Direct and indirect transmission of avian influenza: results from a calibrated agent-based model

This paper develops an agent-based model of the spread of highly pathogenic avian influenza in a regional poultry sector. Spatially located flocks and flock owners interact with each other within a realistic regional geography. A specific regional poultry industry in Thailand shapes the model details: production practices that affect bio-containment, key transportation routes and methods, commercial and social networks, and an unusually detailed tabulation of flock types, flock owners, and human behaviors. Model simulations follow disease spread from an initial infection through both direct and indirect transmission pathways. This demonstrates how to realistically calibrate a model of disease spread to a regional poultry industry, while attending to the relative importance of different pathways. Despite its region-specific calibration, the model design facilitates easy adaptation to similar settings. This supports public policy by providing a modeling framework that may be used to simulate interventions to reduce or prevent the regional transmission of avian influenza.