Development of Healthy Eating and Active Lifestyles for Diabetes, a culturally tailored diabetes self‐management education and support programme for Black‐British adults: A participatory research approach

To develop an evidence‐based, culturally tailored, diabetes self‐management education and support programme for Black‐British adults, called Healthy Eating and Active Lifestyles for Diabetes (HEAL‐D), using participatory methods to engage key stakeholders in the intervention design process.


| INTRODUCTION
Tackling ethnic inequalities in type 2 diabetes is a global healthcare priority. 1 In populations of Black African ethnicity, rates of type 2 diabetes are around 3 times higher than in populations of White European ethnicity, 2 and it develops, on average, 10 years earlier. 3 Furthermore, poorer glycaemic control is evident at the time of diagnosis, and greater medical management is required to achieve good control. 45 Poor access to diabetes healthcare is a significant issue for minority ethnic groups. 6 Specifically, Black communities report that healthcare professionals lack cultural understanding 7 and their advice lacks cultural relevance 8 or is poorly adapted to their culture and needs. 9 Socio-economic barriers to healthcare engagement are also often cited. 10 In the United Kingdom, healthcare is free at the point of delivery. Despite this, UK minority ethnic groups report finding it relatively difficult to access first-line primary care services for diabetes management. 11 Furthermore, they are more likely to report that they have not had the opportunity to attend a diabetes education course, as recommended in national management guidelines, 12 compared with White populations. These issues may contribute to the poorer diabetes outcomes and increased morbidity experienced by Black patients.
Culturally tailored healthcare is proposed as a primary means by which to tackle ethnic inequalities in health 1314 and is identified as a priority by patients. 15 Culturally tailored education has resulted in greater improvements in diabetes control and knowledge than usual care, and the benefits are maintained long-term. 16 Cultural tailoring of healthcare can take many different forms, for example, tailoring of materials to specific languages or cultural foods. 17 To date, culturally tailored interventions for Black communities have largely been based in America and may not translate to other countries. 16 In the United Kingdom, for example, the healthcare structures are different and so may be the cultural needs of the communities. 14 Participatory methods are an effective means by which to develop healthcare interventions for ethnic minority communities. 18 Through establishing partnerships and fostering engagement of those who are affected by the intervention, participatory methods enable a sustained and open dialogue between stakeholders and researchers to produce a more complete understanding of their needs, ensuring its acceptability and sensitivity to the needs of both service users and providers, whilst also enabling incorporation of culture in the intervention. 1819 These methods focus on sharing power, whereby all stakeholders are considered equal, and have been used in the effective development of culturally tailored health promotion interventions for African-American communities. 192021 We aimed to undertake an intervention development study, using participatory methods to engage key stakeholders in the design and development of a culturally tailored, diabetes self-management education and support (DSMES) programme for Black-British adults, called Healthy Eating and Active Lifestyles for Diabetes (HEAL-D).
Our objectives were the following: • To use participatory methods to understand what is needed, in terms of structure, format, delivery and content, to ensure cultural sensitivity of a DSMES programme for Black-British communities. • To develop HEAL-D as a DSMES programme that is sensitive to the needs of service users and providers and is implementable within the healthcare system. • To refine the programme theory for the HEAL-D intervention.

| PARTICIPANTS AND METHODS
An intervention development study was conducted using participatory methods to engage key stakeholders in the design and development of HEAL-D as a culturally tailored DSMES programme for Black-British communities.

Novelty Statement
• Culturally tailored diabetes self-management education offers greater benefits than standard care for people from minority ethnic backgrounds. To date, culturally tailored interventions for Black communities have largely been based in the United States. • Black-British adults living with type 2 diabetes, healthcare professionals and community leaders participated in co-developing a culturally tailored diabetes self-management education and support programme to ensure its acceptability and sensitivity for both service users and providers. • Participatory methods provide a means to understand the needs of specific communities. This approach enables the development of healthcare interventions that are sensitive to the needs of service users and providers.

