This interesting proof-of-concept study by Smieszek and Salathé addressed social interactions within a high school, which is an important focus for seasonal and pandemic influenza transmission [12
]. As acknowledged by the authors, a key limitation of this study is the lack of validation against epidemiological data from real school outbreaks. The simulation model used to evaluate the performances of the method is a conceptualized version of disease transmission, and although it is driven by real contact information, it remains one step removed from the actual disease-transmission process. A previous study combining outbreak data in an elementary school with contact-network information highlighted the importance of gender on influenza transmission, with children of the same gender infecting each other more frequently (reflecting assortative mixing) [4
], an issue that was not considered by Smieszek and Salathé. Interestingly, school outbreak data have also shown that the exact location of children within the classroom does not matter, which supports the use of simple class-schedule information as proposed by Smieszek and Salathé [11
] rather than the use of more detailed seating charts. Although there has been good progress overall in elucidating social interactions among school-age children, more studies are needed to address whether contact patterns, and hence transmission links, might differ between elementary and high schools.
Another limitation of the school-based study by Smieszek and Salathé [11
] relates to the contribution of other units to disease transmission. About one-third of all influenza secondary-transmission events are believed to occur within households [13
], whereas only 7 to 20% are thought to occur in schools [14
]. Hence, estimating the relative infection risk of individuals in a variety of settings relevant for disease transmission, including schools, households, conferences, and transportation systems, will be important in future research. It is not clear how the method proposed by Smieszek and Salathé [11
] could be generalized to household and work environments, where systematic 'schedules' are more difficult to obtain.
As noted by the authors, the transmission mode of influenza and other respiratory pathogens is not clearly understood, but probably involves a combination of direct contact and transmission by fomites and aerosols, which makes it difficult to capture the social network relevant for disease transmission. Because the transmissibility of influenza has been shown to be associated with environmental conditions [15
], actual transmission rates could vary within the same school, house, or office building, owing to local differences in the environment. In the future, more elaborate studies should collect local environmental variables such as room ventilation rates to better quantify influenza transmission potential in confined settings [17
In summary, Smieszek and Salathé [11
] have introduced a promising and practical method to identify individuals with high infection potential who can be targeted for outbreak detection and control. Future studies should employ consistent methodological approaches to measure contact networks in different settings, in parallel with careful disease monitoring. Technological advances in contact-network sensing devices and pathogen identification methods (for example, multiplex PCR), combined with innovative approaches for disease surveillance (for example, web-based and smart-phone technologies [18
]), have huge potential to increase our understanding of infectious disease transmission and to suggest novel ways of detecting and controlling outbreaks.