Optimization models for patient allocation during a pandemic influenza outbreak.

Optimization models;Resource education;Patient allocation;Pandemic response

Li Sun

University:University of Louisville

Department:Industrial Engineering

关键词: Optimization models;    Resource education;    Patient allocation;    Pandemic response;   

Others  :  https://ir.library.louisville.edu/cgi/viewcontent.cgi?article=2404&context=etd

美国|英语

来源: The Universite of Louisville's Institutional Repository

【 摘 要 】

Pandemic influenza has been an important public health concern. During the 20th century, three major pandemics of influenza occurred in 1918, 1957, and 1968. The pandemic of 1918 caused 40 to 50 million deaths worldwide and more than 500,000 deaths in the United States. The 1957 pandemic, during a time with much less globalization than now, spread to the U.S. within 4 to 5 months of its origination in China, causing more than 70,000 deaths in the U.S., and the 1968 pandemic spread to the U.S. from Hong Kong within 2 to 3 months, causing 34,000 deaths. Pandemic influenza is considered to be a relatively high probability event, even inevitable by many experts. During a pandemic influenza outbreak, some key preparedness tasks cannot be accomplished by hospitals individually; regional resource allocation, patient redistribution, and use of alternative care sites all require collaboration among hospitals both in planning and in response. The research presented in this dissertation develops optimization models to be used by decision makers (e.g. hospital associations, emergency management agency, etc.) to determine how best to manage medical resources as well as suggest patient allocation among hospitals and alternative healthcare facilities. Both single-objective and multi-objective optimization models are developed to determine the patient allocation and resource allocation among healthcare facilities. The single-objective optimization models are developed to optimize the patient allocation in terms of minimizing the travel distance between patients and healthcare facilities while considering medical resource capacity constraints. During the pandemic, the surge demand most likely would exhaust all the medical resources, at which time the models can help predict the potential resource shortage so an appropriate contingency plan can be developed. If additional resource quantities become available, the models help to determine the best allocation of these resources among healthcare facilities. Various methods are proposed to conduct the sensitivity analysis to help decision makers determine the impact of different level of each type resource on the patient service. The multi-objective optimization model not only considers the objective of minimization of the total travel distance by patients to healthcare facilities, but also considers the minimization of maximum patient travel distance. A case study from Metro Louisville, Kentucky is presented to demonstrate how the models would aid in patient allocation and resource allocation during a pandemic influenza outbreak. A web-based application based on the optimization models developed in this dissertation is presented as an initial tool for decision makers.