Mass gatherings as natural experiments: travel pulses reveal determinants of SARS-CoV-2 epidemic synchrony and predictability in U.S. states and counties

项目摘要 美国社区之间疫情的时间和强度的差异加剧了至关重要的酒吧的短缺- LIC卫生资源，从检测和接触者追踪到医院床位和医护人员。理解 社区在什么时候以及为什么最容易受到流行病的影响仍然是一个关键的未得到满足的需求 以及未来的流行病。我们的研究将利用一组五个大规模集会(例如，斯特吉斯摩托车拉尔- 位于2020年和2021年)，以评估旅行脉冲在美国一系列社区中的不同影响。我们的 这项研究围绕两个主要目标展开：目标1.量化美国各县的疾病发病率 从2020年的重点群众集会中传播脉冲，然后确定不同反应的预测因素。为了实现这一目标， 我们将使用统计模型来量化各县对游客的不同反应作为城市化的函数， 地理隔离、年龄结构和经济地位。目标2.量化疾病的可预测性和关联性- 2020年和2021年期间，美国各州的活动情况各不相同。我们将构建和参数化机械模型 县级传播，然后评估短期时间范围内的模型预测技能。这项研究 将提供急需的国家和州摘要，详细说明哪些因素对疾病影响最大 连通性和可预测性。我们的工作将提供对各个县的详细评估，包括 可公开访问的交互式地图，说明地理、人口统计学、感染史、 以及接种疫苗。这些产品将帮助利益相关者识别处于风险中的社区，并定制控制努力，以 根据当地情况，并将突出需要加强监测的地区。

Project Summary Variation in epidemic timing and intensity among U.S. communities has exacerbated shortages of vital pub- lic health resources, from testing and contact tracing to hospital beds and health care workers. Understanding when and why communities are most vulnerable to epidemic disease remains a critical unmet need in this and future epidemics. Our study will leverage a set of five mass gatherings (e.g., the Sturgis Motorcycle Ral- lies in 2020 and 2021) to assess the variable impact of travel pulses across a spectrum of U.S. communities. Our study is organized around two principle aims: Aim 1. To quantify how disease incidence in U.S. counties responds to travel pulses from focal mass gatherings in 2020 and then identify predictors of differing responses. To achieve this, we will use statistical models to quantify counties' differing response to travelers as a function of urbanicity, geographic isolation, age structure, and economic status. Aim 2. To quantify how disease predictability and connec- tivity vary among U.S. counties across 2020 and 2021. We will construct and parameterize mechanistic models of county-level transmission, and then assess model forecasting skill over short-term time horizons. This research will provide much-needed national and state summaries detailing which forces most strongly affect disease connectivity and predictability. Our work will provide detailed evaluations of individual counties, including publicly accessible interactive maps illustrating the relative effects of geography, demography, infection history, and vaccination. These products will help stakeholders identify at-risk communities and tailor control efforts to local conditions, and will highlight regions where increased surveillance is warranted.