Hybrid expert-system approach to nurse scheduling

Area of Science:

• Healthcare Management
• Artificial Intelligence in Healthcare
• Nursing Informatics

Background:

• Traditional nurse scheduling is time-consuming and can lead to suboptimal work conditions.
• Existing scheduling methods may not fully incorporate ergonomic principles, impacting nurse performance.
• There is a need for automated, efficient, and high-quality nurse scheduling solutions.

Purpose of the Study:

• To develop NURSE-HELP, a hybrid expert system for flexible and effective nurse scheduling.
• To integrate ergonomics guidelines into the scheduling process to improve nurse performance.
• To combine linear zero-one goal programming with expert systems to optimize run time and schedule quality.

Main Methods:

• Development of the NURSE-HELP hybrid expert system.
• Integration of linear zero-one goal programming and expert system components.
• Comparative analysis of schedules generated by NURSE-HELP versus manual methods by head nurses.

Main Results:

• NURSE-HELP generated schedules in under 20 minutes, compared to 2-4 hours for head nurses.
• The system demonstrated superiority in meeting minimum staff levels, granting day off requests, avoiding backward rotation, and limiting consecutive night shifts.
• Schedule quality was significantly improved across all evaluated criteria.

Conclusions:

• The NURSE-HELP system provides a highly effective and efficient solution for nurse scheduling.
• Integrating expert systems with optimization techniques enhances scheduling performance and quality.
• Adherence to ergonomics guidelines through automated scheduling can positively impact nurse well-being and performance.