Any parent knows that colds spread like wildfire, especially through schools. New research using human-networking theory may give a clearer picture of just how, exactly, infectious diseases such as the common cold, influenza, whooping cough and SARS can spread through a closed group of people, and even through populations at large. With the help of 788 volunteers at a high school, Marcel Salathe, a faculty member in the Department of Biology at Penn State, developed a new technique to count the number of possible disease-spreading events that occur in a typical day. This research will be published in the print edition of the journal Proceedings of the National Academy of Sciences. To view a video about the research, visit http://www.youtube.com/watch?v=5rWKlN_nz5Y online.
“Theoretically, we know that every day people come into contact with many other people, that interactions vary in length, and that each contact is an opportunity for a disease to spread,” Salathé said. “But it’s not like you can take a poll and ask people, ‘How many different people have breathed on you today, and for how long?’ We knew we had to figure out the number of person-to-person contacts systematically.”
Using a population of high-school students, teachers, and staff as a model for a closed group of people, Salathe and his team designed a method to count how many times possible disease-spreading interactions occurred during a typical day. Volunteers were asked to spend one school day wearing matchbox-sized sensor devices — called motes — on lanyards around their necks. Like a cell phone, each mote was equipped with its own unique tracking number, and each mote was programmed to send and receive radio signals at 20-second intervals to record the presence of other nearby motes. Volunteers then were asked to simply go about their day by attending classes, walking through the halls, and chatting with other people. At the end of the day, Salathe’s team collected the motes and recorded how many mote-to-mote interactions had occurred, and how long each interaction had lasted.
“An interaction isn’t necessarily a conversation,” Salathe explained. “Even when people aren’t talking, they might be sneezing and coughing in each other’s direction, bumping into each other, and passing around pathogens.” To record even these non-conversational events — any kind of spatial closeness that would be enough to spread a contagious disease — each mote used a 3-meter maximum signaling range, extending outward from the front of the person’s body.
Defining a single interaction as any 20-second or longer event of mote-to-mote proximity, Salathe and his team found that the total number of close-proximity events was 762,868. “The same two people may have had many very brief interactions,” Salathe said. “Still, we have to count each brief interaction individually, even between the same two people. From a pathogen’s point of view, each interaction is another chance to jump from person to person.” In addition, the team found peaks of interactions at times between classes, not surprisingly, when mote-wearing volunteers were physically closer to one another, moving around in the halls on their way to the next class.
Salathe and his team found that, at the end of the day, most people had experienced a fairly high number of person-to-person interactions, but they also found very little variation among individuals. Strikingly, they did not find any individuals who had an extraordinarily high number of contacts when compared with the rest of the group. Such individuals — called super-spreaders — are known to be very important in the dynamics of disease spread.
“For example, in sexual-contact networks, one often finds a group of people with a much higher potential to contract and spread a virus such as HIV,” Salathe said. “This potential is due to these individuals’ extremely high number of interactions. But in our experiment, while there may have been popular kids with a few more interaction events, for the most part, everyone had about the same high level of interaction.” Salathe explained that while schools may indeed be “hot beds” for colds and the flu, individual students do not seem to vary with regard to exposure risk due to their contact patterns.
Salathe said that data from his motes also confirmed an important social-networking theory — that contact events are not random because many “closed triangles” exist within a community. “If person A has contact with person B, and person B has contact with person C, chances are that persons A and C also have contact with each other,” Salathe said. “Real data illustrating these triangles provide just one more piece of information to help us track how a disease actually spreads.” Salathe also said that networking data such as his may help guide public-health initiatives such as vaccination strategies and prevention education.
In addition to Salathe, other researchers who contributed to this study include an interdisciplinary team at Stanford University: Maria Kazandjieva and Jung Woo Lee, graduate students in computer science, and faculty members Philip Levis of the computer science department, Marcus W. Feldman of the biology department, and James H. Jones of the anthropology department.
This research was funded by the National Science Foundation, the National Institutes of Health, and a Branco Weiss fellowship.
Barbara Kennedy, Penn State University