A multi-agent system approach for monitoring the prescription of restricted use antibiotics

Area of Science:

• Medical Informatics
• Artificial Intelligence in Healthcare
• Pharmacoeconomics

Background:

• Hospitals utilize expensive antibiotics for restricted use (ARU) reserved for specific conditions.
• Pharmacy departments routinely review ARU prescriptions to identify opportunities for cost savings and reduced patient harm.
• Alternative, less aggressive, and more affordable antibiotics can often achieve similar therapeutic outcomes.

Purpose of the Study:

• To develop and describe a multi-agent system (MAS) for assisting in the review of medical prescriptions involving antibiotics of restricted use.
• To enhance the efficiency and accuracy of identifying cost-effective and therapeutically equivalent antibiotic alternatives.
• To integrate a patient-centered MAS within a distributed hospital information system.

Main Methods:

• A multi-agent system approach was designed, with individual agents assigned to each patient to monitor their therapy.
• A dedicated pharmacy agent analyzes patient-specific data and prescribed ARU to suggest alternative treatments.
• The system architecture was specified using the Electronic Institutions design methodology within a distributed hospital environment.

Main Results:

• The proposed MAS facilitates the systematic review of restricted antibiotic prescriptions.
• The system enables the identification of potential substitutions with cheaper and less aggressive antibiotic options.
• The patient-centered agent design allows for personalized therapy assessment within a complex hospital IT infrastructure.

Conclusions:

• The developed multi-agent system offers a valuable tool for hospital pharmacies to optimize the use of restricted antibiotics.
• Implementing this system can lead to significant cost reductions and improved patient safety by promoting the use of appropriate and affordable treatments.
• The Electronic Institutions framework provides a robust method for specifying complex, distributed agent-based systems in healthcare settings.