Ethical approval was obtained for the CaFI trial from the Health Research Authority in the United Kingdom (19/NW/0385; IRAS ID 254857, approved July 24, 2019). Participants were provided with an information sheet and consent form ahead of the workshops to ensure informed consent was obtained. Participants were reimbursed for their time during workshops in line with National Institute for Health Research rates. Identifiable information was removed from workshop feedback to ensure the data was anonymized, in line with privacy and confidentiality guidance.

Participants were recruited through the participating National Health Service (NHS) community mental health teams and collaborating third-sector organizations. Participants were approached by email via established networks. The sampling method was purposive-opportunistic, and all participants who responded to email approaches were invited to attend the workshops. Service user and carer respondents were included if they were individuals of sub-Saharan African and Caribbean descent and either had lived experience of psychosis or were a carer or family member of an individual who had experienced psychosis. Therapist respondents were included if they were trained health care professionals who had clinical experience of working with people with psychosis. The mixed stakeholder group comprised a combination of people with lived experience of psychosis and health care professionals. Demographic data of participants in each research activity were not recorded consistently; this is a noted limitation of this project. This was a patient and public involvement and engagement activity rather than a research participant activity, with feedback collected from the groups and iteratively updated in the designs. The website was co-designed using an iterative, user-centered process over a 6-month period. Three workshops were held consisting of (1) service users and carers (n=7), (2) therapists (n=4), and (3) a mixed stakeholder workshop. Two individual semistructured interviews were also conducted separately with therapists during the design process. Research team meetings (n=3) were also held to capture and prioritize end-user feedback on the clinician- and service-user–facing components of the website. To ensure the team had time to engage with all participants in the virtual workshops, the maximum number for each session was set to 8 members. Staff time for the therapist workshops was difficult to obtain; therefore, the therapist group workshop was supplemented with additional individual interviews. The mixed stakeholder group was designed to accommodate participants who were unable (because of time or other resource constraints) to attend the single stakeholder groups. The sample size was guided by the concept of “information power” [21], which emphasizes the quality and comprehensiveness of data over the quantity of participants. Rather than focusing solely on data saturation, researchers are encouraged to aim for data sufficiency, particularly when studies have narrow aims, allowing for smaller yet richer samples [22].

Workshops were conducted using Microsoft Teams and lasted between 60 and 90 minutes. The workshops followed a similar structure, starting with an overview of the project followed by a discussion of key features, requirements, and priorities for CaFI:Digital. Design prototypes were presented at each workshop to allow participants to see what the website would look like and enable the participants to provide feedback on each iteration of the designs. The prototype user interface was designed in Adobe XD [23], and interactivity was added using Marvel [24]. Usability testing was conducted using Maze [25].

A co-design approach was combined with Agile software methodologies [26] to support collaboration between people with lived experience, software developers, researchers, and designers. The co-design and user research approach followed the UK Government Service Design Guidelines across the 3 setup stages: Discovery, Alpha, and Beta [27]. Workshops were facilitated by the project manager, the project technical team lead, and a UX designer. In the Discovery phase workshop, service users and carers (n=7) were asked to comment on key requirements and features for the website (see Table 1).

The Alpha phase involved 3 stages; after each stage, the prototype was updated to incorporate feedback from the research activity:

The beta-test phase involved usability testing with 9 participants (including service users, carers, therapists, researchers, and software team staff) to ensure the final website, co-designed through the Discovery and Alpha phases, was defect-free and ready for use. The beta-test phase was completed by sending email requests to service users from the NHS and voluntary sectors and therapists with details of exercises to complete in their own time (for details of the tasks, see Multimedia Appendix 1). The tasks were designed to provide feedback to the team on the real-world usability and accessibility of the platform. Each task was designed to last roughly 1 hour. Figure 2 shows the development stages of the prototype.

The website prototypes shown in the workshops had been designed and tested to be Web Content Accessibility Guidelines 2.1 compliant [28] as a baseline approach to accessibility. An accessibility discussion was included in the focus group to identify and review the specific accessibility needs participants might experience when using the website. Participants were specifically asked whether they anticipated encountering any barriers to accessing or using the website and which devices they would find most useful for accessing the website. The focus of accessibility discussions in the workshops was to identify specific additional accessibility needs that may have been particularly relevant for this target population.

Aligned with Agile approaches [26], we adopted the MoSCoW (Must have, Should have, Could have, Won’t have) prioritization system [29] to triage requirements. The MoSCoW approach is widely used in software development to determine the most important features and deliverables across mixed stakeholder groups. In the MoSCoW approach, design features of the website were ranked as “must have,” “should have,” “could have,” and “won’t have.” Features that were ranked as “must have” were considered essential by stakeholders; “should have” features were deemed important but would not affect the functionality of the website if they were not included; “could have” features were not necessarily related to the key aims of CaFI:Digital but were seen as additional desirable features; and features classified as “won’t have” were either not priorities for stakeholders or not feasible to deliver within the scope of this study. Workshop facilitators supported discussion when different priorities were identified, aiming to arrive at a consensus decision. Data from the workshops were then analyzed by the project research associate, project technical team lead, and UX designer. Notes were taken during the workshops by the project research associate. Following the workshops, the notes were discussed in detail by the project team and developed into itemized requirements lists for prioritization and implementation in the software. Similar to our approach taken in other co-design projects [30], the technical team analyzed the workshop notes, identifying common requirements and themes in the data. They grouped key themes from the notes after each workshop and identified their technical feasibility. Key requirements were then circulated to the wider research team for review. The technical team and research team then met to review and discuss the workshop insights collectively. The agreed key requirements were then prioritized for technical implementation by the software team. Where decisions were outstanding after the workshops, the research team and technical teams made the final decisions, guided by the necessity to deliver the project on time and within budget.

