We conducted thematic analysis of all semistructured interviews using an inductive approach that was subsequently mapped to the COM-B framework. Codes emerged organically from participant narratives and were then organized according to COM-B domains (capability, opportunity, motivation). The coding team consisted of AA (ACHD cardiologist with qualitative research training), MJO (pediatric-to-adult transition specialist and qualitative researcher), 1 medical student (K Macholl), and 3 research assistants (KB, JM, PA). All team members completed structured training in the COM-B or TDF framework application and qualitative coding principles, including group practice coding of pilot transcripts to establish shared code definitions. The multicoder consensus process, overseen by AA and MJO, ensured coding rigor while providing methodological training for less experienced team members. Two team members independently coded each transcript in Microsoft Word to identify themes and subthemes related to barriers and facilitators to ACHD care and toward key aspects of the digital tool. After independent coding, coders (AA, K Macholl, KB, JM, PA) met in iterative group sessions to compare coded transcripts and discuss discrepancies. We prioritized resolving discrepancies through consensus discussion among the coding team, rather than quantitative intercoder reliability metrics. The finalized codes were compiled into a single coded transcript for each interview. We then created a structured data matrix using rapid qualitative analysis [43], an action-oriented method for qualitative data analysis conducted in Microsoft Excel, to efficiently produce results for real-world interventions. Coded data from finalized transcripts were transferred into the matrix, which organized codes across participants, allowing systematic cross-case comparison. Sample codebook entries showing how themes were mapped to COM-B and an excerpt of the rapid qualitative analysis matrix structure are provided in Tables S4 and S5 respectively in Multimedia Appendix 1. We summarized the matrix findings via iterative group meetings (AA, K Macholl, KB, JM, PA, MJO). Through ongoing, iterative reviews of CAB meeting discussions documented in study logs, we informed, planned, and refined the intervention components concurrent with their development.

Ethical approval for this study was obtained from the UCSF institutional review board (IRB number 22‐36667). Verbal informed consent was obtained from interview participants. The National Clinical Trial number for this study is NCT06581484. All interview participants and CAB members were offered US $50 per hour in the form of an electronic gift card to compensate them for their time.

We interviewed 54 participants (n=37 patients and n=17 clinicians) (Table 2) and partnered with 21 CAB members. The CAB comprised 11 patients or family members who are racially diverse and from all regions of the United States, 6 pediatric or ACHD physicians, 1 social worker, 1 ACHD nurse, 2 nurse practitioners, and 6 advocacy representatives, some of whom represent more than 1 role. Advocacy organizations included the Adult Congenital Heart Association, Conquering CHD, Mended Hearts, and Team Uncle Joe [44-47].

The challenges uniquely faced by patients with ACHD centered around users’ perspectives, loneliness, feasibility, accessibility, credibility, and scalability (Table 3). They then helped identify design objectives and intervention features that could address these challenges. The interview participants also identified four patient-centered goals to consider during the design of the digital tool (Table S6 in Multimedia Appendix 1): (1) easy access to credible resources (How can we make it easier to connect patients with pre-existing resources?); (2) uplifting of patient voices (How can we uplift patient voices to create a personal connection?); (3) customization to patient needs (How can we cater to different needs within the community?); and (4) centering positivity and joy (How can we bring positivity and joy into our tool?).

Using the BCW framework, we mapped barriers to specialized care (identified through interviews and CAB discussions) to behavioral targets and intervention functions (Table 4). This systematic approach enabled us to select 6 BCW functions (education, training, environmental restructuring, modeling, persuasion, and enablement) and translate them into specific intervention components. For example, educational modules with analogies explaining CHD conditions address capability barriers (lack of knowledge), while peer narratives simultaneously address social opportunity barriers (modeling self-advocacy) and motivation barriers (building confidence and hope). We further mapped these intervention components to intervention objectives to develop a theoretical framework showing how intervention functions are hypothesized to impact outcomes (Figure 2).

The Eureka Platform uses 128-bit secure socket layer encryption for data transmission between the app and servers. Data are stored on Health Insurance Portability and Accountability Act–compliant servers managed by UCSF with restricted access, multifactor authentication, and role-based permissions. Survey responses and engagement metrics are stored as separate comma-separated values files using unique identifiers without personal identifiable information. Only the authorized research team can link the deidentified data to individual participants. Participants can withdraw at any time by contacting the research team. Upon withdrawal, they may request the deletion of their data or allow deidentified data to remain for research purposes. Participants control whether their stories can be shared as Peer Empowerment content (explicit opt-in required). Wearable device data (Fitbit) and patient portal connections require separate participant authorization and are governed by those platforms’ privacy policies in addition to EMCH protections. Study data will be retained per UCSF IRB and NIH data sharing policies, after which the data will be securely destroyed or permanently deidentified.

We describe a theory-based, community co-designed approach to developing EMCH, a digital intervention promoting confidence and skills to enhance engagement in lifelong ACHD care. To our knowledge, this is the first cross-platform (Android and iOS) digital tool specifically designed to address ACHD-specific barriers to lifelong specialized engagement. Our approach integrated evidence-based behavioral frameworks (COM-B or BCW) with extensive qualitative research (n=54) and community-based participatory design to map barriers to engagement and develop targeted intervention components. Our approach ensured that patient and clinician perspectives shaped all design decisions. EMCH enables convenient, passive content access using user-friendly design and behavioral economics (nudge) principles while fostering peer connections and community contributions.

