RAPID/Collaborative Research: Agent-based Modeling Toward Effective Testing and Contact-tracing During the COVID-19 Pandemic

这项快速反应研究（Rapid）赠款将支持研究，以提高我们对COVID-19的传播和城市一级潜在缓解战略的理解，促进科学进步，为国家健康和繁荣做出贡献。随着COVID-19的持续传播，不同检测策略和预测模型的有效性受到质疑。检测策略包括使用免下车设施，这在其他地方取得了成功，但对老年人和低收入人群可能不切实际；以及使用医院，这进一步增加了医疗系统的负担，并可能带来更高的传染风险。预测该疾病传播的数学模型对于告知地方和全球决策者应该采取何种行动以减轻疫情并向民众提供救济至关重要。然而，由于早期阶段没有症状、复杂的活动模式和有限的检测资源，这些模型经常被混淆。该奖项支持对数学模型的基础研究，该模型将通过动力学和控制的进步来克服这些混淆因素。通过明确建模社会和流动性约束，这项研究将有助于提高社区的总体福祉，减少人口之间的差距。该模型将提供关键假设情景的模拟，并将包括对不同测试政策和缓解行动的评估，从而为参与遏制和根除该流行病的决策者提供宝贵支持。研究成果将提交给公众，包括卫生专业人员和当局，以便为当前危机中的公共政策提供信息。该研究将通过基于精细代理和数据驱动的模型实时应对COVID-19疫情，旨在提供前所未有的城市级病毒传播和潜在缓解策略的洞察力。该方法将对正在实施的和潜在的缓解战略的有效性进行彻底的假设分析。基于代理的模型将包括COVID-19的特定特征，如检测类型和时间、无症状发生和住院阶段。该框架将以来自纽约州新罗谢尔的公开可用的人口普查和地理参考数据为基础。将理性和非理性因素相关的社会行为纳入基于主体的模型在多个时空尺度上的流动性模式，以增加预测的粒度。网络理论和数据驱动的控制策略将通知增强的测试协议，包括在测试站点收集的可用接触数据库的基础上进行主动试验。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

This Rapid Response Research (RAPID) grant will support research that will improve our understanding of the spread of COVID-19 and potential mitigation strategies at the city level, promoting scientific progress and contributing to national health and prosperity. As COVID-19 continues to spread, the effectiveness of different testing strategies and predictive models are brought into question. Testing strategies include the use of drive-through facilities that have found success elsewhere but may prove impractical for elderly and low-income sections of the population, and the use of hospitals, which adds further burden to the healthcare system and may carry the risk of higher contagion. Mathematical models that forecast the spread of the disease are of paramount importance to inform local and global policy makers on the course of action that should be undertaken to mitigate the outbreak and give relief to the population. However, such models are often confounded by the absence of symptoms in early stages, complex mobility patterns, and limited testing resources. This award supports fundamental research toward a mathematical model that will overcome these confounding factors, through advancements in dynamics and control. By explicitly modeling social and mobility constraints, this research will help increase the general well-being of communities and reduce disparities across the population. The model will afford the simulation of critical what-if scenarios and will include the evaluation of different testing policies and mitigation actions, thereby constituting a valuable support to policy makers involved in the containment and eradication of the epidemic. Research outcomes will be presented to the public, including health professionals and authorities to inform public policy in the ongoing crisis.The research will respond to COVID-19 outbreak in real time through a fine-resolution agent-based and data-driven model that aims at providing unprecedented insight in the spread and potential mitigation strategies of this virus at the city level. The approach will afford thorough what-if analysis on the effectiveness of ongoing and potential mitigation strategies. The agent-based model will include COVID-19 specific features, such as the type and timing of testing, asymptomatic occurrence, and hospitalization stages. The framework will be grounded in publicly available census and geo-referred data from New Rochelle, New York. Social behavior associated with rational and irrational factors will be included in the mobility patterns of the agent-based model at multiple spatial and temporal scales to increase the granularity of the predictions. Network-theoretic and data-driven control strategies will inform enhanced testing protocols involving active trials on the basis of available contact databases collected at testing sites.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.