 Management of resource sharing in emergency response using data-driven analyticsJifan Zhang,
Salih Tutun,
Samira Fazel Anvaryazdi,
Mohammadhossein Amini,
Durai Sundaramoorthi () and
Hema Sundaramoorthi
Annals of Operations Research, 2024, vol. 339, issue 1, No 25, 663-692 Abstract: Abstract This research models a medical resource-sharing problem with three powerful techniques in analytics: mixed-integer linear programming (MILP), deep q-learning (DQL), and double deep q-learning (DDQL). These models can be utilized by a group of collaborating entities in emergency situations to decide how they share resources and mitigate adverse effects on human lives. Due to the sudden spikes in demand, the allocation of healthcare resources becomes dynamic in nature. Medical resource management underwent an unprecedented level of difficulty in the modern history of our healthcare system because of the Coronavirus pandemic. At the beginning of the pandemic, we witnessed a severe shortage of ventilators in several countries including in the U.S. Followed by that, there was a severe shortage of Oxygen supply in treating patients who tested positive for the virus in India. Many lives were unnecessarily lost because of the lack of a resource-sharing framework to use in an adverse situation. When a hospital or region experiences sudden spikes in demand for resources, the key decision involves sending resources from other places. Especially, a swift optimal decision has to be made regarding who will send the resources and when will it be sent to the places of need. To tackle this problem, we developed three analytics models—MILP, DQL, and DDQL—based on the Institute for Health Metrics and Evaluation (IHME)’s public data of ventilator allocation. To the best of our knowledge, DQL and DDQL are introduced in this context for the first time. Our models can be used either to share critical resources between collaborators or transfer patients between the locations of the collaborators. We devise two policies for DQL and DDQL models while optimizing the parameters in them. We call these policies the normal policy and the just ship policy. The just ship policy promptly ships healthcare resources to save patients in other locations even at the risk of not meeting its own demand in the future. The normal policy allows saving resources at a location so that future demands are met at the location rather than satisfying the current demand at other locations. To demonstrate the performance of our models, we applied our models to the IHME’s ventilator data and present the results for all fifty states in the U.S. We show how these models optimize ventilator allocations assuming that all the fifty states collaborate as a task force and implement the solution that emerges from the analytics. The experimental results show that the proposed models are effective in handling the resource sharing problem presented by Covid-19. This research provides a decision-making framework for resource allocation during any disease outbreak, such as, new variants of coronavirus and other outbreaks like monkey-pox. Keywords: Resource sharing; Deep learning; Artificial intelligence; Reinforcement learning; Healthcare management (search for similar items in EconPapers) Downloads: (external link) Related works: Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text Persistent link: https://EconPapers.repec.org/RePEc:spr:annopr:v:339:y:2024:i:1:d:10.1007_s10479-023-05702-x Ordering information: This journal article can be ordered from DOI: 10.1007/s10479-023-05702-x Access Statistics for this article Annals of Operations Research is currently edited by Endre Boros More articles in Annals of Operations Research from Springer |
Is your work missing from RePEc? Here is how to contribute.
Questions or problems? Check the EconPapers FAQ or send mail to .
EconPapers is hosted by the
School of Business at Örebro University.