MIT researchers have analysed which airport hubs would be key to spreading a virulent disease, with some surprising results
When the next outbreak of Sars or Swine flu hits, New York’s John F Kennedy airport and Los Angeles’s airports will likely be the key spreaders of disease, according to a new study. But while the influence of these super-hubs may not come as much of a surprise, the third most outbreak-friendly airport in the states is far smaller, and far less obvious – Honolulu International.
In a paper published Monday in the journal PLoS One, a team of researchers from MIT outlined a new computer model that predicts how the 40 largest American airports may contribute to the diffusion of contagious disease within the first few days of a potential epidemic.
They looked at which hubs may be key “early spreaders” because knowing where epidemics may begin is key to stemming an outbreak, Marta Gonzalez, professor in the department of civil and environmental engineering at MIT, and one of the contributors to the new model, told the Guardian.
Factors other than sheer volume of travelers may contribute to the spread of disease.
While Honolulu is neither the busiest airport in the US (that would be Atlanta International), nor even among the top 20 biggest hubs (such as Chicago O’Hare and Minneapolis St Paul, both of which featured in the film Contagion), volume is not the key factor in disease spreading. Atlanta’s airport, which sees the highest volume of travelers in the US was ranked just eighth in its ability to cause contagion.
Honolulu airport “combines three important features that catalyze contagion spreading”, the study reports. Its geographical positioning in the middle of the Pacific Ocean makes it a prime layover between the US west coast and large Asian hubs; it’s also “well connected” to other powerful spreader airports, such as LAX; and it sees a high volume of long-range travel; all of which would help to spread a disease outbreak.
To fine-tune their new model, Gonzalez and her team analyzed cellphone data on top of passenger itineraries to determine real-world travel patterns, including layovers and re-routing.
“The spread of a disease is not random, just as human travel patterns are replicable and not random (particularly when taking into account return flights),” Gonzalez said. “We are able to create more accurate models due to our ability to analyze big data.”
Though computer models may not predict precisely when or if a new outbreak will hit, they can prepare officials in high-risk areas – such as Hawaii.
“This can improve the measures for containing infection in specific geographic areas and aid public health officials in making decisions about the distribution of vaccinations or treatments in the earliest days of contagion,” Gonzalez noted.
“Techniques such as multi-scale computer modelling … can make a contribution to strengthening our societies’ adaptiveness, resilience, and sustainability.”