Network Intervention Planning without Actual Network Data for Infectious Disease Control

项目摘要(摘要) 接触网络流行病学是一个引人注目的流行病学框架，旨在对动态交互作用进行建模 通过他们的社交网络对人们进行监测，以便跟踪感染的级联，特别是传染性疾病。 接触网络流行病学中的基于网络的模拟可以结合人的属性和 行为(例如，年龄、种族/民族、戴口罩)、他们的互动模式(例如，同性关系或 分类)和社会结构(例如，社会规范和政策，包括非药物干预 [NPIS]))。尽管获取精确的网络数据具有挑战性，但它可以指导我们识别潜在的工作 网络干预战略，这可能被证明有助于应对新冠肺炎大流行。 利用网络干预的框架，一项试点模拟研究提出了替代NPI策略 呆在家里的秩序，在人们的社会经济活动持续的同时，传播得到缓解 (Nishi等人，2020，PNAS)。在最有效的划分+平衡群体策略中，一个社会群体(如 同一工作场所的员工和同一所学校的学生)被随机分为两个小组 一个相等的数字，以减少身体接触的数量。如果它是以空间方式运行的，则额外的空间 为分组做好准备；如果以临时方式运作，两个分组将进行其 在不同的营业时间内进行活动。因此，该战略将允许人们从事同样的工作 经济活动的规模。所提议的策略的优点在于它不需要实际的网络 数据，这在大多数情况下是很难获得的。 在初步研究之后，这项研究试图创建其他新的新的NPI策略来控制传染病 (目标包括新冠肺炎和其他新出现的疾病)(目标1)。这项研究还试图创作小说 疫苗分配的网络干预战略(目标2)。为减轻流行病而提出的策略 原则上，优化疫苗分配不需要实际的网络数据。因此，他们的潜力 需要使用具有现实假设的基于网络的模拟或使用其他方法来检查效果 方法，包括数学建模。所使用的社交网络将基于样本城市 10,000人(Nishi等人，2020，PNAS)和各种公开可用的网络结构(使用 二次数据)。此外，本研究将分析早期预警信号(EWS)的作用，该信号已被 在传染病控制背景下发展的非线性动力系统。我打算用76号 加利福尼亚州县新冠肺炎数据(AIM 3)。

PROJECT SUMMARY (ABSTRACT) Contact network epidemiology is a compelling epidemiologic framework that aims to model dynamic interactions of people over their social networks in order to track infection cascades, especially for communicable diseases. Network-based simulations in contact network epidemiology can incorporate variations in people’s attributes and behaviors (e.g. age, race/ethnicity, wearing a facial mask), their interaction patterns (e.g. homophily or assortativity), and social structures (e.g. social norms and policies including non-pharmaceutical interventions [NPIs]). Although obtaining precise network data is challenging, it can guide us to identify potential working network intervention strategies, which may prove beneficial in addressing the COVID-19 pandemic. Using the framework of network interventions, a pilot simulation study proposed alternative NPI strategies to the stay-at-home order, in which transmission is mitigated while people’s socioeconomic activities are sustained (Nishi et al, 2020, PNAS). In the most effective dividing + balancing groups strategy, a social group (e.g. employees of the same workplace and students of the same school) is divided randomly into two subgroups with an equal number to reduce the number of physical contacts. If it is operated in a spatial manner, additional space for the subgroups is prepared; if it is operated in a temporal manner, the two subgroups will engage in their activities during different business hours. Therefore, the strategy would allow people to engage in the same magnitude of economic activities. The strength of the proposed strategy is that it does not require actual network data, which is difficult to obtain in most cases. Following the pilot study, this research seeks to create other novel NPI strategies for infectious disease control (the targets are both COVID-19 and other emerging diseases) (Aim 1). This research also seeks to create novel network intervention strategies for vaccine allocation (Aim 2). The proposed strategies for mitigating an epidemic and optimizing vaccine allocation will not, in principle, require actual network data. Therefore, their potential effect needs to be examined using network-based simulations with realistic assumptions or using other approaches, including mathematical modeling. The utilized social network will be based on a sample city of 10,000 individuals (Nishi et al, 2020, PNAS) and various network structures that are publicly available (the use of secondary data). Moreover, this research will analyze the role of early warning signals (EWS), which has been developed in non-linear dynamical systems in the infectious disease control context. I plan to use the 76 California County COVID-19 data (Aim 3).