
Only one FSM dissented from the overwhelmingly positive evaluation, indicating that, from the their perspective, CaFI had not met the needs of one service user. According to the FSM, the service user, who had deep-rooted fears of services due to experiences of racism, perceived the therapy content to be ‘White’. Specifically, the FSM did not believe that CaFI addressed the issue of racism sufficiently. Although this was not reflected in qualitative interviews with other participants (including the service user in question) in the acceptability study (see Chapter 7), it raises an important issue about whether this FSM’s perception of CaFI’s as ‘White’ related to its content or to the fact it was delivered by White therapists.
Our exploration of ‘ethnic matching’ of therapists and families reflect previously reported ambivalence about if and to what extent shared ethnicity between therapists and clients affects the nature of therapeutic engagement.211 Although conceding the potential value of ‘ethnic matching’ for some families, participants concluded that this was impracticable in a NHS in which the majority of therapists are White or Asian. Moreover, endorsing findings from a meta-analysis which reported almost no benefit to treatment outcomes from racial/ethnic matching of clients with therapists,211 they suggested that the most important factor in therapeutic alliance was establishing trust as the basis of open communication.
Culturally adapted Family Intervention therapists reported that they valued the opportunity to gain in-depth knowledge about and experience of working with members of an ethnic group known to have complex needs. This enabled them to utilise both advanced and newly acquired skills. Although those who received the full training package (all non-psychologists) reported significant increase in cultural awareness and confidence (see Chapter 5), this was not reflected in the trial or acceptability data. Psychologists, although more recently qualified than non-psychologists, were more likely to deliver CaFI within the therapy window and were more positively rated by family members. This may reflect clinical psychologists’ greater familiarity with the model from which CaFI was derived and more advanced training such as cognitive analytic therapy, which afforded greater flexibility and skills to work with complex client groups. However, lack of protected therapy time and competing clinical priorities undoubtedly contributed to non-psychologists’ self-reports of sometimes insufficient time to reflect, plan and prepare for sessions.
It is also noteworthy that the clinician who independently rated CaFI sessions (see Chapter 6) noted that non-psychologists tended to conceptualise psychosis using ‘illness model language’, which did not facilitate discussion about cultural appraisals or alternative models of mental illness. This is important because this and previous studies36,212 indicate that, in common with other BME groups, African-Caribbean people’s conceptualisations often differ from those of White Westerners, who make up the majority of NHS therapists. Cultural competence and ‘psychological mindedness’ are therefore important considerations for delivering CaFI and other interventions, suggesting that capacity building is warranted if such interventions are to be delivered by non-psychologists.
Across all groups, the most frequently mentioned benefits derived from CaFI were improved understanding of schizophrenia and psychosis and communication between service users and health-care professionals, as well as with their families (see Chapter 7). Some CCs also reported improved communication with families. For example, one CC reported that a service user’s partner had begun to engage with her for the first time and was demonstrably more supportive and less dismissive of his illness. This is important given the well documented lack of engagement and adverse relationships between mental health services and African-Caribbean people. We know that this leads to increased carer burden, more family conflict (often resulting in families involving the police) and service access via adverse care pathways.8,16
Delayed access is also costly for the NHS. The mean inpatient stay of African-Caribbean people is around 2.5 times longer than that of their White British contemporaries. The former also are more frequently hospitalised and more likely to be released on costly CTOs.11 Improved communication between service users, their families and health-care professionals might facilitate earlier access to care, with an associated reduction in costs. Our findings suggest that evaluation of the clinical effectiveness and cost-effectiveness of culturally specific interventions, such as CaFI, in fully powered trials is warranted.
An important finding from our acceptability study was participants’ view that CaFI should be rolled out across the NHS and made available to all ethnic groups versus limiting its perceived benefits to African-Caribbean people. Participants further suggested that, as it was not feasible to implement different culturally adapted versions of FI for each ethnic/cultural group in the UK, a culturally adaptable form should be developed. This marks an interesting and important departure from the increasing number of culturally adapted psychological interventions.58,61,213
Although culturally adaptable approaches were more often suggested by African-Caribbean service users and health-care professionals, they were also endorsed by other participants. Given recent changes in the UK population’s ethnic make-up,214 NICE guidance on implementing FI30 and the changing commissioning landscape, a culturally adaptable FI is appealing. However, the service barriers we experienced, principally a service virtually denuded of FI expertise, would need to be addressed for the benefits of CaFI and other forms of family work to be fully realised.
