Clinical Decision Support System for Diabetes Based on Ontology Reasoning and TOPSIS Analysis

Area of Science:

• Endocrinology
• Medical Informatics
• Pharmacology

Background:

• Existing Clinical Decision Support Systems (CDSS) for diabetes management often overlook patient attitudes towards treatment strategies and HbA1c targets.
• Effective management of type 2 diabetes mellitus (T2DM) requires personalized treatment plans, including appropriate antidiabetic medication selection and glycemic control.

Purpose of the Study:

• To develop and evaluate a CDSS that integrates patient-specific factors for recommending antidiabetic medications and setting HbA1c targets.
• To assess the system's utility and user satisfaction among clinicians managing T2DM patients.

Main Methods:

• Utilized guidelines from the American Diabetes Association (ADA) and European Association for the Study of Diabetes (EASD) for HbA1c targets.
• Employed antidiabetic medication profiles from the American Association of Clinical Endocrinologists (AACE) and American College of Endocrinology (ACE).
• Applied the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) for ranking antidiabetic medications.

Main Results:

• The CDSS demonstrated high performance, with 87% of virtual patients indicating it was useful.
• The antidiabetic medication recommendation system achieved an 85% satisfaction rate among clinicians.
• The system is deemed suitable for outpatient management and assists in selecting appropriate medications for T2DM.

Conclusions:

• The developed CDSS effectively aids clinicians in managing T2DM by providing evidence-based medication recommendations.
• The system can support both specialty and non-specialty physicians, including junior doctors, in prescribing medications.
• Integrating patient attitudes and personalized targets enhances the potential of CDSS in diabetes care.