Tags for studying the spread of epidemics

Until now, data concerning the dynamic of contacts was vague: it was often assumed that within a group, each individual could be confronted with another according to an average estimation of the frequency and duration of contacts. In the case of an infectious risk, the determined probability of infection was consequently not representative: in fact, in epidemiology, a single contact lasting one minute does not have the same impact as a repeated or longer contact.

As part of the SocioPatterns project, a team gathering physicists and epidemiologists from CNRS laboratories and other organizations took the opportunity of a two-day congress to measure the contacts between the 500 participants, each of whom agreed to wear a RFID (radiofrequency identification device). These tags made it possible to detect when two individuals were close together and faced each other. The devices were able to intercommunicate through the emission of very low intensity waves (1,000 times weaker than a mobile telephone) at a distance comprised between 1 and 2 meters. After collecting and processing the data by computer, the researchers succeeded in modeling the dynamic of contacts.

This has enabled the team to precisely define the relations between individuals in terms of temporal data, namely duration and frequency. Modeling the dynamic of actual interactions in a population makes it possible to simulate the spread of infection and, in the long term, envisage a better response in the event of an epidemic. To refine the model, the team intends to gather complementary information, which is why contacts between pupils at a school are currently being analyzed and the measurement infrastructure has already been deployed for a study in a hospital in Rome. In the future, it will be possible to envisage integrating this new inter-tag communication within businesses or in new technologies, since smartphones are already very well equipped with sensors.

More information: Simulation of an SEIR infectious disease model on the dynamic contact network of conference attendees. J.Stehlé, et al. BMC Medicine, 11 July 2011

Provided by CNRS