| Intervention development
To ensure its clinical utility and potential for embedment within current healthcare services, it was established that the intervention curriculum should align with existing UK diet and lifestyle management recommendations and guidelines 1222 : • Achieve 5%-10% weight loss or weight maintenance in those of healthy weight. • Undertake 150 min/week of moderate-to-vigorous intensity aerobic physical activity plus 2 sessions/week of strength training. • Balance carbohydrate intakes through portion control and promotion of low glycaemic index and wholegrain sources. • Limit saturated fat intake (<10% of energy intake), replace with mono-unsaturated fats. • Limit salt intake (<6 g/day). • Consume oily fish at least twice per week.
The structure, format and delivery of the curriculum, and accompanying resources, were developed in three distinct phases of participatory research (Figure 1). A stakeholder advisory group (SAG) consisting of stakeholder representatives was set up to inform and guide all stages of the intervention development, particularly to seek the opinion of key stakeholders relating to interpretation of the qualitative data.

| Participants
Three stakeholder groups were engaged in the research.

| People living with type 2 diabetes
Men and women (aged 18-75 years) were recruited of selfdefined Black-British, Black African or Black Caribbean ethnicity, residing in London, and with type 2 diabetes (documented diagnosis by healthcare provider). As HEAL-D was intended in its first incarnation only for delivery in English, non-English-speaking individuals, as well as people with other chronic conditions affecting their diet and lifestyle needs, e.g., chronic kidney disease, were excluded.
Recruitment was via letters of invitation sent to people who had participated in previous diabetes research and to eligible participants identified through general practice database searches within the London boroughs of Lambeth and Southwark. Selfreferral was also facilitated through advertisements in the local press and networking with churches, mosques and community groups with predominantly Black-British communities/congregations in the London boroughs of Lambeth and Southwark.
Eligibility was confirmed via a short telephone screening questionnaire.

| Healthcare practitioners
Healthcare practitioners (e.g. general practitioners, general practice nurses, diabetes specialist nurses and dietitians, and commissioners) who provide primary and intermediate care services to Black-British communities in the London boroughs of Lambeth, Southwark, Brent (boroughs recognised for significant representation of Black communities), were invited to participate in the study by email and telephone.

| Community leaders
Community leader participants were identified from the research team's networks and active partnerships, for example, through the Diabetes UK Community Champions network, which is a volunteer health advocacy initiative, and several London churches and mosques. Subsequently, the reach of our recruitment was expanded through 'snow-balling', whereby participants introduced us to other individuals within their networks. The emphasis was on faith leaders, activists and volunteer health advocates, serving or working with Black-British communities/congregations from across London.

| Phase 1: Formative research
Focus groups and interviews were conducted, aiming to explore the relevance, acceptability and feasibility of cultural tailoring methods that have been used in health promotion interventions targeted at Black communities from the literature, whilst also identifying other issues and priorities that might be relevant to the design and context of HEAL-D.
The focus groups were conducted with people living with type 2 diabetes. They were held in local accessible community venues, e.g., church halls, as well as on university premises. Participants were purposively sampled to get a spread of socio-economic position, generational status and ancestral origins, and separate focus groups were conducted with men and women and patients of direct African versus Caribbean ancestry. The focus groups aimed to explore key themes identified in the literature relating to barriers to healthcare engagement, as well as exploring perceptions of diabetes and barriers/motivations to diet and lifestyle behaviour change. A topic guide was used, and prompts such as food models, video clips and body weight images were used to steer discussions and ensure coverage of key themes whilst encouraging free discussion of opinions/perspectives (Data S1). Focus groups were selected to enable us to understand normative needs, as suited to the development of a community intervention.
Semi-structured interviews, conducted with healthcare practitioners, explored issues relating to healthcare needs and engagement. Topics for discussion included experiences of delivering healthcare to Black-British patients and barriers and facilitators to working in partnership with community groups to deliver care for Black-British communities (Data S1). Interview methods were chosen to enable us to gather a full range of experiences and therefore optimise implementation.
Semi-structured interviews with community leaders focused on issues relating to the role of community networks in promoting health and healthcare engagement in Black-British communities and to sustaining health amongst community members (Data S1).
The focus groups and interviews were digitally recorded and transcribed verbatim. The data were analysed using the framework approach in NVivo (QSR International). We also used the Behaviour Change Wheel and COM-B framework (reported elsewhere 23 ) to identify behaviour change techniques that would be relevant to enhance self-efficacy and behaviour change and could be implemented within the delivery of the curriculum. Whilst our initial framework matrix was deductively developed to identify the relevance of existing themes from the literature, we were sensitive to emergent themes in our data (e.g., contextual themes relating to the health system), which helped inform our programme theory, as well as understanding issues of implementation, e.g., favoured settings and timings.
Following initial analysis of the qualitative data, a feedback meeting was held with the SAG. Anonymised interview extracts were presented to illustrate the key themes and issues that were identified. Open discussion/debate was encouraged to enable the research team to understand alternative interpretations.

