Researchers in the Department of Civil and Environmental Engineering (CEE) at Massachusetts Institute of Technology (MIT) write about their findings in a paper published online on 19 July in PLoS ONE.
Model Focuses On Early Stages of OutbreakIn the past ten years we have seen a number of disease outbreaks that have spread around the world. In 2003, the SARS outbreak took merely a few weeks to spread from Hong Kong to 37 countries, killing nearly 1,000 people in its wake. In 2009, the H1N1 "swine flu" pandemic killed nearly 300,000 people worldwide.
Such outbreaks heighten awareness that new pathogens could spread quickly around the world with the help of air travellers.
To investigate such contagion patterns, scientists are building mathematical models that incorporate ideas from complex network systems and how information spreads in social networks.
Up to now, these models have focused on the final stages of outbreaks, looking at places that ultimately develop the highest infection rates.
But the MIT researchers took a different approach: they decided to focus on the early stages of epidemics and compare the likelihoods of spread from their home cities to other places through the largest 40 airports of the US.
Thus their model takes into account the travel patterns of individuals, the geographic location of airports, the differences in connectedness between airports, and the waiting times at individual airports. Bringing these factors together, the model then tries to predict where and how fast a disease might spread.
The researchers suggest this way of looking at the problem could help decide the best ways to contain infection and distribute vaccine and treatments in the first few days of an outbreak.
Senior author Ruben Juanes, the ARCO Associate Professor in Energy Studies at MIT's department of CEE, told the press:
"Our work is the first to look at the spatial spreading of contagion processes at early times, and to propose a predictor for which 'nodes' - in this case, airports - will lead to more aggressive spatial spreading."
"The findings could form the basis for an initial evaluation of vaccine allocation strategies in the event of an outbreak, and could inform national security agencies of the most vulnerable pathways for biological attacks in a densely connected world," he added.
New Model Is More RealisticThe model brings together two contrasting mobility patterns: one geophysical and the other human. The first comes from Juanes' studies of the flow of fluids through fracture networks in underground rocks, and the second comes from CEE's Marta González's studies that model human mobility patterns and trace contagion processes in social networks using cellphone data.
Incorporating these two sources of knowledge, the new MIT model departs from the conventional approach that assumes humans travel in a random diffusion pattern when moving from one airport to another.
The new model is more realistic. People don't travel randomly. They tend to repeat patterns.
The team applied Monte Carlo simulations to González's studies of human mobility patterns to determine the likelihood of any single traveler flying from one airport to another.
And they also replaced the conventional random flow model with an "advective fluid" model that assumes the transport process relies on the properties of the substance that is moving.
A conventional random flow model would show that the biggest airport hubs in terms of traffic volume would be the most influential spreaders of disease.
But the team, with their more realistic model, showed that is not the case.
Honolulu Airport: Less Traffic But Big InfluenceA random diffusion model would look at Honolulu airport, which has only 30% of the air traffic of New York's Kennedy International Airport, and conclude half its travellers would go on to San Francisco and half to Anchorage, taking the disease to those airports, passing it onto other travellers, who then in turn pass it on in further random travel patterns.
But the new MIT model looks at Honolulu airport and predicts, despite it having 70% less traffic, that in terms of disease spread, it is nearly as influential as New York's Kennedy International Airport.
This is because Honolulu airport occupies a unique position in the air transportation network. It is located in the Pacific Ocean and is well connected to distant, large and well-connected hubs. So it comes third, ahead of San Francisco, in the list of 40 US airports in terms of contagion-spreading influence.
Of the 40 US airports the model examined in terms of influence on disease spread, it puts Kennedy Airport in first place, followed by airports in Los Angeles, Honolulu, San Francisco, Newark, Chicago (O'Hare) and Washington (Dulles).
The top airport in terms of number of flights is Atlanta's Hartsfield-Jackson International Airport, but the model ranks it eighth in contagion influence. Boston's Logan International Airport ranks 15th.
González is the Gilbert W. Winslow Career Development Assistant Professor of CEE at MIT. She said the method they used is relatively new but very robust.
"The study of spreading dynamics and human mobility, using tools of complex networks, can be applied to many different fields of study to improve predictive models," said González, suggesting that the "incorporation of statistical physics methods to develop predictive models will likely have far-reaching effects for modeling in many applications".
A Vergottis Graduate Fellowship and awards from the NEC Corporation Fund, the Solomon Buchsbaum Research Fund and the US Department of Energy helped to fund the study.
Written by Catharine Paddock PhD