From random walks to epidemic spreading: Compartment model with mortality for vector transmitted diseases
Résumé
Epidemic compartment models have become a popular fashion for describing the propagation of an epidemic. To combine compartment models with random walk simulations gives valuable insights into the microscopic mechanisms of the spreading. Here we focus on the propagation of vector-transmitted diseases in complex networks such as Barab\'asi-Albert (BA) and Watts-Strogatz (WS) types. The class of such diseases includes Malaria, Dengue (vectors are mosquitos), Pestilence (vectors are fleas), and many others. There is no direct transmission of the disease among individuals. Individuals are mimicked by independent random walkers and the vectors by the nodes of the network. The walkers and nodes can be either susceptible (S) or infected and infectious (I) representing their states of health. Walkers in compartment I may die from the infection (entering the dead compartment D) whereas infected nodes never die. This assumption is based on the observation that vectors do not fall ill from their infection. A susceptible walker can be infected with a certain probability by visiting an infected node, and a susceptible node by visits of infected walkers. The time spans of infection of walkers and nodes as well as the survival time span of infected walkers are assumed to be independent random variables following specific probability density functions (PDFs). We implement this approach into a multiple random walkers model. We establish macroscopic stochastic evolution equations for the mean-field compartmental population fractions and compare this dynamics with the outcome of the random walk simulations. We obtain explicit expressions for the basic reproduction numbers $R_M, R_0$ with and without mortality, respectively, and prove that $R_M < R_0$. For $R_M,R_0>1$ the healthy state is unstable, and the disease is starting to spread in presence of at least one infected walker or node. For zero mortality, we obtain in explicit form the stable endemic equilibrium which exists for $R_0>1$ and which is independent of the initial conditions. % % The random walk simulations agree well with the mean-field solutions for strongly connected (small world) graph topologies, whereas for weakly connected graph architectures (large world) and for diseases with high mortality the agreement is less well or occurs only after a long observation time. We also investigate the effect of confinement measures on the spreading of the disease. % % Our model has a wide range of interdisciplinary applications beyond epidemic dynamics, for instance in the kinetics of certain chemical reactions, the propagation of contaminants, wood fires, and population dynamics. .
Origine | Fichiers produits par l'(les) auteur(s) |
---|