Young people were recruited from Child and Adolescent Mental Health Services (CAMHS) and from the Cardiff Early Prediction of Adolescent Depression (EPAD) study [23]. They were required to be at least 13 years of age and to have either current or past history of depression (recruited from CAMHS), or be at high risk of depression because of a family history (at least one parent with recurrent depression, recruited from the EPAD study).

Parents, guardians and carers of these young people were also recruited for interviews and focus groups. Professionals (from health, education, social, and youth services and charities) were approached if they worked with young people who present with mental health difficulties. Participants were not eligible if they had severe health difficulties that would make it difficult for them to contribute, or make it more likely they would become distressed, or if they were unable to understand the intervention or questions or discussions (eg, because of insufficient understanding of English).

Participants were recruited from south and west Wales and included areas of high deprivation, as well as more affluent areas. Within the interviews and focus groups, we aimed to have a balance of participants with regard to age, gender, primary language (Welsh or English), severity of depressive symptoms, and service involvement.

Participants from CAMHS and the EPAD study were sent invitation letters and information sheets (Figure 2) and asked to reply if interested in participating. Young people were consulted during the design of information sheets and consent and assent forms. These were developed with reference to a readability test [24] and with the help of an experienced graphic designer so that they could be engaging for young people. The documents and the program were developed in Welsh and English in line with depression guidelines [5] and to ensure inclusivity for the population of Wales [25]. Participants were offered gift vouchers for their participation, time, and travel.

All interview and group participants described how there was a need for this program, especially given the lack of specialist CAMHS and Web-based resources. Parents, carers and professionals also felt there was not enough time during CAMHS clinic sessions, and the waiting lists were long, as illustrated in the following quotes:

Participants across the interviews and groups noted that using digital technologies was a valid approach to engagement, as young people use these in everyday life, although there were concerns about those without internet access:

Participants in the interviews and groups stated the program could increase awareness and understanding of depression in young people, promote self-management, address the diversity in the target group, and help those who find it difficult to talk to others:

It was clear that there was a need to cater for parents, carers, and others concerned about a young person:

The overall design was discussed the most in the interviews and groups, especially by young people. This included discussions on how multimedia should engage users and communicate information:

Participants suggested that the tone, language, and terminology should be at the level of the young person and discussed how much text would be appropriate. They also advised there should be clear structure and navigation:

Most recommended tailoring or personalizing the program, for example, by using a log-in, tailoring or modifying the content according to the needs of the user, allowing the user to save or upload information, and to set goals:

Interview participants suggested that the program should be multiplatform, and there should be a mood monitor and an app. Group participants agreed and recommended developing other interactive elements (eg, to set goals and to save links to helpful resources). The introduction of designs for the program, including elements such as illustrations, characters, metaphors, moving images, and audio, helped to guide group discussions:

Security and confidentiality were key issues in the interviews and groups, especially for young people. There were also concerns about including a forum or links with social media:

Participants in the interviews stated there should be different levels of information and a clear explanation of the program and its aims. There should be information on mood, the signs, symptoms and effects of depression, the difference with “normal sadness,” and related issues such as anxiety and accounts of personal experiences. Many young people and parents wished to highlight to adolescents that they are “not alone” in their experiences of depression:

Possible reasons for depression were discussed, including environmental and biological factors and also how there are sometimes no clear reasons for feeling low. Participants felt that self-management approaches should build on existing strengths and attributes:

A personalized tool kit was suggested to include short-term measures dealing with stress and healthy living:

Professionals suggested that the young person should take control and think about issues from an outsider’s perspective:

Participants discussed where to get help for depression, including from trusted friends and family, from school and health services, and the different possible treatments:

They felt that there should also be information for parents or carers and others, and professionals suggested there was a need for reframing:

Focus group participants agreed there should be levels of information, with a hierarchy of sections and subsections. The overall structure and content evolved with each group and the program prototype included sections on mood and depression, possible reasons, self-help, and where to get help. Although there were differing views, overall, it was felt there should be a separate section on “other health issues” such as anxiety, bipolar disorder, and physical illness, although there was caution:

To help with adherence and encourage support, participants in the interviews and groups suggested it would be helpful for the young person to be able to use the program with others, including family, carers, friends, or professionals, as well as independently. The creation of a separate user pathway and section for parents, carers, friends, and professionals was supported by all focus groups—with information on signs, symptoms, and triggers; supporting the young person; and dealing with difficulties.

