Collaborative Research: Patient Triage in the Aftermath of a Mass Casualty Event - A Dynamic Programming Approach

本研究项目的目标是使用运筹学工具，特别是随机动态规划，更好地理解大规模伤亡事件后患者的分流和优先决策。此类事件(如重大事故、恐怖爆炸或自然灾害)可能会造成大量人员受伤，导致对手术室、救护车、X光机等医疗资源的需求突然跃升。因此，这些医疗资源变得严重不堪重负，因此它们的有效使用对整个应急工作的成功至关重要。一个基本的问题是，这些资源应该如何分配给需要治疗的病人，以便为最多的人做最大的好事。在这项研究中，我们将建立数学模型，以确定有效的患者分流和优先决策的关键特征，这将有助于制定可在现场使用的更好的政策。我们会特别注重在知情的情况下作出优先次序的决定，并考虑到不同类别、不同健康状况和治疗需要的病人数目等因素。在第一阶段，我们将研究最常用的分类程序，即在响应过程中只对患者进行一次分类。使用动态规划公式，将获得危重患者的分析性特征，并将制定考虑患者组合的近乎最佳的优先排序策略。该项目的第二阶段将是动态分诊程序，根据该程序，在整个应对工作中定期对患者进行重新评估。将调查动态分流的潜在好处，并为其实施提出良好的政策。最后，在第三阶段，将进行广泛的模拟研究，以在更现实的环境中检验在项目的前两个阶段提出的基本原则和政策。

The objective of this research project is to use Operations Research tools, in particular stochastic dynamic programming, to develop a better understanding of patient triage and prioritization decisions in the aftermath of mass casualty incidents. Such incidents (e.g., a major accident, a terrorist bombing, or a natural disaster) may cause a significant number of injuries creating a sudden jump in demand for medical resources such as operating rooms, ambulances, X-ray machines, etc. As a result, these medical resources become seriously overwhelmed, and thus their effective use becomes extremely important for the success of the entire emergency response effort. One basic question is how these resources should be allocated to the patients in need of treatment so as to do the greatest good for the greatest number. In this research, we will build mathematical models to identify key characteristics of effective patient triage and prioritization decisions that will help develop better policies that can be used on the field. We will particularly focus on giving informed prioritization decisions that take into account various factors such as the number of patients in different classes with different health conditions and treatment requirements.The project will consist of three main stages. In the first stage, we will study the most commonly applied triage procedure that classifies patients only once during the response effort. Using a dynamic programming formulation, an analytical characterization of the critical patient will be obtained and near-optimal prioritization policies that consider the patient mix will be developed. The second stage of the project will be on dynamic triage procedures under which patients are reevaluated periodically throughout the response effort. The potential benefits of dynamic triage will be investigated and good policies for its implementation will be proposed. Finally, in the third stage, an extensive simulation study will be carried out to test the basic principles and policies that are proposed in the first two stages of the project, in more realistic settings.