Improving pandemic mitigation policies across communities through coupled dynamics of risk perception and infection

Abstract

Capturing the coupled dynamics between individual behavioural decisions that affect disease transmission and the epidemiology of outbreaks is critical to pandemic mitigation strategy. We develop a multiplex network approach to model how adherence to health-protective behaviours that impact COVID-19 spread are shaped by perceived risks and resulting community norms. We focus on three synergistic dynamics governing individual behavioural choices: (i) social construction of concern, (ii) awareness of disease incidence, and (iii) reassurance by lack of disease. We show why policies enacted early or broadly can cause communities to become reassured and therefore unwilling to maintain or adopt actions. Public health policies for which success relies on collective action should therefore exploit the behaviourally receptive phase ; the period between the generation of sufficient concern to foster adoption of novel actions and the relaxation of adherence driven by reassurance fostered by avoidance of negative outcomes over time.

Group size and modularity interact to shape the spread of infection and information through animal societies

“…The difference will be greatest when the density of connections in a group is higher, as might be expected, for example, in a social insect colony or when comparing smaller human communities (higher network fragmentation) to large cities (networks likely still modular but with much larger groups). Future research that models direct interactions between information and disease spreading simultaneously (Funk et al 2009 ; Silk et al 2021 ) could also contribute to understanding patterns of social behaviour in animal groups.…”

Balancing timeliness of reporting with increasing testing probability for epidemic data

“…Self-testing has been proposed as a viable method of identifying infections and tracing exposures ( Catania et al., 2021 ), but, even with access to free tests, some people are unwilling to utilize tests ( Thunström et al., 2021 ). Unwilling people may represent a large minority (e.g., 31%) ( Thunström et al., 2021 ) that may also be unresponsive to reporting ( Grossman et al., 2020 ) until outbreaks have progressed beyond a behaviorally receptive phase ( Silk et al., 2021 ).…”

“…We built upon a model detailed elsewhere that relied on dynamic multiplex disease and communication layers ( Supplementary Methods ) ( Silk et al., 2021 , 2022 ). A diagram for the general progression of the model is included in Fig.…”

“…Furthermore, the emergent use of protective behaviors can result in the construction of a new social norm ( Bavel et al., 2020 ; Cialdini & Trost, 1998 ; McAlaney & McMahon, 2007 ), which may normalize behaviors that are more effective at mitigating a disease than in a norm-naive public. Prior work has indicated that the promotion of concern, and the associated adherence to protective behaviors, is most effectively implemented during a behaviorally receptive phase ( Silk et al., 2021 ). This phase signifies the period of time within which the normalization of protective behaviors can occur, without reassurance eroding their adoption prior to an epidemic peak.…”

“…Here we examined how variation in the timeliness of reporting and probability of testing may influence the generation of concern and subsequent adherence to protective behaviors, when confronted with an outbreak. We built on a previously published simulation-based dynamic multiplex network model ( Silk et al., 2021 , 2022 ) that coupled a disease network layer with a communication network layer. We included a population of 2000 individuals, distributed among three age categories and ten communities.…”