| Phase 2: Co-development workshops
Following evaluation of the focus groups and interviews, the same stakeholders (patients, healthcare practitioners and community leaders) were invited to participate in 2 half-day (3.5 hr) workshops, held in community locations.
The workshops sought to gain stakeholder involvement in developing the details of the intervention, including the setting, the media channels, structure and delivery, as well as steering the research team to understand patients' literacy and numeracy needs and health literacy. The workshops were structured to begin with a lunch (African and Caribbean cuisine, sourced from specialist caterers) to foster interaction among the participants, followed by small group discussions/tasks in which four to five mixed stakeholders, each with a facilitator from the research team, were provided with tasks.
In the first workshop, scenarios were used to understand important themes that arose from both the formative research and the existing evidence base, namely, setting/location, educator credentials, session timing and literacy/numeracy needs. After discussing the scenarios in their small groups, a volunteering group spokesperson (not the facilitator) presented their feedback to the rest of the participants. A broader discussion was co-facilitated by the spokesperson and a member of the research team to brainstorm key issues, pros and cons relating to alternative scenarios presented to them, for example, regarding the intervention setting, noting feedback on flip charts. The attendees were also asked to review existing educational/support materials, e.g., leaflets, and provide feedback on, for example, language/phrasing, content, pitch and understanding. The research team then facilitated cross-discussion between groups.
The second workshop focused mainly on engaging with people living with type 2 diabetes and aimed to seek input into the development of the HEAL-D education and support materials. Following a lunch, participants were engaged in a series of activities. Participants were shown a selection of educational videos, ranging in detail and complexity, that aimed to explain about diabetes and its management. For each one feedback was sought as to what aspects were clear and understandable. In other activities, participants were given samples of the draft resources that had been developed, for example, educational booklets, and were asked to provide verbal or written feedback on which information was easy to read versus too complex, whether the pictures/images were helpful/ motivating, and what messages of advice they perceived were being conveyed. Following small group discussions, the researchers facilitated feedback and a group discussion to clarify/confirm the researchers' understanding.

| Phase 3: Materials development
During Phase 3 the research team met weekly to discuss the iterative intervention development. The curriculum and appropriate activities/materials, based on relevant behaviour change techniques that were identified through the COM-B framework, were mapped to each behavioural goal. At regular intervals, stakeholder feedback was sought through the SAG, and refinements were made.

| Phase 1 (formative research)
Eight focus groups were conducted, involving 41 people living with type 2 diabetes (14 men, 27 women; 18 Caribbean, 23 African ethnicity; three groups with African women, two each with Caribbean women and Caribbean men and one with African men). The mean age of participants was 62 ± 11.7 years, ranging from 30 to 84 years. The majority of participants (88%) were first generation migrants; the most common countries of birth were Nigeria (n = 17) and Jamaica (n = 13). Semi-structured interviews were conducted with 11 healthcare practitioners, including two primary care physicians, three primary care nurses, four diabetes specialist nurses and two diabetes specialist dietitians. Nine community leaders, including five faith leaders (three Christian and two Muslim) and four volunteer health advocates, participated in semi-structured interviews.
The main findings that arose from the different stakeholder groups, including the emergence of themes and priorities, and the learning we took for our intervention development are shown in Table 1, along with illustrative extracts from the qualitative data that led to them. In relation to existing recommendations in the literature, nurturing collectivism was strongly supported. Powerful social connectedness was evident amongst our focus groups participants; we observed participants being motivated by story sharing and the accounts of others, and participants frequently shared contact details to remain connected outside of the research. Other recommendations that were supported were the relevance of social networks and provision of social support, although these emerged as mainly relevant to women. The findings relating to use of community gathering places, culturally concordant educators and lay educators were less clear. In some cases, there was rejection of these; for example, people living with diabetes were equivocal as to the importance of having culturally concordant healthcare professionals/educators, whilst healthcare practitioners recognised clear benefits. Another frequently cited issue, for which we got mixed opinions, was distrust of conventional medicine and healthcare settings. Whilst there was a strong desire to avoid taking medications and a preference for natural therapies, our patients did express trust in the healthcare system, and healthcare professionals were seen as credible individuals. Whilst our patients did not discuss a rejection of medical settings, our community leaders supported the assertion that community gathering places are a means by which communities can be accessed.