Study outputs
We have produced a series of new resources that can fill a gap for new practice and capacity building in NHS:
Service user, carer and community involvement
A major strength of this study was the collaboration with key stakeholders to co-produce the intervention. Service user involvement in identifying our research question, study design and membership of the research team has been acknowledged at a national level with the NIHR Mental Health Research Network’s 2014 award for outstanding contribution to patient and public involvement.
This level of engagement marked the culmination of several years of the PI’s work within this community, for example, delivering community mental health conferences to raise awareness of mental illness and to counter stigma. During one of these conferences, our service user co-investigator spoke openly about his experiences of being ‘sectioned’, police involvement in his care and several episodes of hospitalisation. The fact that he is now a church minister and married with children gave credibility to his assertion that recovery is possible. It also countered the tendency of some community members to attribute mental illness to deviance and/or demonic possession. Such open dialogue and self-disclosure is rare in a community renowned for unwillingness to discuss mental health issues.36
A long history of negative experiences of mental health care has made members of the African-Caribbean community fearful of engaging with mental health services and practitioners.23 Service user participants’ narratives indicate that many of them have experienced adverse care pathways and coercive care. For example, one service user (a single parent with no family locally) spoke about being ‘sectioned’ while her children were at school. She reported that six police vehicles came to her home, reinforcing negative associations among her neighbours between mental illness and dangerousness and deviance. She is now actively engaged in research and training health professionals based on her experience.
Families also spoke about routinely being excluded from care planning and feeling unable to advocate on behalf of service users (see Chapter 3). Nevertheless, motivated by a strong desire to improve mental health care, they were willing to help develop and test culturally appropriate ‘talking treatment’, the absence of which they highlighted as a significant gap in service provision during our community engagement events. In particular, they raised concerns about White therapists being able to understand African-Caribbean people’s cultural norms and the context of their lived experiences, which many associated with elevated rates of schizophrenia diagnosis within their community. A therapy that makes explicit such issues and provides cultural competency training for therapists was regarded by study participants (including service providers) and wider community as much needed and long overdue.
African-Caribbean service users, their families and community members have been actively involved in every stage of the research process:
participating in community conferences and focus groups to identify research priorities
developing the research questions and grant application
as co-applicants and collaborators on the CaFI study
serving as members and chairpersons of the RMG, RAG and the Study Steering Committee (SSC)
disseminating findings at conferences and co-authoring publications.
Our RAG was particularly actively involved in all aspects of developing CaFI. We sought their advice on strategies to maximise recruitment and retention and feedback on all study materials, including recruitment information and materials to support delivery of the intervention. This authentic partnership was integral to establishing CaFI and the study’s credibility in the African-Caribbean community.217 African-Caribbean people’s involvement in determining the content of the therapy, developing and delivering therapists’ training, co-producing study materials and research management (RAG, RMG, SSC), undoubtedly contributed to our success in recruiting above target in our feasibility trial.
Community engagement also facilitated successful recruitment of FSMs, ‘trusted individuals’ who ‘came alongside’ service users to enable them to access CaFI in the absence of families. Service users who could not nominate family members could nominate friends, key workers or others to participate in CaFI with them. When individuals were unable to nominate anyone, they were able to select from a group of individuals recruited to fulfil this role, subsequent to police and other relevant checks. Recruited FSMs provided one-page profiles and met service users to facilitate the ‘matching’ process. Half of the service users (n = 13, 50%) would not have been able to participate without this wholly novel aspect of our study.
There were reservations and lack of clarity about the role of community-based FSMs in delivering CaFI. Although conceding that many service users would not have been able to access the therapy without them, therapists reported that their involvement meant some sessions were more akin to individual CBT than FI. Greater success was reported where FSMs were CCs or key workers as they could support service users in addressing issues related to the delivery of their care. This should be incorporated into future trials. Further work is needed to determine if, as suggested by some participants, community-based FSMs could form a ‘bridge’ to reconnect service users to their families and wider communities. For example, service users spoke movingly about the normalising and enabling effect of working with FSMs such as being able to go to the barber shop and eat in a café. Another service user who had severed all ties with his family had begun to reconnect with them via FSM’s acquaintance with members of his extended family.
