Mapping spread and risk of avian influenza A (H7N9) in China

Abstract

The outbreak of human infections with an emerging avian influenza A (H7N9) virus occurred in China in early 2013. It remains unknown what and how the underlying risk factors were involved in the bird-to-human cross-species transmission. To illustrate the dynamics of viral spread, we created a thematic map displaying the distribution of affected counties and plotted epidemic curves for the three most affected provinces and the whole country. We then collected data of agro-ecological, environmental and meteorological factors at the county level, and used boosted regression tree (BRT) models to examine the relative contribution of each factor and map the probability of occurrence of human H7N9 infection. We found that live poultry markets, human population density, irrigated croplands, built-up land, relative humidity and temperature significantly contributed to the occurrence of human infection with H7N9 virus. The discriminatory ability of the model was up to 97.4%. A map showing the areas with high risk for human H7N9 infection was created based on the model. These findings could be used to inform targeted surveillance and control efforts in both human and animal populations to reduce the risk of future human infections.