Programme must embed within NHS systems and align with current incentives and metrics.
Note: Relevance was rated as STRONG when the majority of the participants within a stakeholder group provided support for the theme, MODERATE when there was support from around half the stakeholder groups and WEAK when there was no support or support from less than a quarter of the stakeholder group.

T A B L E 1 (Continued)
10 of 16 | GOFF et al.

T A B L E 2
The HEAL-D programme: session learning outcomes, content, materials and behaviour change techniques

Diabetes -It's in your hands! An introduction to T2D self-management
Understand what T2D is and how it develops. Understand the role of insulin, glucose and blood glucose in health and T2D. Be introduced to key ways in which diet and lifestyle influence T2D. Understand the principles of a healthy, balanced diet. Understand how to set personal goals. healthcare professionals and commissioners, the importance of an intervention embedding within current primary care structures and aligning with incentivised targets/metrics was clear.

| Phase 2 (co-development workshops)
Two workshops were conducted, approximately four weeks apart. In the first, 27 stakeholders (three healthcare professionals, four community leaders and 20 patients) attended. The scenario discussions led to the following conclusions: • Intervention setting: avoidance of medical settings was agreed by all stakeholders. • Educator credentials: patients and community leaders were unanimous that the programme educators needed to be appropriately trained/qualified in diabetes management, although they didn't have strong opinions on different practitioner types (e.g., nurses and dietitians). The need for educators to share the same cultural/racial background was rejected but the importance of having cultural knowledge, particularly relating to dietary practices, was recognised. • Timing of sessions: stakeholders stressed the difficulty of attending appointments alongside work and carer commitments and the need for the programme to be delivered at times when they were free to attend. • Literacy/numeracy needs and communication preferences: basic levels of numeracy, literacy and health literacy were evident and a need to avoid medical terminology. Findings from Phase 1 relating to preferences for verbal messages/ advice rather than written documents was explored, with some patients keen to have detailed written information and others finding this off-putting.
After the scenarios, the groups provided feedback on samples of existing educational leaflets. The main learnings were that participants were seeking very clear and prescriptive advice on how much carbohydrate they should be consuming and that guidance on portions should avoid using weight measures but instead focus on household measures, such as a mackerel tin or handfuls, which are more relatable. Other important learnings were the need for leaflets to feature relevant cultural foods and to avoid medical terminology.
The second workshop was attended by 11 participants (eight people living with T2D and three community leaders), all of whom had attended the first workshop. Draft HEAL-D educational materials, including information booklets and educational slides, were discussed by the participants. The main learnings were technical terms (e.g., hyperglycaemia) were off-putting, and simple phrasing was preferred; the use of visual images to illustrate information (e.g., hands to illustrate portion sizes) were favoured as more accessible. The participants liked the colourful nature of the leaflets and the logo and professional quality of the graphics gave participants confidence and trust in the materials. In response to findings from Phase 1 and the first workshop, in which a preference for visual and verbal information was expressed, the research team considered whether film/ video media might be an effective means to convey important programme information. This was explored in the second workshop, whereby participants were asked to view existing educational diabetes-focused films and provide feedback on their length, language and content, to inform the making of a bespoke HEAL-D educational film. Participants were positive about the use of films and recommended that they should contain information specific to Black-British groups (e.g., prevalence rates and foods that affect blood glucose) and the use of animation and visual images increased their understanding.

