Automating Guidelines for Clinical Decision Support: Knowledge Engineering and Implementation.

Area of Science:

• Health Informatics
• Clinical Decision Support Systems
• Knowledge Engineering

Background:

• Clinical decision support (CDS) systems are increasingly utilized in healthcare.
• Accurate translation of clinical practice guidelines (CPG) into computable formats is crucial for effective CDS.
• Existing methods for CPG translation require validation and extension.

Purpose of the Study:

• To validate and extend the 13-step model for translating CPG knowledge into a computable form for CDS.
• To refine the process of integrating complex CPG recommendations into existing clinical dashboards.
• To develop an updated model for translating CPG knowledge into CDS integrated with health information technology.

Main Methods:

• Validation and extension of Shiffman et al.'s 13-step model for CPG translation.
• Encoding complex CPG recommendations for five common chronic conditions during an ATHENA-CDS implementation project.
• Recording and categorizing major decisions made during the implementation process.

Main Results:

• 119 decisions were categorized according to the 13-step model.
• 8 new categories were identified as necessary to complete the CPG translation project.
• An updated model for CPG knowledge translation into CDS was developed.

Conclusions:

• The 13-step model provides a robust framework for CPG to CDS translation.
• The updated model enhances the process of integrating CPG knowledge into health information technology.
• Refined translation methods are essential for advancing the utility of clinical decision support.