Using GEM-encoded guidelines to generate medical logic modules

Area of Science:

• Computer Science
• Medical Informatics
• Health Services Research

Background:

• Computer-mediated decision support is crucial for improving clinical care.
• Existing knowledge representation models include Guideline Elements Model (GEM) and Arden Syntax for Medical Logic Modules (MLMs).

Purpose of the Study:

• To describe an XML-based application for automated generation of MLMs from GEM-encoded guidelines.
• To facilitate the use of these MLMs in Arden-compliant information systems for clinical decision support.

Main Methods:

• Knowledge extraction from guideline documents using GEM.
• Mapping GEM elements to MLM slots.
• XSL transformation of GEM-encoded guidelines to generate MLMs.

Main Results:

• Successfully generated 15 MLMs from a sample guideline.
• Each MLM corresponded to a conditional or imperative element within the GEM structure.
• Demonstrated automated MLM population from GEM-encoded guidelines.

Conclusions:

• The developed XML-based application automates MLM generation from GEM guidelines.
• This facilitates the integration of clinical guidelines into decision support systems.
• Further development is needed for linking MLMs to represent complex guideline logic.