Usage and engagement analytics were embedded within the platform following the AMUsED (Analyzing and Measuring Usage and Engagement Data) framework for digital health interventions [31]. The main goal was to track analytics within the platform that could help in understanding which elements of content were most visited by participants.

Feedback in the service user and carer workshops (Table 1) focused on the need to ensure that the website was easy to use and simple to navigate. An emphasis was placed on accessibility: making the font size large to support people with visual impairments, making the text easy to read, and providing adequate support and training for people unfamiliar with digital technologies. Diversity in the images portrayed on the website was also requested, particularly in terms of diversity of skin tones depicted.

The main themes from the first workshop with service users and carers (Table 1) related to (1) the accessibility of the website in terms of navigation, disability, digital literacy, and lack of hardware; (2) the content of the website being diverse and representative of people of sub-Saharan African and Caribbean descent; and (3) positive feedback in terms of ease of use of the website and the look and feel of the content. It is important to note that discussions relating to accessibility also raised the possibility of providing audio versions of all materials; however, this was not supported in this project due to time and cost limitations. Each of the feedback items was agreed by a majority within the workshop group as a need or priority for the software.

The main themes from the second workshop with therapists (Table 2) related to (1) the accessibility of the website in terms of navigation, training therapists to use the content, lack of hardware, and a preference from some service users to work with physical materials and (2) the website being visually appealing to therapists.

The main themes from the mixed stakeholder workshop related to (1) the website being visually appealing and (2) accessibility of the website across different devices (ie, phones and laptops).

Feedback in the service user and carer workshops (Table 1) focused on the need to ensure that the website was easy to use and simple to navigate. An emphasis was placed on accessibility: making the font size large to support people with visual impairments, making the text easy to read, and providing adequate support and training for people unfamiliar with digital technologies. Diversity in the images portrayed on the website was also requested, particularly in terms of diversity of skin tones depicted.

Feedback in the therapist workshops included recommendations to prioritize simplicity of design; ensure ease of site navigation; embed training within the site; and ensure supplementary printed materials were provided to participants. The service user and carer workshop identified the positives of using digital technology to supplement therapy delivery, including reporting how digital technology “offered a way to engage with the world” even when people felt very unwell. Poor accessibility of some digital platforms was raised as a concern by service user participants in the workshops, several of whom mentioned the need to ensure the website worked well for those with visual impairments.

Requirements from the workshops were then actioned by the software team or wider research team, as appropriate. The aim was to implement as many requirements as possible within the time and budget available. A small number of items were deemed “out of scope” because of the time and cost implications of implementation, detailed in Tables 1-3.

Participants across all workshops expressed concern about people who had limited or no access to digital technology and/or had poor digital literacy skills, stressing the need for equity in the delivery of digital solutions and therapeutic interventions more widely.

After all the changes from the co-design workshops had been addressed, as described in Tables 1-3, 9 volunteers were asked to test the website on their own devices. The purpose of this phase was to identify any critical or major technical defects with the website that needed to be resolved. Volunteers represented (1) service users and families, (2) therapists, and (3) networks of the research team, including people with therapeutic, clinical, and/or digital backgrounds. A major problem was reported on a specific version of the iPhone, which blocked users from logging in. This was resolved by the software team, and the website was retested on this iPhone version. No other critical or major defects were reported by beta testers. Positive feedback on the beta testing is reported in the table below.

Accessibility was identified as a key area of focus throughout the co-design and development work packages. Additional accessibility needs were identified through the workshops. In the co-design workshops, visual impairment was identified as a common problem among the target population. To respond to this, a larger-than-usual font size was used as the default, and training was included with the trial rollout to advise how to increase the font of the website on common web browsers. Color contrast was also assessed to ensure the website worked well for people with visual impairments, a point raised by service users in Workshop 1 and by the wider research team in team meetings.

To our knowledge, this is one of few studies to co-design a culturally informed digital mental health intervention to address the disparities in psychosis care experienced by people of sub-Saharan African and Caribbean descent. Another example of this is from the Care Pathways therapeutic intervention in America, which, like CaFI:Digital, was co-designed with Black American women and mental health professionals [32]. Unlike CaFI:Digital, the intervention in America was a mindfulness-based application to support anxiety and depression [32]. The CaFI:Digital co-design activities ensured the priorities of the end users of the platform are met. The target population for the CaFI trial has historically been underserved digitally through lack of provision of digital mental health tools as well as in terms of access to evidence-based psychological interventions. CaFI:Digital is an attempt to redress this imbalance. The CaFI:Digital intervention will be evaluated as part of the process evaluation of the randomized controlled trial. Metrics that include login frequency, duration, module completion, engagement with materials, accessibility, and user-friendliness will be used to evaluate usage and engagement.

