medicX-KG: a knowledge graph for pharmacists' drug information needs

Area of Science:

• Pharmacology and Health Informatics
• Artificial Intelligence in Healthcare

Background:

• The role of pharmacists is expanding beyond dispensing to comprehensive pharmaceutical care.
• Access to integrated, accurate medicinal product information is crucial for this evolution.
• Knowledge Graphs (KGs) offer a powerful approach to managing complex health data and uncovering relationships.

Purpose of the Study:

• To present medicX-KG, a pharmacist-oriented knowledge graph designed to support clinical and regulatory decision-making.
• To create a semantic layer for the medicX platform, enabling advanced pharmacy services.
• To address the lack of a unified drug information repository for pharmacists.

Main Methods:

• Integration of data from the British National Formulary (BNF), DrugBank, and the Malta Medicines Authority (MMA).
• Development of an ontology and semantic mapping for data integration.
• Informed KG design through interviews with practicing pharmacists.

Main Results:

• medicX-KG successfully integrates data from multiple sources, catering to specific regulatory landscapes (e.g., Malta).
• The KG effectively supports queries related to drug availability, interactions, adverse reactions, and therapeutic classes.
• Demonstrates reduced reliance on fragmented information sources for pharmacists.

Conclusions:

• medicX-KG provides a robust foundation for data-driven pharmaceutical services.
• The knowledge graph enhances pharmacists' ability to access and utilize critical medicinal product information.
• Future work will focus on addressing limitations such as detailed dosage encoding and real-time updates.