Synthetic contact networks to assess covid-19 protocols in schools

ABSTRACT
Close proximity contact data capture mixing patterns in a population. Publicly available data sets often cover short recording windows, and must be synthetically extended in time before serving as inputs in infectious disease spreading models. Focussing on school populations, I will show how synthetic contacts can be constructed from empirical data in order to preserve memory effects related to friendships between children. Using numerical simulations of Covid-19 spread, we found that these memory effects are relevant to capture plausible infection pathways between students, and that two days of data appear sufficient to produce robust contacts.
In order to model the spread of Covid-19 under a pooled testing program in Swiss schools, we developed additional methods to synthetically vary the population size of these synthetic contacts. Fitting and informing a Covd-19 spreading model under pooled testing with multi-source data from the implemented program, we compared the impact of the pooled testing program to protocols commonly applied in schools during the pandemic. In the second part of this talk, I will discuss results of this modelling study showing that pooled testing is a resource-efficient strategy while limiting infections in children at school, offering a viable alternative to more resource-intensive individual weekly screening protocols.