| Phase 3 (materials development)
The programme resources were developed with continual SAG feedback to ensure relevance and acceptability for participants. For each session, activities and materials were developed to draw on relevant behaviour change techniques, as identified through the COM-B mapping of the patient focus group data. 23 The HEAL-D sessions, learning outcomes, content, materials and behaviour change techniques are detailed in Table 2.
Key aspects of the programme design and the rationale for their use were the following: • Group-based delivery, fostering collectivism and social connectedness, inviting inclusion of a 'significant other' to draw on social support structures, whilst aligning with current healthcare services to increase potential for implementation. • Curriculum aligning with evidence-based diet/lifestyle guidelines, nurturing patients' desire to manage diabetes through 'natural' means, whilst aligning with current healthcare services to increase potential for implementation. • Delivery in non-medical, community settings to overcome issues of distrust and increase convenience. • Delivery by a diabetes specialist dietitian in partnership with a trained Black-British lay educator, ensuring credibility of advice and information, whilst conveying cultural understanding. • Flexible attendance including programme switching to enable participants to attend missed sessions and sessions at convenient times and venues. • Curriculum delivered in seven sessions, on a fortnightly basis, to allow sufficient time for participants to implement and experiment with the behaviour change of focus. • Group-based physical activity classes, delivered by trained exercise instructors, to foster social cohesion whilst demonstrating a range of activities that are culturally acceptable (e.g., walking and dance) and at appropriate levels of exercise intensity. • A practical 'cook and taste' session, delivered by Black-British cooks/chefs, to overcome barriers to changing cooking practices. • Short films, drawing on story sharing and role modelling, to motivate behaviour change relating to limiting carbohydrate portions and increasing physical activity levels. • Culturally tailored information booklets, using photography of cultural foods to illustrate appropriate portion sizes, to reinforce educational messages and behavioural goals. • A culturally tailored animation film, 'diabetes explained', to inform patients of the basic principles of diabetes selfmanagement. This video was publicly available on the programme website to allow patients to re-watch. • Provision of exercise videos led by Black-British exercise trainers to encourage participation in physical activity. • Curriculum delivery using specific behaviour change techniques that were chosen through detailed analysis of our focus group data using the COM-B framework. 23 The components and activities of the HEAL-D programme, mapped to the proposed outcomes, are depicted in a logic model (Figure 2).

