Contacts in context: large-scale setting-specific social mixing matrices from the BBC Pandemic project

Petra Klepac; Adam J Kucharski; Andrew JK Conlan; Stephen Kissler; Maria Tang; Hannah Fry; Julia R Gog

The largest eigenvalue of the reciprocal contact matrix is the average number of different people contacted during one day by someone who has just been contacted [29] (this value is proportional to R 0 ). The relevant type of contact for transmission (physical or all contact) will depend on the type of pathogen. For pathogens that require close sustained contact for transmission (such as commensal skin colonisers associated with nosocomial infections like Staphilococcus aureous [12] ), physical contacts might be more relevant. For very easily transmissible pathogens like measles, some combination of physical and conversational contacts would better represent transmission. Figure 8A illustrates the difference between the average number of people contacted for a whole range of matrices ranging from physical contacts only to all contacts (by gradually adding conversational contacts in increments of 0.1). The average number of physical contacts in the POLYMOD dataset are more than twice that of the BBC dataset. However on average participants in the BBC dataset speak to two more persons a day than POLYMOD participants (7.9 compared to 5.9).