Abstract
BACKGROUND: Emerging adults living with type 1 diabetes (T1D) need targeted support to equip them with the knowledge and motivation required for self-management, particularly as they transition from pediatric to adult care. While multicomponent digital health interventions have shown promise in addressing their multifaceted needs, traditional effectiveness studies provide little, if any, insights into which components work effectively, how they function, and for whom. OBJECTIVE: This study aims to explore the implementation of a multicomponent, text message-based digital intervention (Keeping in Touch; KiT) to provide early insights into which components may shape participants' transition experiences and how. The secondary objective was to explore which subgroups, defined by individual characteristics, may benefit most from the intervention. METHODS: Embedded within a broader randomized controlled trial, we conducted a qualitative realist evaluation with intervention-arm participants who had engaged with KiT for a minimum of 3 months. One-on-one semistructured realist interviews were conducted in a teacher-learner cycle to test the initial program theory. The initial program theory included several pathways through which the 5 intervention components (ie, T1D self-management information and suggestions, transition support information, problem-solving support, stress management strategies, and transition reminders) were hypothesized to influence a range of theorized outcomes. RESULTS: A total of 16 interviews were completed with intervention participants. All 5 KiT intervention components were reported to shape participants' transition experiences positively but to varying degrees. T1D self-management information and suggestions presented a universal positive impact across all participants. However, the effectiveness of problem-solving support and stress management strategies varied depending on participants' individual characteristics (eg, duration of diabetes, perceived access to information, and baseline diabetes distress). Rather than acting through parallel independent mechanisms, KiT appeared to support participants' transition experiences via multiple chains of interconnected mechanisms, often beginning with knowledge or reinforcement and contributing to changes in motivation (eg, self-efficacy and diabetes distress). Interview participants described tangible improvement in mechanisms and proximal outcomes (eg, diabetes knowledge and self-efficacy). CONCLUSIONS: A multicomponent, text message-based digital intervention could support emerging adults living with T1D during their transition to adult care by enhancing their knowledge and motivation for self-management. Participant subgroups responded differently to various intervention components, which highlights that one-size-fits-all approaches are likely inadequate. Digital interventions should be developed and studied in a variety of subgroups and contexts to optimize their reach. Interventions for emerging adults living with T1D might benefit from targeting those who are more recently diagnosed with relatively lower baseline levels of diabetes knowledge and self-efficacy or higher levels of diabetes distress. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): RR2-10.2196/46115.