The CaFI trial offered an opportunity to develop and evaluate a digital mental health intervention that was co-designed. Digitization also provides an opportunity to engage family members in therapy who may live far away from each other. This is particularly relevant for the CaFI trial target population, many of whom have family members located in different locations around the world. Moreover, unlike face-to-face interventions, the CaFI:Digital platform can be readily adapted and scaled for worldwide delivery. This aligns with national and international policy and practice to address ethnically based inequalities, aligning with both NHS priorities in the United Kingdom [33] and global approaches to enhancing delivery of health care using digital tools [34].

The findings of this study are consistent with recent policy shifts in the United Kingdom, notably the NHS Long Term Plan (2019) and NHS England’s Inclusive Digital Healthcare Framework (2023), which advocate for the increased use of digital tools to enhance accessibility and equity in health care delivery [35,36]. Furthermore, CaFI:Digital aligns with NHS inclusive research practice foregrounding cultural competence, stakeholder engagement, and the reduction in health inequalities [37]. The co-design approach used in the platform’s development reflects the NHS commitment to involving diverse, underrepresented communities in research and developing interventions tailored to meet the needs of Black, Mixed heritage, and other racially minoritized populations.

The combination of co-design and Agile software approaches used for CaFI:Digital enabled the software team to identify and respond to requirements and refinements identified by service users, their families and carers, and the therapists, despite the constraints on face-to-face meetings during the COVID-19 pandemic. The use of the MoSCoW prioritization system supported the rapid development cycle required to respond to the pandemic and stakeholder needs. This was particularly helpful when time-consuming or resource-intensive recommendations were identified. A key concern for participants during the development phase of the website was ensuring the website was simple to use and accessible to all, and dedicated workstreams were developed to address specific accessibility concerns raised. The beta test ensured that the software was defect-free and ready to use in the live trial. Inevitably, more suggestions were made at the co-design workshops than there was time or budget to implement. Future versions of the software could seek to address these recommendations (eg, by providing audio versions of all the website materials).

At the time of writing, the CaFI:Digital is being used in a live trial, so it is not possible to report trial outcomes or to present analyzed data on usage and engagement. We are therefore unable to comment on real-world impact and effectiveness. Another limitation is that not all suggested features, such as providing audio versions of all materials, were included because of resource constraints. Finally, although the platform has been co-designed to meet accessibility needs in terms of simplicity, conveying information within the CaFI:Digital platform can also depend on the competency of the therapist. The contribution of the CaFI trial to evidence-based practice will be evaluated as part of the ongoing trial.

The co-design and development phases have sought to make the website as usable and accessible as possible by working with service users, their families, and therapists throughout the co-design process. CaFI:Digital provides access to psychological therapies for people in underrepresented populations. Analysis from the CaFI trial using the engagement analytics embedded within the website combined with qualitative research on user experiences of the platform will enrich our understanding of how participants experienced the CaFI:Digital platform during the trial. Evaluating online group therapy interventions with underserved populations will enhance our understanding of how technology enhances care or perpetuates inequalities for already underserved and “seldom heard” populations.

The CaFI:Digital platform presents unique opportunities to respond to NICE recommended policy guidelines in the United Kingdom [13] by enabling remotely delivered, culturally adapted family therapy. Policy recommendations stemming from this research should prioritize the integration of digital health solutions within national and international health care frameworks, particularly in relation to ethnic minority groups who experience disproportionately high rates of serious mental illnesses such as schizophrenia and related psychoses and systemic health care exclusion.

CaFI:Digital, an evidence-based digital therapy platform, aspires to be at the forefront of efforts to address the gap in access to mental health care for UK Black and other minoritized communities, working toward conforming to the NICE evidence standards framework for health technologies. There is potential for the platform to be rolled out internationally for people of sub-Saharan and Caribbean descent, at scale and at a competitive cost, to provide access to treatment to historically excluded and underserved groups. As with all digital mental health interventions, the potential for CaFI:Digital to widen inequalities through “digital exclusion” (by further restricting access to treatment only to people who have capabilities to use digital tools) must also be considered. It is therefore important to situate CaFI:Digital within a landscape of options for the target population rather than to promote it as a single solution that will be relevant for everyone.

However, CaFI:Digital also has the potential to address longstanding digital inequalities, aligned with 2023 NHS England guidance [35]. The NHS England guidance outlines ways to make digital health care more accessible by maximizing the relatability of the technologies, conducting mixed methods research in the digital space, and actioning user feedback, many of which have been addressed in the development of CaFI:Digital. Moreover, for a population who often have family members located internationally, the remote therapeutic delivery enables the inclusion of family members previously excluded. This is important given the disproportionate rates of diagnosis and lack of access to psychological therapies experienced by this population. Our hope is that CaFI:Digital can begin to address this historic gap in provision of CaFIs for Black populations.