Existing CHD digital interventions focus narrowly on symptom monitoring, exercise support, or transition readiness in young adults and lack theoretical grounding or rigorous evaluation [23,24]. EMCH addresses the full adult lifespan (18+ years) and targets the complete spectrum of barriers to ongoing specialist care for all patients, including those who never transitioned from pediatric care or were lost to follow-up. The intervention’s systematic foundation in behavioral theory (COM-B or BCW) distinguishes it from education-only approaches by addressing psychological capability, social opportunity, and motivation simultaneously. If effective, EMCH could reduce preventable ACHD complications, emergent hospitalizations, and mortality associated with care gaps by promoting sustained engagement with specialist care.

We aim to recruit patients with ACHD at high risk for care gaps (young adults, men, rural residence, lower socioeconomic status, those living far away from CHD centers, etc) and those not previously engaged in research. Mobile phones are nearly ubiquitous—96% of Americans (and 94% of patients with ACHD) in their twenties and thirties own smartphones, with particularly high reliance among young adults, non-Whites, and lower-income Americans [22,49,50]. EMCH cross-platform, low-bandwidth design, and digital health equity framework enable reach across diverse populations. EMCH was designed to augment, not replace, patient-provider interactions by building engagement skills outside clinic visits. Unlike clinic-based navigator programs requiring ongoing personnel costs, EMCH automated content delivery and peer-generated stories provide sustainable engagement with minimal marginal cost per additional user. We selected the Eureka Digital Research Platform for its sustainability (NIH-supported, UCSF-owned), integration of recruitment through intervention delivery, and cost-efficiency in supporting hundreds of concurrent studies.

We selected BCW over alternative frameworks (eg, Intervention mapping or the Consolidated Framework for Implementation Research) [51,52] because it provides flexible, evidence-based guidance for digital behavior change interventions without being overly prescriptive [25,53]. The BCW process enabled transparent documentation of our intervention’s theory of action by systematically mapping behavioral barriers to specific components using shared taxonomic language [32]. Integrating a community-engaged approach with behavioral theory proved essential for moving beyond identifying barriers [30]. While we knew many patients with ACHD lack the awareness of specialist care needs [5], interviews and CAB feedback showed that evidence-based information alone was insufficient. Participants needed peer stories and real-life scenarios to internalize the relevance. This insight led to “Empowerment” and “Peer Empowerment” quotes becoming core components, using modeling to build confidence and demonstrate care’s importance when patients feel well.

The combined qualitative, theory- and community-engaged approach can be resource-intensive but adaptable. Teams with limited resources can supplement qualitative work with rapid stakeholder consultation while maintaining the framework’s core strengths: preventing the reinvention of existing content, simultaneously addressing engagement and implementation barriers, and ensuring intervention components meet users’ actual needs.

EMCH addresses patient-level barriers but cannot overcome system-level challenges (ACHD specialist shortages, insurance gaps, geographic distance from centers). The intervention is most appropriate for stable patients capable of self-management with remote guidance; patients with complex acute needs or severe cognitive impairment require in-person clinical support. While 94% of patients with ACHD own smartphones [22] and EMCH is available on web and native iOS or Android platforms with low-bandwidth design, the most vulnerable patients may face barriers beyond digital intervention scope. However, by reducing routine educational burden for digitally engaged patients, EMCH may enable clinical teams to prioritize in-person resources for those with greater needs. The digital passport currently captures baseline data only; update functionality is under development, and we are exploring electronic health record integration to auto-populate verified clinical data. Although we incorporated diverse perspectives during development, we may not have captured all viewpoints. The EMCH infrastructure supports continuous adaptation, enabling rapid implementation of new components (eg, mental health modules, culturally adapted content). Currently, individuals with intellectual disabilities are excluded due to unique accommodation needs requiring caregiver-focused approaches.

EMCH was launched in September 2024 and has enrolled more than 500 participants. Ongoing process evaluation is assessing acceptability, usability, and preliminary engagement patterns across diverse ACHD populations. We are developing a Spanish-language adaptation to reach the 15% to 20% of patients with ACHD who are primarily Spanish-speaking.

Following process evaluation, we will conduct a randomized controlled trial evaluating EMCH’s effectiveness in completing ACHD specialist visits and improving knowledge, self-efficacy, and social connectedness. If proven effective, EMCH will be scaled to reach patients with ACHD nationally through partnerships with community advocacy organizations and integration into diverse clinical settings.

The theory-based, community co-designed EMCH digital tool equips patients with ACHD with evidence-based resources, peer support, and practical skills to sustain engagement with lifelong specialized care. Ongoing process evaluation will identify the acceptability, usability, and engagement patterns to inform a planned randomized controlled trial. If proven effective in reducing care gaps, EMCH has the potential to reduce preventable ACHD complications, emergent hospitalizations, and mortality—outcomes currently affecting the majority of patients with ACHD. The behavioral frameworks and intervention strategies address universal barriers to care engagement and can be adapted for other chronic conditions requiring lifelong specialized care. The EMCH infrastructure also supports the rapid recruitment of diverse ACHD populations for future research and health policy initiatives.