Strengths and limitations
This study had several unique strengths:
The qualitative methods (focus groups and expert consensus conference) we employed were essential for establishing CaFI’s credibility with service users, their families and NHS professionals who would commission and deliver it.
Working collaboratively with members of the African-Caribbean community required high levels of engagement to build trust and develop an authentic partnership with health-care professionals.
Service users and carers, ‘experts by experience’, worked alongside ‘experts by profession’ (health and care staff, academics, advocates, voluntary sector agencies, police), in a non-hierarchical way to develop CaFI. The significance of having their voices heard and expertise by experience acknowledged cannot be overstated for a group so often labelled ‘hard to reach’, but who regard themselves as ‘seldom heard’.
We developed a cultural adaptation framework that we applied systematically to an evidence-based model of FI, demonstrating, we believe for the first time, how to culturally adapt an intervention compared with focusing on what to adapt.
The multidisciplinary composition of the research team and health professionals (social workers, OTs, psychologists, RMNs, psychiatrists) involved in CaFI’s development and testing is an additional strength of the study. This enabled a range of different perspectives to be considered, including ‘antipsychiatrists’, voice hearing network members and people committed to abolishing the schizophrenia label. The fact that consensus was reached on the content, delivery and desired outcomes of receiving CaFI by such a diverse group is a key strength of the intervention, facilitating its acceptability and implementation.
The main limitation in terms of developing and undertaking the research in this fashion is that the process was time and labour intensive. Community engagement requires significant investment of time, personal and professional resources. We are also mindful that although we successfully engaged a range of African-Caribbean service users, families and community members, their views (which underpin much of CaFI) might not reflect those of the entire constituency; in particular, service users with the most limited social function or family/community members whose experience of stigma and fear of sanctions, for example, might have hindered participation. Trialling the intervention and undertaking process evaluation in other geographical locations and clinical settings, including forensic services wherein African-Caribbean people are over-represented, would provide opportunities to address these limitations.
Another limitation of the study is that the initial cohort of CaFI therapists’ training in cultural awareness and confidence building was not delivered to subsequent therapists. Instead, they received detailed orientation to the study and guidance in using the therapy manual from the PI, clinical supervisor and RPM. This was due to the service delivery issues at the host NHS trust highlighted earlier in the report (see Chapter 5, Therapist recruitment, First wave of recruitment). Although delivering the same training to all therapists would have improved consistency of approach, our response to the service delivery issues encountered serendipitously enabled us to collect acceptability data on delivering CaFI by both psychologists and non-psychologists and to undertake initial assessment of the potential impact of training.
Conclusions
Equality of access to care and treatment is a founding principle of the NHS. Nevertheless, despite numerous policy initiatives and recommendations from bodies such as NICE,11,30 the mental health care of African-Caribbean people continues to be suboptimal. Lack of access to evidence-based care and to psychological interventions in particular has been identified by NICE and the Department of Health and Social Care over many years.30,153,206 This is now urgently needed, not least to enable service providers and commissioners to fulfil their public sector equality duty under The Equality Act.132
Despite the African-Caribbean community being labelled ‘hard to reach’, we have successfully partnered with members’ community to culturally adapt an extant model of FI. Our findings provide evidence that delivery of CaFI in hospital and community settings is possible, despite considerable service barriers. CaFI is acceptable to all key stakeholder groups who recommended that it should be implemented in the NHS. This is important because acceptability of any new intervention is fundamental to its implementation and embedding in routine practice.
Considering UK mental health care delivered in a multicultural context, we suggest that the learning from CaFI should be applied to developing new ways of working with minority ethnic communities. We advocate going beyond cultural adaptation for individual groups to developing culturally adaptable models capable of meeting the needs of all groups. Measures to evaluate the cultural competence of both individual practitioners and services are warranted.
To increase the feasibility of doing so, our findings suggest that further work is needed to evaluate the clinical effectiveness and cost-effectiveness of this and other culturally adapted interventions. The capacity of services to deliver culturally appropriate care will also require investment in data recording and retrieval as well as training and workforce planning.
