Centralized Ambulance Diversion Policy Using Rolling-Horizon Optimization Framework to Minimize Patient Tardiness

Area of Science:

• Healthcare Management
• Operations Research
• Public Health

Background:

• Ambulance diversion (AD) is a strategy to manage emergency department (ED) overcrowding by redirecting patients.
• In multi-hospital systems, AD at one facility can exacerbate congestion at neighboring hospitals.
• Minimizing patient tardiness across an entire system requires careful consideration of AD impacts.

Purpose of the Study:

• To propose a centralized ambulance diversion policy using a rolling-horizon optimization framework.
• To evaluate the impact of centralization, mathematical modeling, and AD strategy on patient tardiness.
• To provide an efficient management strategy for coordinated hospital operations.

Main Methods:

• Developed a centralized optimization model using a mixed-integer linear programming approach.
• Implemented a rolling-horizon methodology to iteratively address uncertainty.
• Compared the proposed centralized AD policy against decentralized optimization, decentralized priority, and No-AD models.

Main Results:

• The centralized AD policy significantly outperformed other models, reducing patient tardiness by 30%, 37%, and 44%.
• Centralization, mathematical modeling, and the AD strategy were all identified as contributing factors to reduced patient delays.
• Numerical experiments based on Seoul, South Korea data from 2017 validated the proposed policy's effectiveness.

Conclusions:

• The proposed centralized AD policy offers an efficient strategy for managing ambulance diversions in multi-hospital systems.
• Active coordination between hospitals is crucial for improving local healthcare system performance.
• The study demonstrates the benefits of integrated optimization for emergency medical services.