Compartmental Model for the Ebola Virus Disease
The 2014 EVD outbreak reached unprecedented levels in part due to the general public in affected countries being uninformed about the disease. As a result, a large portion of the population did not take necessary precautions to avoid infection and many infected individuals did not seek medical treatment in a timely manner. This behavior exacerbated the spread of the disease because when a highly infectious individual refuses medical treatment, they expose additional uninformed community members, which has a compounding effect. These dynamics are what lead researchers to predict that over 1.4 million cases of EVD would occur by mid-January 2015. However, a massive international eort began in September 2014 that largely focused on educating the general public about key aspects of EVD transmission. This immense response can be credited with limiting new EVD cases and highlights how education and resulting behavior change can have a profound impact on a disease epidemic.
In 2014 I began working with African researchers as part of the MASAMU Advanced Study Insti- tute (MASI). The MASI program aims at connecting African researchers with American researchers to collaborate on meaningful projects. One project we started was a compartmental disease model for the Ebola virus disease (EVD). I later attended a rapid response workshop in Atlanta, Georgia titled "Modeling the Spread and Control of Ebola in West Africa." After learning more about the disease at the workshop, I guided the MASI group in using a dierent model structure in order to more accurately model EVD. The model focuses on educating the general public about the disease as an intervention strategy.
The EVD model distinguishes infected community members from infected individuals in medicalfacilities, with new infections arising from interactions in both settings. The population is further separated by levels of understanding about EVD transmission. A common intervention strategy used by international health organizations to reduce the spread of EVD is to educate individuals about the disease and best practices for avoiding infection. We model this with flow from the uneducated compartments to the educated compartments to represent dissemination of information by the government, health care institutions, word of mouth and the media. Although the model is fairly complex, unique data from two historic outbreaks can be used to estimate most parameters. By reducing my model to a single population and using data from the 1976 Sudan outbreak, I can estimate parameters related to individuals uneducated about EVD. We can then use the reduced model paired with data from the 1979 Sudan outbreak to train parameters related to those who are educated about EVD. The estimated parameters can then be brought back to the full model to view interactions between the two populations and analyze the importance education plays in limiting an EVD outbreak.
This work was published in Infectious Disease Modeling. The full paper is located in my publication list.