RAPID: Data-Based Spatiotemporal Models of Ebola Epidemics and Control

Arizona State University, Scottsdale AZ

Investigators

Abstract

DMS-1518529 Kuang The Ebola epidemic in West Africa continues to cause significant morbidity and mortality. The World Health Organization declared the Ebola epidemic in West Africa a Public Health Emergency of International Concern on 8 August 2014. As the epidemic continues to spread at an alarming rate in some areas of West Africa, there is an urgent need to develop, test, and calibrate novel mathematical models of Ebola transmission and control with the goal of generating real-time, science-based forecasts of the epidemic in West Africa in order to quantify the times, locations, type and intensity of interventions that would be required to achieve control. To this end, the investigators and their colleagues formulate and validate models that can be used to predict the number and location of new cases, in order to facilitate decision-making on the allocation of resources that will allow timely medical treatment and isolation of the sick and to implement effective and sensible travel controls. This project provides first-hand educational experience in cross-disciplinary communication and exploration and cutting-edge research opportunities for undergraduates and graduate students. It also provides professional development for graduate students. The investigators disseminate their findings to a diverse range of mathematicians, modelers, and public health and biomedical researchers. The investigators divide the project efforts into three modeling tasks. At first, they formulate dynamical models of the evolving infection reproduction number, comparing how well a simple SI model and the logistic model fit past and current real Ebola epidemics data. In this initial task, instead of treating the whole population as susceptible, they treat the susceptible population size and the final epidemics size as parameters and allow the reported data to inform these parameter values. Subsequently, they present a novel delay differential equation based modeling framework that allows models to treat susceptible population size as a dynamical variable, which is highly correlated to current infectious population size. More specifically, they initialize the susceptible population size as zero instead of the current thinking that susceptible population size equals the total population size initially. This is simply due to the fact that individuals become susceptible to Ebola only if they had a direct and close contact with infectious Ebola patients. In the second half of this project, they formulate and validate partial differential equation-based models in order to estimate the Ebola spread speed and to predict the locations and numbers of new Ebola cases subject to various treatment options and travel control policies. The project is supported by the Division of Mathematical Sciences and the Division of Environmemtal Biology.