Time use data and contact patterns

Emilio Zagheni, University of California, Berkeley
Piero Manfredi, Dipartimento di Statistica & Matematica Applicata all'Economia
Francesco C. Billari, Università Bocconi
Joel Mossong, Laboratoire National de Santé, Luxembourg

Social contact patterns are the critical explanatory factor of the transmission dynamics of infectious diseases. Both indirect (via observed epidemiological data) and direct (via diaries concerning “at risk” contacts) approaches to the measurement of contacts by age have been proposed in the literature. In this paper, we systematically discuss the possibilities offered by Time Use Surveys (TUS) for the estimation of contact patterns. Indirectly, TUS provide finely “time structured” data on the activities undertook by sampled individuals, i.e. they provide valuable information on the “arenas” where contacts occur. Focusing on the notion of “mixing by activity” we develop two estimation approaches for mixing matrices: the first one is based on the amount of time people spend together, while the second one on the aggregation of people’s daily activities. We show that the allocation of contacts to activities appears to be the major source of assortativeness of contacts by age.