Face to Face With Human Mobility Research

TURIN, ITALY—The 2009 H1N1 flu pandemic kept scores of virologists busy, but it also attracted interest from the growing number of scientists studying how people move and interact with one another. "It's our mobility patterns that spread epidemics," notes Alessandro Vespignani of Indiana University, Bloomington, who moderated an ESOF session today on how probing human movement can elucidate the dynamics of infectious diseases.

The talks included one on the well-known effort to track human mobility with the circulation of U.S. dollar bills, a project run by Dirk Brockmann of Northwestern University in Evanston, Illinois. (Last year, the data was used to predict the spread of H1N1, although the results were far from perfect.) And Marta Gonzáles of the Massachusetts Institute of Technology in Cambridge described following people over the course of a day via their cell phones. (Here's a recent paper from Science on that research avenue.)

Both Brockman's and Gonzales's efforts track large-scale movements of people, but Alain Barrat of the Center of Theoretical Physics in Marseille, France, works on a more intimate project called SocioPatterns. He and his colleagues monitor the quantity and duration of face-to-face interactions among people, key data for models of infectious diseases spread by human contact. To gather such information, the researchers have gone into schools, hospitals, museums, and academic conferences and solicited people to wear badges with radio-frequency identification tags that register when another RFID tag is within 1 or 2 meters. The radio signal from a tag can't travel through a person's body. But to confirm that tags were registering face-to-face contact, not just close proximity, the researchers compared hits during coffee breaks at a conference and during talks. The former far exceeded the latter, as one would expect, if the tags recorded primarily face-to-face interactions.

At the Science Gallery in Dublin, says Barrat, the experiment even became part of an exhibit on infectious diseases; some RFID tags started out "infected," and people could see if they had become infected during their visit. Over the course of 2 months, more than 30,000 people had their face-to-face interactions monitored in the museum.

Barrat and his colleagues have started to tease out some trends in the data. (A paper will soon appear in PLoS ONE, he says.) In the venues studied so far, most contacts last between 40 to 60 seconds; 85% last less than 1 minute, 95% last less than 2 minutes, and under 2% last more than 10 minutes. The researchers are also documenting networks of contacts for each venue; conferences show dense networks, as people are eager to meet, whereas museum networks exhibit few contacts, as people spend more time viewing exhibits than meeting new people. In the hospital setting, patients rarely meet other patients, which is good, says Barrat, as that should stymie the spread of communicable diseases.

Vittoria Colizza, who organized the ESOF session and is head of computational epidemiology at the ISI Foundation in Turin, Italy, says that using cell phones, RFID tags, and dollars bills represent a "new wave of research" in human mobility studies. "Until now, we've not been able to track people in a detailed way. ... Now we have the possibility of doing that," she says.