With regard to implementation, schools were considered an important setting for the intervention, particularly personal, social, and health education sessions. It was also felt that the program could be integrated into health and other services:

The name and promotion of the program was discussed, and the use of the term “mood” was considered more acceptable than “well-being” to young people.

A key issue relating to the program was that it should be person-centered and follow guidelines for the development of digital health-related interventions [20-22]. The framework by Mohr et al [21] describes considerations for the theoretical approaches (aims [“why”] and conceptual or behavior change strategies [“how”]), as well as the instantiation and technical approaches (elements [“why”], characteristics [“how”], and workflow [“when”]). Another approach followed was the Persuasive System Design [22] that identifies four general design features: (1) primary task support, including tailoring or personalization and self-monitoring; (2) dialogue support, including positive reinforcement and suggestions; (3) credibility, by conveying trustworthiness and expertise, in the general tone and references to evidence and guidelines, supporting institutions, and user input in the development process; and (4) social support, by encouraging use and discussions with a trusted person and providing links to other sources of help.

The program was personalized where possible, so that engagement was a two-way process, whereby the participants could take ownership of the program rather than be a passive recipient of information [28,29]. Diversity and learning preferences are key considerations in education, and the program was developed to be accessible and inclusive. A range of media helped to engage and adapt to the user’s learning style or preference, and key themes and messages were repeated in various ways [30]. This linked in with the VARK approach, which states that individuals prefer to learn through different sensory modalities: visual, auditory, read or write, and kinaesthetic.

Other theories and approaches that informed the development of the program included a blended learning approach, Bloom’s taxonomy, Kolb’s experiential learning theory, the constructivist approach, and approaches to help with concentration, such as segmenting sections and interactive elements [31-34].

Information on talking or psychological therapies for depression was presented in the help sections. Elements of CBT were suggested by professionals, possibly because many practiced this and because this approach is recommended in guidelines for treating depression [5]. CBT theory was referred to in the program when discussing possible symptoms, effects, reasons, negative thoughts, and metaphors related to mood, depression, and other issues. The goal-setting component and self-help subsections used elements of behavioral activation theory [35].

Aspects of positive psychology were incorporated in the general tone of the program, particularly where the user was encouraged to think of their “strengths and positives,” and how these can help them to overcome difficulties. Interpersonal and family systems theories were referenced in help sections, encouraging the user to consider their relationship or roles with friends and family and those advising others to support the young person.

The development and content of the intervention was influenced by several behavior change theories including information, motivation, behavior theory, self-regulation theory, self-determination theory, and social cognitive theory [36-38].

This manuscript describes the collaborative design and development of a Web-based psychoeducation program for adolescent depression (MoodHwb, or HwbHwyliau in Welsh). This program could help young people with (or at risk of) depression, as well as their families, carers, friends, and professionals. It has the potential to be integrated into a range of services including health, education, social, and youth services and charities.

MoodHwb meets the need outlined in the guidelines for depression in young people [5,6] for good information and evidence-based psychosocial interventions for the young person and family or carer. The review conducted as part of this work on PIs for adolescent depression [13] concluded that there were few existing programs, particularly Web-based, that have been developed and evaluated using rigorous methods. This program will help to fill this important gap.

The program also fits with the UK government’s push to provide access to Web-based therapies to young people through information and communication technology [39-42]. Over 90% of young people aged 16 to 24 years have internet access in their homes [43], and the use of technological devices (such as computers and smartphones) is similar in individuals with mental health disorders compared with the general population [44]. eHealth approaches offer a valuable opportunity to provide a novel intervention widely and at low cost compared with face-to-face therapies. Web-based interventions also address issues of accessibility, waiting lists, and treatment flexibility.

The findings from the development work suggested the program might be particularly helpful in the early stages of depression when a young person starts to experience difficulties. This could be when they first present to professionals in education, health, or other sectors. MoodHwb could be integrated into each of these services at the lower levels of the stepped care approach for depression in young people [5]. It could also complement (and be an adjunct to) other approaches, for example, in the management of more severe or chronic difficulties.

MoodHwb fits with the guided self-help approach as it has been designed so that it can be used either independently or with another person. PIs could be delivered by a range of professionals, and Colom [9], a pioneer in psychoeducation for mood disorders, states that facilitators need to be “an expert on the disorder not the technique,” avoiding the “complex training” and associated funding required, for example, for CBT. Web-based PIs could therefore help in areas where there is a lack of skilled alternative approaches, including in low and middle-income countries [45].

