A transport model for the spread of bird flu uri icon


  • Highly Pathogenic Avian Influenza (HPAI) is a major threat to the poultry industry worldwide. In developing countries backyard poultry production is often an important livelihood activity for women who use the additional income for unexpected household expenses. The disease can therefore have significant social impacts. To evaluate different options governments and public health authorities could employ to reduce the negative effects of HPAI outbreaks, it is necessary to have a means for simulating outbreaks of the disease. Models can aggregate the available information about the disease dynamics for conducting such simulations. The underlying assumption of the model we propose is that the fastest way for HPAI to spread is through transport of infected material or chickens. Undoubtedly, other infection routes, like direct transmission between neighbouring villages or disease spread by migrating birds, play a role in the disease dynamics, too, but these are likely to be slower than transport of infectious material through trade. The speed of transport depends on the topography of the landscape: transport will be faster along roads and slower off roads. In the model, the landscape is divided into quadratic areas termed grid cells. Grid cells contain information about the poultry population. Each grid cell is assigned a travel friction, which is the time it takes to cross the grid cell with the appropriate means of travel, i.e. by truck on roads and on foot off-road. The travel friction is computed from data on road type, vegetation, slope and other landscape features. The model calculates which area can be reached within 12 hours travel time for each grid cell. Data of past outbreaks of HPAI in Nigeria were used to validate the model. The data were also used to calculate the expected number of secondary outbreaks resulting from an infected farm during the period the farm is infective. The probability distribution of the number of secondary outbreaks is used to predict the future spread of disease and to access the quantity of poultry affected. The model was then used to evaluate different control options

publication date

  • 2010