The Ebola virus outbreak in West Africa is the largest Ebola outbreak in history and has captured global attention as the disease continues to spread. The current outbreak is the first Ebola epidemic the world has ever known, according to the U.S. Centers for Disease Control and Prevention (CDC), and is currently affecting the countries of Guinea, Liberia, Nigeria, and Sierra Leone.

The disease is severe and often fatal, and there is no vaccine proven to be effective against the virus. The numbers are constantly being revised, but as of Sept. 19, health officials have counted 5,864 Ebola cases with 2,811 deaths in the current epidemic.

On Sept. 23, the CDC released an article stating that without additional interventions or changes in community behavior, by January 20, 2015, there will be a total of approximately 550,000 Ebola cases in Liberia and Sierra Leone or 1.4 million if corrections for underreporting are made. A recent article by the World Health Organization (WHO) Response Team also concludes “that without drastic improvements in control measures, the numbers of cases of and deaths from Ebola virus disease are expected to continue increasing from hundreds to thousands per week in the coming months.”

The outbreak is being studied by Grant Brown, a fourth-year graduate student from Clive, Iowa, who is seeking his doctoral degree in biostatistics form the UI College of Public Health. Working with Jacob Oleson, associate professor of biostatistics, Brown is developing open source software called libspatialSEIR to model and predict the spread of the Ebola epidemic in the region.

“This is a living document, as the Ebola epidemic is rapidly evolving,” Brown says of the project, which includes interactive maps illustrating the estimated and predicted spread of the virus over time. The document will continue to be updated as the epidemic progresses, reflecting new data and potentially additional analysis techniques.

Brown and Oleson described the project in more detail:

libspatialSEIR is an evolving open source software library that provides tools designed to fit a class of epidemic models known as spatial SEIRS models. The acronym SEIRS stands for Susceptible-Exposed-Infectious-Removed-Susceptible, and describes the assumed disease progression experienced by members of a population. The disease state categories, or compartments, are defined as follows:

• Susceptible: capable of contracting a pathogen
• Exposed: infected by a pathogen, but not yet infectious
• Infectious: capable of spreading a pathogen to others
• Removed: recovered or perished, not capable of re-contracting the disease

The SEIRS framework not only allows researchers to model diseases which have a latent period, but can accommodate pathogens that confer only temporary immunity, such as influenza.

Our work within this family of epidemic modeling techniques estimates relationships to, and provides future prediction of, disease behavior over both time and space allowing for interactions between and within discrete spatial locations (e.g., counties, nations, or other geographic and political areas).

In addition to the novel statistical and computational methodology underlying the specification and generalization of the spatial SEIRS statistical model family, two specific guiding development goals of libspatialSEIR are computational efficiency and ease of use. The first is of critical importance as the computational burden of fitting spatial epidemic models increases rapidly with the number of locations and time points included, while the second aims to increase the libraries’ usefulness to researchers.

The 2014 outbreak of Ebola in West Africa has provided a dramatic, though tragic, opportunity to develop these methods. The libspatialSEIR software was used to fit relatively simple spatial SEIR models with incidence data released by the World Health Organization and U.S. Centers for Disease Control and Prevention, among others, and compiled on Wikipedia.

Preliminary results include estimates of the current and future number of infectious individuals in the countries of Guinea, Liberia, Sierra Leone, and Nigeria; estimates of the basic reproductive number over time in each country; and interactive maps of the current new case incidence and predicted epidemic size.  Updates are released regularly at the following link: http://grantbrown.github.io/libspatialSEIR/doc/tutorials/Ebola2014/Ebola2014.html