MODELING SEASONALITY IN AVIAN INFLUENZA H5N1

摘要

The number of cases of H5N1 avian influenza in birds and humans exhibit seasonality which peaks during the winter months. What causes the seasonality in H5N1 cases is still being investigated. This article addresses the question of modeling the periodicity in cumulative number of human cases of H5N1. Three potential drivers of influenza seasonality are investigated: (1) seasonality in bird-to-bird transmission; (2) seasonality caused by wild bird migration or seasonal fluctuation of avian influenza in wild birds; (3) seasonality caused by environmental transmission. A framework of seven models is composed. The seven models involve these three mechanisms and combinations of the mechanisms. Each of the models in the framework is fitted to the cumulative number of humans cases of H5N1. The corrected akaike information criterion (AICc) is used to compare the models and it is found that the model with periodic bird-to-bird transmission rate best explains the data. The best fitted model with the best fitted parameters gives a reproduction number of highly pathogenic avian influenza R_0 = 1.06. The best fitted model is a simple SI epidemic model with periodic transmission rate and disease-induced mortality, however, this model is capable of very complex dynamical behavior such as period doubling and chaos.