A multi-layered framework for disseminating knowledge for computer-based decision support.

Area of Science:

• Health Informatics
• Clinical Decision Support Systems
• Knowledge Representation

Background:

• Implementing clinical guidelines across diverse healthcare settings faces challenges due to variations in clinical decision support (CDS) tools, patient data, and workflows.
• Heterogeneity in these systems hinders the seamless integration and portability of guideline knowledge.

Purpose of the Study:

• To develop a novel multi-layered knowledge representation framework for structuring guideline recommendations.
• To enhance the portability and implementability of guideline knowledge within various CDS contexts.

Main Methods:

• A four-layered framework was created to progressively structure narrative guideline recommendations into CDS system inputs.
• The framework was applied to implement rules for a CDS service using three distinct clinical guidelines.
• A preliminary evaluation involved CDS experts from four institutions assessing the implementability of six guideline recommendations.

Main Results:

• The developed framework facilitates the transformation of narrative guidelines into structured, implementable knowledge for CDS.
• Preliminary expert evaluations suggest the framework's potential for cross-organizational guideline implementation.
• The multi-layered approach addresses the heterogeneity challenges in CDS integration.

Conclusions:

• The multi-layered knowledge representation framework shows promise for creating portable and usable structured knowledge for CDS systems.
• This approach can improve the consistency and efficiency of guideline implementation across different clinical sites.
• Further validation is recommended to confirm the framework's broad applicability and impact on clinical practice.