Machine learning based integrated scheduling and rescheduling for elective and emergency patients in the operating...

Area of Science:

• Healthcare Management
• Operations Research
• Artificial Intelligence in Medicine

Background:

• Operating Room (OR) scheduling is a critical factor in hospital revenue and cost management.
• Existing scheduling models often lack integration, leading to inefficiencies.
• Balancing elective and emergency patient needs within ORs presents a significant challenge.

Purpose of the Study:

• To develop a fully integrated model for optimizing OR scheduling.
• To effectively manage both elective and emergency patient surgeries.
• To improve overall OR efficiency and patient throughput.

Main Methods:

• A hierarchical three-phase model (weekly, daily, rescheduling) was proposed.
• Machine learning (random forest) and GIS were used for emergency patient arrival and duration prediction.
• Genetic algorithms and particle swarm optimization were employed to solve the scheduling models.

Main Results:

• The integrated model successfully balanced elective and emergency patient scheduling.
• A reserved capacity system for emergency patients was implemented.
• The three-phase model demonstrated a significant positive impact on OR scheduling efficiency in a real-world case study.

Conclusions:

• The proposed integrated OR scheduling model enhances operational efficiency.
• The model effectively prioritizes emergency patients while optimizing elective surgeries.
• This approach offers a viable solution for improving hospital resource management.