A framework for automated conflict detection and resolution in medical guidelines

Area of Science:

• Medical Informatics
• Computational Medicine
• Health Systems Research

Background:

• Patients with multiple chronic conditions (multimorbidity) often face conflicting treatment recommendations from various clinical guidelines.
• Simultaneous application of multiple guidelines can lead to adverse effects due to inconsistent or incomplete medication management.
• Existing approaches lack automated methods for conflict detection and resolution in multimorbidity treatment plans.

Purpose of the Study:

• To present an automated formal framework for detecting, highlighting, and resolving treatment conflicts in multimorbidity, with a focus on medications.
• To develop a system that visualizes conflicts and suggests alternative treatments based on formal models of clinical guidelines.
• To optimize conflict resolution by considering medication efficacy, severity of adverse effects, and temporal overlaps.

Main Methods:

• Guidelines are captured in a standard modeling language and transformed into formal models.
• A theorem prover and constraint solver are used to check guideline correctness and identify inconsistencies.
• An optimizing constraint solver searches for optimal solutions to resolve or minimize conflicts, considering medication efficacy and severity.

Main Results:

• The framework successfully detects and highlights conflicts within combined clinical guidelines for multimorbidity.
• It provides visualized conflict information and suggests alternative, optimized treatment plans.
• The use of an optimizing constraint solver allows for the prioritization of solutions based on clinical relevance and patient safety.

Conclusions:

• The developed automated framework offers a robust solution for managing complex medication regimens in patients with multimorbidity.
• This approach enhances patient safety by proactively identifying and resolving potential treatment conflicts.
• The system's ability to suggest optimized alternatives represents a significant advancement in personalized and safe multimorbidity care.