RAPID: Modeling the Coupled Social and Epidemiological Networks that Determine the Success of Behavioral Interventions on Limiting Spread of COVID-19

摘要：思想和病毒可以通过不同形式的互动在人类群体中传播。社会距离是一种理念，一旦付诸实施，就可以降低疾病传播风险，减缓感染在人群中的传播。就目前的新冠肺炎疫情而言，在缺乏现成的疫苗和医疗手段的情况下，社交距离是我们最好的防线。然而，社交疏远行为的普遍程度可能取决于社会和地理社区及其社会规范的组合，从而影响学校、社交媒体、工作环境以及朋友和家人之间的传播动态。一些社交社区(如社交媒体朋友群)的成员可能会在一起分享价值观和信仰，而不一定在地理上很接近。或者，人们可以在工作、商店、海滩、体育赛事中亲近，而不分享坚定的信念。有时，即使是陌生人也可能会模仿某些看得见的行为，比如在杂货店戴上防护口罩。流行病既是一种生物现象，也是一种社会现象。这项工作将开发实用的工具(模型)来预测集体行为和疾病传播的时空动态之间的相互作用。这将使人们能够更准确地预测随着时间的推移人口将需要的医疗资源。公共卫生措施不仅可以针对个人行为，也可以针对集体行为，这可能需要不同的激励和督促，从而使公共卫生信息能够最大限度地受益。该项目的结果还将通过关于数学在预防大流行中的作用的公开网络研讨会分享。这项消除这些差距的工作将涉及两种不同类型的数学建模工作。第一种类型将依赖于设计一个系统常微分方程组(ODE)来捕捉疾病动力学和社会影响。这个ODE模型将假设群体行动疾病和信念状态之间的平均转换率足以获得洞察力，产生量化特征来描述随着时间的推移，信念动态如何与社区中的疾病流行相互作用。第二种类型将依赖于设计耦合的多层网络，其中一层捕捉社会影响，另一层捕捉身体接触和疾病传播。这个模型将探索每一层中个人之间的动态联系，其中联系的强度可以根据另一层中同一个人的状态和邻居而变化。第二个模型，通过专注于特定的网络结构，将补充ODE模型获得的关于平均行为的见解，并提供对个人在改变社区认知和/或传播感染方面可能扮演的不同角色的洞察。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Abstract:Ideas and viruses can spread in human populations through different forms of interaction. Social distancing is an idea that, when enacted, can lower disease transmission risk and slow the spread of infection in a population. In the case of the current COVID-19 outbreak, in the absence of ready vaccines and medical treatments, social distancing is our best line of defense. The prevalence of social distancing behaviors can depend, however, on the mix of social and geographic communities and their social norms, influencing spread dynamics in schools, social media, work environments, and among friends and family. Members of some social communities (e.g. social media friend groups) may share values and beliefs together without necessarily being in close geographic proximity. Alternatively, people can come into physical proximity—at work, shops, beaches, sporting events— without sharing strongly-held beliefs. Sometimes, even strangers may copy certain visible behaviors, such as wearing protective masks at the grocery store. A pandemic is both a biological and social phenomenon. This work will develop practical tools (models) that predict the interaction between collective behavior and the spatiotemporal dynamics of disease spread. This will enable more accurate predictions of medical resources the population will need over time. Public health measures can target not only individual behavior but also collective behavior, which may require different incentives and nudges, such that public health messaging can be maximally beneficial. Results from the project will also be shared through a public webinar on the role of mathematics in pandemic preparedness.This work to address these gaps will involve two different types of mathematical modeling efforts. The first type will rely on designing a system ordinary differential equations (ODEs) to capture both disease dynamics and social influence. This ODE model will assume that mass action average rates of transition between both disease and belief states are sufficient to gain insight, producing quantitative characterizations to describe how belief dynamics interact with disease prevalence in a community as both progress over time. The second type will rely on designing coupled multi-layer networks in which one layer captures social influence and the other captures physical contact and disease transmission. This model will explore dynamic connections among individuals within each layer, where the strength of contact can shift based on the state and neighbors of the same individual in the other layer. This second model, by focusing on particular network structures will complement the insights about average behaviors gained by the ODE model and provide insight into the different roles individuals may play in shifting community perception and/or spreading infection.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.