Individual and neighborhood based socioeconomic factors relevant for contact behaviour and epidemic control.

Background Identifying sources of heterogeneity in contact patterns is key to inform disease transmission models. Recent works have investigated how individual-based socio-economic factors, besides age, affect contact behavior, but neglected the individuals' area of living. Here, we aim at estimating contact matrices stratified by both individual-based and area-based socio-economic factors.Methods We use social contact data from Switzerland collected in 2021, combined with a neighborhood-based index of socio-economic position (SEP). Despite lacking socio-economic information on the contacts, we develop a method to reconstruct contact matrices fully stratified by age, education level, and SEP, with varying assortativity levels.Results We find a positive association between education level and number of contacts in the elderly, and, notably, a negative association between SEP level and number of contacts in adults. Compared to homogeneous mixing, accounting for heterogeneous contact patterns leads to higher attack rates in groups with high education level, especially for adults living in low SEP areas and seniors living in high SEP areas. Adults and young individuals living in high SEP areas are the main contributors to transmission. Including socio-economic factors into model parameterization has limited effect on the basic reproduction number but substantially influences the effectiveness of control strategies. The more assortative contacts are, the higher the control effort required by a targeted strategy to be successful in preventing disease spread.Conclusions Our results shed light on contact behavior in previously neglected socio-economic groups, enable model integration of socio-economic indicators, and provide insights to improve disease control.Patterns of human interaction (also known as contact patterns) that impact disease spread can vary depending on age and education level, but less is known about the influence of the economic conditions of the area where people live. In addition, socio-economic information of the contacts (such as income, education or occupation) is often not collected in traditional surveys. In this study, we characterize mixing patterns across different socio-economic groups, even when some information is missing. Using mathematical modeling, we identify which groups generate most infections, and which groups experience the highest disease burden. We also show how diverse contact patterns can determine the success of a control strategy. Our findings can help improve the design of public health measures to contain epidemic spreading.