Global dynamics of heterogeneous epidemic models with exponential and nonexponential latent period distributions (Q6575469)

Of concern in this paper is the impact of different latency distributions upon the long-term dynamics of an infectious disease modeled via a multi-group \(SEIHR\) model. For both the exponential and gamma distributions, the basic reproduction number is derived via the next generation method as a spectral radius of a certain matrix, although its exact value is found to be computationally intractable, upper and lower bounds involving isolated and average reproduction numbers defined \textit{ad hoc}, respectively, being provided instead. The global asymptotic stability of the disease-free and of the endemic equilibria, respectively, are then established via M.Y. Li's graph theoretic approach towards constructing Lyapunov functionals.\N\NIt is then determined numerically that the gamma distribution may lead to large epidemic peak sizes and longer peak times than the exponential distribution, although the impact is smaller than in the corresponding short-time models. Also, the effect on the cumulative infected size is negligible, leading to the idea that in long-term epidemics dynamics the choice of a latency distribution does not significantly alter the quantitative outcome. For the sake of simplicity, it then suffices to use the exponential distribution to model latency.