Monitoring Public Behavior During a Pandemic Using Surveys: Proof-of-Concept Via Epidemic Modelling

Implementing a lockdown for disease mitigation is a balancing act: Non-pharmaceutical interventions can reduce disease transmission significantly, but interventions also have considerable societal costs. Therefore, decision-makers need near real-time information to calibrate the level of restrictions. We fielded daily surveys in Denmark during the second wave of the COVID-19 pandemic to monitor public response to the announced lockdown. A key question asked respondents to state their number of close contacts within the past 24 hours. Here, we establish a link between survey data, mobility data, and, hospitalizations via epidemic modeling. Using Bayesian analysis, we then evaluate the usefulness of survey responses as a tool to monitor the effects of lockdown and then compare the predictive performance to that of mobility data. We find that, unlike mobility, self-reported contacts track the immediate behavioral response after the lockdown's announcement, weeks before the lockdown's national implementation. The survey data agree with the inferred effective reproduction number and their addition to the model results in greater improvement of predictive performance than mobility data. A detailed analysis of contact types indicates that disease transmission is driven by friends and strangers, whereas contacts to colleagues and family members (outside the household) only played a minor role despite Christmas holidays. Our work shows that an announcement of non-pharmaceutical interventions can lead to immediate behavioral responses, weeks before the actual implementation. Specifically, we find that self-reported contacts capture this early signal and thus qualify as a reliable, non-privacy invasive monitoring tool to track the implementation of non-pharmaceutical interventions