Synthesizing data and subject matter expert opinion to establish simulation model input parameters: Avian influenza in poultry as an example

Transboundary Animal Diseases (TADs) are defined as highly contagious epidemic diseases with the potential for rapid spread across national borders causing serious socioeconomic and possibly public health consequences. Research efforts directed at understanding the transmission and control of TADs are important for disease control planning and may help prevent a widespread sustained epidemic in livestock and/or humans. Simulation models can provide an artificial but realistic representation of complex livestock and poultry systems in which to formulate and test hypotheses. They can be useful tools for response planning, especially when information regarding a disease in the region of interest is scarce or unavailable. The reliability of model outputs is highly dependent on the quality of model input parameters. Examples of disease spread model inputs include animal population information, disease epidemiology, the frequencies and distances of animal and human movements, and disease control measures. Model outputs often include information such as the numbers of animals or herd/flocks infected, mechanisms of infection, and other characteristics of disease outbreak and control. Establishing realistic model input parameters is a rigorous and iterative process involving informaticists, veterinarians, animal husbandry experts, economists, and epidemiologists. Methods of gathering and synthesizing data and expert opinion to develop realistic model input parameters will be discussed using the example of recent research simulating the spread and control of highly pathogenic avian influenza in the southeastern United States using the North American Animal Disease Spread Model.