Interpretative framework of chronic disease management to guide textual guideline GEM-encoding

Area of Science:

• Medical Informatics
• Artificial Intelligence in Medicine
• Clinical Decision Support

Background:

• Textual clinical guidelines often contain semantic ambiguities.
• Automating the extraction of decision rules from these guidelines is challenging.
• Standardized representation of medical recommendations is needed for computational use.

Purpose of the Study:

• To develop an XML-based application for automated decision rule generation from textual guidelines.
• To formalize treatment steps and standardize decision variables and actions within the Guideline Elements Model (GEM).
• To validate the application using the 1999 Canadian Recommendations for hypertension management.

Main Methods:

• Extended the Guideline Elements Model Document Type Definition (DTD) for standardized representation.
• Formalized guideline-based chronological treatment steps to reduce semantic ambiguity.
• Utilized an XML parser to extract IF-THEN clauses for decision rules from GEM-encoded hypertension guidelines.

Main Results:

• Developed an XML application for automated decision rule generation.
• Successfully encoded the 1999 Canadian Hypertension Recommendations using the extended GEM DTD.
• Generated 104 decision rules, comparable to the 98 rules manually derived in the ASTI project.

Conclusions:

• The developed XML application effectively automates the generation of clinical decision rules from textual guidelines.
• Formalizing guideline elements and extending GEM improves the accuracy and consistency of rule extraction.
• This approach offers a promising method for enhancing clinical decision support systems.