Carlo Giovanni CAMARDA

Carlo Giovanni CAMARDA

Modelling Social Contact Data: a smoothing constrained approach

Abstract: Mathematical modelling of infectious diseases transmitted by the respiratory or close-contact route (e.g. influenza) is increasingly being used to determine the impact of possible interventions. Several authors have shown that informing mathematical models with social contact data is of great value avoiding making a priori contact assumptions with little or no empirical basis. Moreover, it has also been shown that estimating mixing patterns from social contact data is not without difficulties and model choice is likely to affect the mathematical model outcome substantially. In this paper we present a model for estimating age-specific contact rates from social contact surveys. Specifically we propose a smoothing constrained approach to estimate these contact rates between people of possibly different ages. We use a two-dimensional approach where contact rates are assumed smooth from a cohort perspective as well as from the age distribution of contacts. The proposed method uses a combination of penalized likelihood for rectangular arrays for smoothing the contact rates and linear constraints to ensure reciprocity of contacts. We illustrate our approach using social contact data from a population-based contact survey that has been carried out in Belgium in 2006.

Updated on 15/05/2014

The unit is part of the department ’Infection and Epidemiology"




Since August 2006, the unit is localized in building "Biotop", 5th floor