Diabetes Self-Management in the Digital Health Era: A Concept Analysis Using Natural Language Processing

Diabetes self-management is increasingly shaped by digital health technologies, yet existing conceptual definitions do not fully reflect this evolving landscape. This study aimed to refine the concept of diabetes self-management in the digital health era by applying Rodgers’ evolutionary method integrated with Natural Language Processing (NLP). Ninety-seven studies were analyzed through a dual-phase design combining manual review and computational clustering of sentence-level embeddings. The analysis identified core attributes, including daily behavioral routines, cognitive and psychological processes, multilevel social support, personalization, digital health integration, and dynamic patient–provider relationships. Antecedents encompassed physical and psychological demands, emotional readiness, access to education and support, resource and technology availability, and health and digital literacy. Consequences included improved clinical outcomes, enhanced quality of life, empowerment, and reduced healthcare burden. Integrating manual and NLP-based approaches strengthened conceptual clarity and provided a contemporary framework for understanding diabetes self-management within digitally enabled patient-centered care.