RAPID: Developing Social Differentiation-respecting Disease Transmission Models

在该项目中，开发了考虑社交网络和暴露机会差异的传播模型，以深入了解COVID-19在人群中的不平等传播。 一些地区的COVID-19传播缓慢甚至没有传播，而其他地区则不堪重负。 在高流量地区，一些社区的风险比其他社区大得多。 为了解释这种不均匀的传播，这些模型将与社会密度和社会人口特征相关的人口差异纳入其中，这些特征塑造了疾病暴露和社交距离的能力。 这些模型增强了对社会状况如何影响疾病风险和疾病缓解工作成本的一般理解，这将使决策者能够评估不同健康保护干预措施的相对成本，并可能优化干预措施，最大限度地减少经济损害，同时最大限度地提高人身安全。灵活和可扩展的疾病传播模型，可以解释观察到的疾病传播中的社会差异。 我们采用模拟模型来实现这一目标，并利用COVID-19大流行对疾病特异性感染参数以及转为住院、死亡或康复率的最佳估计。 建模发生在两个层次上：基于代理的模型（ABM）和基于小区域单元的仿真模型。 ABM由社交网络数据构建，具有最大的灵活性，可根据不同类型的人群进行调整，从发展中国家的农村社区到人口密集的城市中心。 小区域（人口普查区块组）基于单元的模拟模型将网络结构转换为基于人口和经济相似性概况的相互作用概率，包括人口分化，但扩展到国家一级。 这两种建模策略相互补充，可用于评估不同的缓解策略的健康有效性（挽救的生命）和经济困难。该奖项反映了NSF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。

In this project, transmission models that account for differences in social networks and exposure opportunities are developed to gain insight into the unequal spread of COVID-19 across populations. Some areas have experienced slow to no spread of COVID-19 while other settings have been overwhelmed. Within high-volume locations, some neighborhoods have been at much greater risk than others. To account for this uneven spread, these models incorporate population differences related to social density and sociodemographic characteristics—features that shape disease exposure and ability to social distance. These models augment general understanding of how social situation affects both disease risk and the cost of disease mitigation efforts, which will allow decisionmakers to evaluate the relative costs of different health-preserving interventions and, potentially, optimize interventions that minimize economic harm while maximizing physical safety.This project aims to have accurate, flexible and scalable models for disease transmission that can account for observed social differentiation in disease spread. Simulation models are employed to meet this goal, drawing on best estimates from the COVID-19 pandemic for disease-specific infection parameters and rates of transitioning into hospitalization, death, or recovery. Modeling occurs on two levels: Agent-Based Models (ABMs) and small-area cell-based simulation models. ABMs are constructed from social network data and allow for maximum flexibility, being tunable to different types of populations, ranging from rural communities in developing nations to dense urban centers. Small-area (census block group) cell-based simulation models, which translate network structure to interaction probabilities based on demographic and economic similarity profiles, include population differentiation but scale to the national level. These two modeling strategies complement each other and can be used to evaluate different mitigation strategies for both health effectiveness (lives saved) and economic hardship.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.