Two-stage multi-objective optimization for ICU bed allocation under multiple sources of uncertainty.

Area of Science:

• Healthcare Management
• Operations Research
• Medical Informatics

Background:

• The COVID-19 pandemic overwhelmed intensive care units (ICUs), leading to postponements in elective patient care.
• Effective resource allocation in ICUs is critical, especially when managing diverse patient needs under uncertainty.

Purpose of the Study:

• To develop and evaluate a two-stage model for optimizing intensive care unit (ICU) bed allocation.
• To address uncertainties in patient numbers and lengths of stay (LOS) for emergency, elective, and current ICU patients.
• To minimize the number of required ICU beds and maximize resource utilization while ensuring maximum patient admission.

Main Methods:

• A two-stage optimization model was formulated to handle uncertainties in patient flow and LOS.
• An improved Non-dominated Sorting Genetic Algorithm II (NSGA-II) was employed for ICU bed allocation.
• The NSGA-II algorithm was benchmarked against multi-objective simulated annealing (MOSA) and multi-objective Tabu search (MOTS) using real-world hospital data.

Main Results:

• The improved NSGA-II demonstrated superior performance in ICU bed allocation compared to MOSA and MOTS.
• NSGA-II achieved significant savings in ICU beds, outperforming MOSA by 9.8% and MOTS by 5.1%.
• Across five different scenarios, NSGA-II showed average improvements of 0% to 49% across five key objectives.

Conclusions:

• The proposed two-stage model and improved NSGA-II offer an effective solution for optimizing ICU bed allocation.
• This approach enhances resource utilization and patient throughput, crucial for managing healthcare capacity during high-demand periods.
• The findings provide valuable insights for hospital administrators seeking to improve ICU operational efficiency and patient care delivery.