Predicting contact rates in response to non-pharmaceutical interventions: A validated protocol for generating contact matrices relevant to transmission modelling of respiratory diseases

Transmission models predict infectious disease transmission and inform policy makers. For respiratory diseases, many models incorporate the rate of face-to-face contacts within and between age groups, summarized in contact matrices. Contact rates may change due to non-pharmaceutical interventions (NPIs). To forecast transmission, contact matrices should reflect this impact before NPI introduction. Limited data on which contacts will be prevented forces modellers to assume changes in contact rates, challenging consistency, transparency, and accuracy. We developed a protocol to predict future contact matrices using data on time-use, contacts, and demographics, collected before an epidemic. For each NPI, we identified activities on which less time will be spent. The protocol assumes time reductions translate to proportional reductions in numbers of contacts made during these activities, enabling prediction of future contact matrices. School and work time were stratified by educational level and profession, based on distributions of students across educational levels and the working population across professions. We validated the protocol by applying it to NPIs during the COVID-19 pandemic in the Netherlands, comparing predicted to observed matrices in contact surveys. Predicted matrices correlated well with observed matrices. Stratified by age, predicted contact rates agreed with observed rates, especially for contacts involving adults. For child-child contacts, predicted rates were lower than observed rates, with observed school contacts likely overreported. Predicted rates of workplace contacts matched with office occupancy data. The protocol uses pre-epidemic data, providing consistent and transparent contact matrices for transmission models. It is well-suited for upcoming epidemics when data are periodically collected.