Design, Performance Evaluation, and Optimization for Intensive Care Networks Based on Non-Hierarchical Overflow Loss...

Area of Science:

• Healthcare Operations Research
• Network Optimization
• Critical Care Medicine

Background:

• Intensive care unit (ICU) network design often involves prioritizing high-emergency patients while ensuring Quality of Service (QoS) for others.
• Bed reservation policies are commonly used to manage patient flow and prioritize specific classes.
• Optimizing these reservation policies in non-hierarchical ICU models is computationally challenging.

Purpose of the Study:

• To develop a computationally efficient method for optimizing bed reservation thresholds in non-hierarchical ICU networks.
• To improve the accuracy and speed of evaluating Quality of Service (QoS) metrics for patient allocation.
• To reduce the computational burden associated with complex ICU network optimization problems.

Main Methods:

• Application of Information Exchange Surrogate Approximation (IESA) to analytically approximate key QoS metrics.
• Utilization of an integer Particle Swarm Optimization (PSO) algorithm to search for optimal reservation thresholds.
• Validation using real-world ICU data from Hong Kong.

Main Results:

• IESA provides reasonably accurate approximations for QoS metrics.
• The combined IESA and PSO approach accurately identifies optimal reservation thresholds.
• A significant reduction in computation time (over four orders of magnitude) compared to existing methods was achieved.

Conclusions:

• The proposed IESA and PSO approach offers a highly efficient and accurate solution for optimizing ICU bed reservation policies.
• This method is particularly beneficial for large-scale ICU networks, addressing a critical challenge in healthcare operations.
• The findings pave the way for improved resource allocation and patient management in critical care settings.