Multidisciplinary Development and Initial Validation of a Clinical Knowledge Base on Chronic Respiratory Diseases for...

Area of Science:

• Medical Informatics
• Digital Health
• Respiratory Medicine

Background:

• Current mobile health (mHealth) decision support systems for chronic obstructive respiratory diseases (CORDs) often lack clinical evidence and validation.
• Developing a robust knowledge base is critical for effective mHealth decision support, requiring systematic collection and interpretation of clinical knowledge.

Purpose of the Study:

• To describe the development and initial validation of a clinical knowledge base for mHealth decision support systems targeting patients with CORDs.
• To establish a framework for creating evidence-based mHealth systems applicable to other clinical areas.

Main Methods:

• A multidisciplinary team defined a framework, reviewed scientific sources (e.g., guidelines) for recommendations, and prioritized them based on evidence, scope, and expert perception.
• Selected recommendations were operationalized into logical rules (n=405) with defined variables (n=116) and integrated into an intelligent coaching component.
• Internal validation used observational study data, and external validation assessed high-priority recommendations by independent physicians. A knowledge base update strategy was also defined.

Main Results:

• A total of 358 recommendations were selected and prioritized for the knowledge base.
• The knowledge base was implemented in an intelligent individualized coaching component and pre-tested with an asthma use case.
• Initial internal and external validation demonstrated the appropriateness of the developed knowledge base for CORDs.

Conclusions:

• A consensus-driven, iterative process successfully developed and validated a clinical knowledge base for CORDs, translating guidelines into personalized patient recommendations.
• This validated knowledge base can be integrated into mHealth decision support systems, improving patient self-management.
• The described framework offers a replicable model for developing evidence-based mHealth systems in other clinical domains.