U.S. Health Officials Build 'Very Large Scale' Flu Simulations

Researchers at the U.S. National Institutes of Health (NIH) have been at work at computational simulations that model potential widespread outbreaks of the avian flu in the human population.

"What we have been doing at NIH is build computational simulation models on a very large scale. We have produced two models of Thailand published in Nature and Science, two models of the U.S. published in Proceedings of the Academy of Sciences and Nature," says Joshua Epstein, senior fellow, economic studies at the The Brookings Institution and director of a global epidemic model at NIH.

"Our missions, as it were, are to project the global spread of these diseases, development containment strategies, and offer real-time decision support in crises," Epstein adds.

"But why should you believe any of these models?" he asks.

"One answer to that is that we try to make them credible by using what data exists and calibrating the models to known outbreaks of influenza," he says. "We can try to at least get the social contact dynamics right, and in some cases we know how the bug acted in human hosts and so we can reconstruct these epidemics on computers and say this is how the 1968 global flu really did unfold on the 1968 population transportation system, and that is what we are going to do."

So far, the avian flu has not been transmitted from human to human, only animal to human, but researchers are worried that could change, triggering a worldwide flu pandemic.

"So we are going to say let's reconstruct the 1968 flu, then leaving the 1968 bug, let's update the transportation and population to modern levels and then with modern levels captured let's make the best ... biomedical estimate we can about what a human-to-human variant would look like and plot that into this contact dynamic and make some sort of ballpark estimate of how things might unfold under a huge array of uncertainties, but nonetheless, we think it is a valuable exercise," Epstein says.