Foot and mouth disease model verification and 'relative validation' through a formal model comparison

Researchers from Australia, New Zealand, Canada and the United States collaborated to validate their foot and mouth disease models--AusSpread, InterSpread Plus and the North American Animal Disease Spread Model--in an effort to build confidence in their use as decision-support tools. The final stage of this project involved using the three models to simulate a number of disease outbreak scenarios, with data from the Republic of Ireland. The scenarios included an uncontrolled epidemic, and epidemics managed by combinations of stamping out and vaccination. The predicted numbers of infected premises, the duration of each epidemic, and the size of predicted outbreak areas were compared. Relative within-model between-scenario changes resulting from different control strategies or resource constraints in different scenarios were quantified and compared. Although there were differences between the models in absolute outcomes, between-scenario comparisons within each model were similar. In all three models, early use of ring vaccination resulted in the largest drop in number of infected premises compared with the standard stamping-out regimen. This consistency implies that the assumptions made by each of the three modelling teams were appropriate, which in turn serves to increase end-user confidence in predictions made by these models.