A two-tiered ambulance system, consisting of advanced and basic life support for emergency and nonemergency patient care, respectively, can provide a cost-efficient emergency medical service. However, such a system requires accurate classification of patient severity to avoid complications. Thus, this study considers a two-tiered ambulance dispatch and redeployment problem in which the average patient severity classification errors are known. This study builds on previous research into the ambulance dispatch and redeployment problem by additionally considering multiple types of patients and ambulances, and patient classification errors. We formulate this dynamic decision-making problem as a semi-Markov decision process and propose a mini-batch monotone-approximate dynamic programming (ADP) algorithm to solve the problem within a reasonable computation time. Computational experiments using realistic system dynamics based on historical data from Seoul reveal that the proposed approach and algorithm reduce the risk level index (RLI) for all patients by an average of 11.2% compared to the greedy policy. In this numerical study, we identify the influence of certain system parameters such as the percentage of advanced-life support units among all ambulances and patient classification errors. A key finding is that an increase in undertriage rates has a greater negative effect on patient RLI than an increase in overtriage rates. The proposed algorithm delivers an efficient two-tiered ambulance management strategy. Furthermore, our findings could provide useful guidelines for practitioners, enabling them to classify patient severity in order to minimize undertriage rates.

Ambulance diversion (AD) is a common method for reducing crowdedness of emergency departments by diverting ambulance-transported patients to a neighboring hospital. In a multi-hospital system, the AD of one hospital increases the neighboring hospital’s congestion. This should be carefully considered for minimizing patients’ tardiness in the entire multi-hospital system. Therefore, this paper proposes a centralized AD policy based on a rolling-horizon optimization framework. It is an iterative methodology for coping with uncertainty, which first solves the centralized optimization model formulated as a mixed-integer linear programming model at each discretized time, and then moves forward for the time interval reflecting the realized uncertainty. Furthermore, the decentralized optimization, decentralized priority, and No-AD models are presented for practical application, which can also show the impact of using the following three factors: centralization, mathematical model, and AD strategy. The numerical experiments conducted based on the historical data of Seoul, South Korea, for 2017, show that the centralized AD policy outperforms the other three policies by 30%, 37%, and 44%, respectively, and that all three factors contribute to reducing patients’ tardiness. The proposed policy yields an efficient centralized AD management strategy, which can improve the local healthcare system with active coordination between hospitals.