Historically, pressure to identify effective interventions in public health has led to many interventions being tested for effectiveness in randomized controlled trials (RCTs), with little rigorous development work being completed [46]. This has led to large amounts of resources being invested in interventions that are unlikely to be effective. However, frameworks for the development and evaluation of complex interventions [19] stress the importance of the development phase, which is at least as important as the effectiveness evaluation stage. Proper development involving users, relevant theory, and research evidence is more likely to produce an intervention that is acceptable and feasible to deliver and also potentially effective. This process also facilitates identification of potential problems that may be preventable at a later stage (eg, in a large trial).

In the world of digital health, this development phase is also crucial, and user-centered design is a common approach used [20-22]. As more digital health apps and interventions are being developed, one issue that is important to consider is whether these interventions should be tested in RCTs or whether other evaluation methods should be used. Given the pace of development of digital health, evaluation methods are needed that are more agile and more user-centered than the standard RCT methodology, which may take too long to produce the results needed. However, when feasible, RCTs represent the most internally valid means of establishing the effectiveness of complex public health interventions [47]. In the MRC guidance, emphasis is placed on conducting randomized trials of such interventions [19]. The staged, user-centered approach to development taken here is recommended regardless of whether the effectiveness is evaluated in an RCT or using other methods of evaluation.

There is a lack of literature not only on the overall development process of Web-based (PI) programs but also specific elements such as the approach to graphic design or illustration in the context of youth mental health. This paper gives a detailed account of the process of development and a practical example of how to follow existing guidelines and could help inform other studies and programs in this field. The paper stresses the importance of involving all potential users (young people and their families, carers and professionals) in the development of all aspects of the program—not only the content but also the design and technical elements and why this is crucial especially when attempting to engage young people with mental health issues. The underlying theories associated with all these aspects were considered (not only the psychological theories), and this influenced the development of the logic model. The researchers were also involved in all stages and in regular discussion with the multimedia company and other agencies.

One of the main strengths of the project relates to the rigorous methodology, which followed the MRC guidance for complex interventions and the consultative and person-based approach [20]. Interviews and focus groups progressed in an iterative fashion so that participants could guide the development. The interview and group transcripts were all analyzed and double coded to ensure reliability of coding.

Another strength was the diverse range of participants, including young people from mental health services (primary and secondary care) and volunteers, many of their parents and carers, and a mix of professionals. Young people at high risk were also recruited from a previous study sample. There were a range of recruitment centers in urban and rural areas, including in areas with several ethnicities and areas of deprivation. Participants were also recruited from communities where Welsh was spoken widely.

The results from the review, interviews, and focus groups were complemented by other investigations, including advice from experts, visits to centers of excellence in youth mental health and eHealth in the United Kingdom and overseas, and workshops with multimedia professionals. The development was further strengthened by using user-centered design, educational and psychological approaches, and a logic model, as well as basing the content and design on current research evidence and guidelines. All these approaches helped to create a potentially accessible, engaging, and informative program that the user could personalize, which fits with the increasing interest in precision (personalized) medicine [56]. The detailed account of the development process could inform the development of other programs in the future.

The research or multimedia teams were involved throughout the project, and it is possible that their personal views could have influenced the process of development. However, a range of different groups was consulted, which would help minimize bias. Certain decisions regarding development were made before the interviews and groups (eg, that this would be a Web-based resource targeting adolescent depression), and the key themes that emerged may have been influenced by the topic or discussion guide. However, it was necessary to have some focus to the discussions, and this was not intended as an open scoping exercise. All prior decisions were based on a literature review and expert clinical opinion.

As with all methods of data collection, there are limitations to the interview and focus group approach. Although efforts were made to recruit a diverse sample with different backgrounds and experiences of depression, the overall number of participants was small and may not be representative of the population of adolescents who are experiencing depression or who are at risk. In addition, it is possible there was self-selection, in that participants who volunteered were more likely to have an interest in the project. Although the number of initial interviewees (n=12) was small, many ideas and themes were repeated as the interviews progressed, indicating that saturation might have been reached. The participants were purposively chosen at this early stage of the project to offer a range of viewpoints or ideas to guide later focus groups.

Focus groups were balanced as far as possible with regard to age, gender, and experience of mental health difficulties. However, the number of those who attended the focus groups varied across the study. Recruitment for young person groups 2 and 3, where the majority had experience of mental health difficulties, was difficult. Some participants cancelled at short notice or did not attend, although this is not unusual for focus groups [57,58]. In addition, three fathers had stated they would attend the groups but did not, leading to an all-female parent and carer group. This is consistent with other depression studies involving families, where most participants were mothers [23].