Sunday, May 03, 2009

h1n1 systematization

Northwestern ROCS | In our simulations for the spread in the United States we take Mexico City to be the starting point of the infection. The initial conditions of the model are chosen to match the available information on confirmed cases in the United States and Mexico, and we permit for a bracket of unconfirmed cases. We take into account the travel of individuals between counties in the United States and the incoming flux of airline passengers from Mexico City. We consider a worst-case scenario based on assumptions made from the information we have gathered thus far. The key factors in our modeling approach are very accurate human mobility datasets on scales from a few to a few thousand kilometers. We obtained the underlying multi-scale human mobility network indirectly by our recent investigation on the geographic circulation of dollar bills in the United States, which is an excellent proxy for human mobility and includes small scale daily commuting traffic, intermediate traffic, and long distance travel by air. Our simulations consist of multiple layers, each layer possessing and increasing degree of accuracy and complexity. Our final projections are done with a fully stochastic model that incorporates the inherent randomness in disease dynamics that is particularly important at the onset of an epidemic when the number of infected individuals is small compared to the whole population.