Updated: Tuesday, March 07, 2006
New model to predict spread of disease through air travel
months of research, four scientists at the School of Informatics have
unveiled a mathematical framework that could serve as a way to predict
the spread of disease through air travel.
patterns of illness from air travel
By Matt Cunningham | Indiana Daily Student |
Tuesday, March 07, 2006
The study was published in a recent edition of the "Proceedings of the
National Academy of Science." The computational model was stochastic,
meaning it took into account random variations across a broad range of
factors. The study used census data and a database provided by the
International Air Transport Association.
Never before has such a detailed amount of airline information been
employed in this type of research. The database of the International
Air Transport Association includes all commercial airports in the world
and the flight connections among them, as well as the number of
passengers traveling on a given route.
Using this information, numerical simulations were carried out by more
than 10,000 equations simultaneously to achieve the predictions. The
database included information from 265 airlines, which comprises 99
percent of all international air traffic. Disease patterns from cities
and census information obtained from 3,100 cities in 220 countries
supplemented the database information.
"I think the major implication of the research is the fact that we can
now start studying the reliability of global epidemic forecasts," said
informatics professor Alessandro Vespignani. "This will provide policy
makers with a more informed way to look at risk assessment and strategy
Research associate Vittoria Colizza said the study will be useful in
the development of strategies for containing the spread of an emerging
"This stochastic model can be applied to a variety of diseases,"
Colizza said. "Once you have this model you can apply it to whatever
disease you want."
She cautioned that in order to implement a disease the first
consideration should be "specific infection dynamics of each disease."
"Regarding infection dynamics, one would need to specify the
infectiousness of the virus and the typical evolution of the disease in
an individual by identifying the stages of the disease and the average
periods an individual spends in each of those stages," Colizza said.
Zoltan Toroczkai, deputy director of the Center for Nonlinear Studies
at Los Alamos National Laboratory in New Mexico, is involved with
similar research. He is working on macro simulations of city epidemics.
He complimented IU researchers on incorporating realistic mobility
models derived from the airline database in studying epidemic networks.
Toroczkai said these research efforts are important due to an
increasing connectedness within populations. Cheaper and easier methods
of travel invite more physical contact, which Toroczkai said represents
vulnerability for the world.
"Diseases can easily take advantage of this small-world character of
the contacts and virtually overtake the population in a very short
time," Toroczkai said. "Research is paramount for understanding the
implications of the increased connectivity and for designing strategies
of counteraction and defense. Professor Vespignani and his group's work
is an excellent piece of research targeting to fill this gap."