The Impact of Host-Based Early Warning on Disease Outbreaks

Mark Hernandez; Lauren E Milechin; Shakti K Davis; Rich DeLaura; Kajal T Claypool; Albert Swiston

The SEIR Model One of the most common epidemiological models for simulating disease outbreak scenarios is the deterministic SEIR model. This approach splits a given population into separate compartments defined by their relationship to a disease outbreak [15] : scenarios can be simulated by changing the rate parameters of the ODEs linking the population compartments. Furthermore, this approach allows quantitative projections of how many people are exposed to a pathogen and become sick under different outbreak conditions. While the SEIR model is mathematically rigorous and often has good predictive utility during an outbreak, the SEIR model requires several assumptions. First, the model assumes there is a fixed population N, with no births or deaths other than those resulting from the infectious disease. Second, it assumes the population is homogeneously mixing, that is the transmission between any two individuals is equally likely. Third, the model assumes that exposed individuals become infectious after a fixed incubation period, thus not accounting for individual variability of disease progression for the young, elderly, or immunocompromised. Finally, the model assumes that all recovered individuals are immune to further infection and thus do not re-enter the susceptible class. While any of these assumptions may fail to hold in particular contexts, abiding by them allows for greater mathematical tractability and offers similar relative output trends.