| DISCUSSION
For interventions to be effective, implementable and sustainable, they must be responsive to the needs of both service users and service providers. Participatory methods are an effective means by which to identify and understand the needs of key stakeholders, thus aiding the development of sustainable healthcare interventions. 18 We used participatory methods in the development of HEAL-D, a DSMES programme, specifically tailored to the needs of Black-British adults, to identify key adaptations that were needed in terms of structure and format to ensure cultural relevance and acceptability, whilst aligning the intervention with existing healthcare structures to ensure its implementation potential. Our three phases of participatory work enabled us to develop a detailed understanding of the priorities of our stakeholders. The first phase, our formative research phase, identified key issues among our different stakeholder groups. Service users led us to understand the importance of nurturing collectivism and identified a reliance on informal sources of information/ advice, barriers to attending appointments associated with competing priorities of work, travel and carer commitments, and a preference for directness and simple, clear advice/messages. Our service providers prioritised the need for the intervention to embed within current primary care structures and align with incentivised targets/metrics. Our Phase 2 workshops, which brought our stakeholders together, identified key considerations for the intervention structure, format and delivery, particularly avoidance of medical settings, the need for appropriately trained and culturally knowledgeable educators, flexible appointments, preference for verbal and visual information and avoidance of technical/medical terminology. The final phase focused on stakeholder input into the development of the intervention materials to ensure their cultural relevance and sensitivity.
Tackling ethnic inequalities in type 2 diabetes is a healthcare priority in the United Kingdom. 11 Ethnic minority groups often find it harder to access healthcare; for example, healthcare professionals may lack cultural knowledge or understanding, and patient resources usually align to the cultural practices of the majority population. 24 Culturally tailored interventions are proposed as a means by which to improve engagement amongst minority ethnic groups. 14 There have been many different methodologies used in the development of culturally tailored healthcare interventions across a range of health conditions. Barrera and Castro developed a framework for culturally adapting healthcare interventions from a consensus of existing methodologies, outlining five stages in the process: information gathering; preliminary design; preliminary testing; refinement; and final trial. 25 The work we report here for the development of HEAL-D aligns with Stages 1 and 2 of this framework in which we have conducted extensive formative research to inform the design of our intervention. The next stages of our work are to deliver HEAL-D in a pilot trial (Phase 4, see Figure 1), with community evaluation to identify refinements that are required prior to testing efficacy in a definitive trial (to be reported separately).
To date, culturally tailored DSMES interventions for Black communities have mainly been developed and evaluated in America. 16 Whilst some aspects of the format and structure of HEAL-D are comparable to the American interventions, for example, the use of non-medical settings, there are many aspects of HEAL-D that are distinct. This is largely due to the focus we took in ensuring HEAL-D was sensitive to the needs of the healthcare system, maximising its potential to embed in primary care beyond the research period. Our involvement of healthcare practitioners in our intervention development enabled us to understand the priorities and needs of the healthcare system, whilst identifying scope for change. Key aspects of health system alignment were the need for the programme to meet criteria for performance metrics/incentivised services (i.e., evidence-based curriculum, delivered by healthcare professionals) and for it to embed within the referral pathway for structured education. In some cases, our formative research identified opposing needs/priorities between patients and practitioners, for example, delivery by cultural concordant practitioners; in these cases, our interactive workshops enabled us to work with our different stakeholders to explore these issues and understand how to structure the intervention to meet all needs.
Often intervention tailoring for minority ethnic groups is targeted at observable, superficial traits such as language, where educational materials are translated into the first language of the target population. To date, in the United Kingdom, culturally tailored diabetes structured education programmes have mainly used these types of adaptations, targeted at South Asian communities, which form the largest minority ethnic group in the UK population. 1626 In the Resnicow et al. model of cultural sensitivity, which was developed to inform public health intervention development, traits such as language are considered 'surface' structures of culture, and tailoring at this level is proposed to increase the receptivity or acceptance of messages. However, the model proposes that tailoring to 'deep' cultural structures, such as health beliefs, is more likely to result in an intervention that conveys salience. 27 Our participatory methods enabled us to gain an in-depth understanding of our target communities and the factors that influence their health behaviours and to target these 'deeper' structures of culture in our tailoring. Whilst there is significant diversity in participatory research methods, they all centre around a 'bottom-up' approach, in which research is carried out with the people or communities who will benefit rather than on them. The key aspect of participatory research is partnership working and the location of power, which is shared equally between the research team and the community participants as opposed to conventional research methods that involve limited interaction and decision making with people outside of the research team. Participatory research originated in the social sciences and has been criticised by conventional health researchers because it is less well defined and harder to measure and reproduce than mainstream health research methods. 28 In intervention development, it is the perspectives of the target community, rather than the research team, that define the intervention details. Whilst we strove to remain true to the ethos of participatory research in this intervention development study, at times, we were limited in our ability to do this by competing needs; for example, the requirement for our intervention to deliver an evidence-based curriculum meant that our stakeholders could not make decisions on this aspect.
The strengths and limitations of our work warrant consideration. We have used participatory methods to conduct comprehensive intervention development research. Engaging key stakeholders in this process has enabled us to identify and prioritise key components of our intervention structure and format, ultimately seeking to ensure that our intervention is sensitive to the needs of both service users and providers. The key components of our programme that will effect change are presented in our logic model. Ultimately though we must be cautious in drawing conclusions about the appropriateness of our intervention until it has been implemented and evaluated.
It is important that our future evaluation work tests each part of our programme theory to understand whether it operationalises as proposed; such an approach will require further engagement of key stakeholders. We acknowledge that our intervention was developed with a focus on the UK healthcare context and our findings may not translate to broader contexts. Whilst we believe many of our formative findings are relevant to Black communities in other regions, there are aspects of our intervention that are specific to UK healthcare provision. Our methods, which provide an in-depth insight into intervention barriers and facilitators, have demonstrated the importance of understanding the context in which an intervention is trying to effect change.
Participatory methods, engaging key stakeholders in the intervention design process, enable the development of healthcare interventions that are sensitive to the needs of service users and providers and enable intervention designers to understand the needs of their target communities. Engaging communities who might otherwise have poor access to healthcare in intervention development offers an opportunity to tackle health inequalities through the development of culturally sensitive healthcare interventions.