EAGER: Collaborative Research: Combining Community and Clinical Data for Augmenting Influenza Modeling

这种渴望代表了及时和必要的探索性工作，评估了社区来源的数据在传染病建模工作中的价值。社区生成的数据可能会因缺乏关于参考人口的信息而受到影响，这阻碍了流行率估计。从理论上讲，实时和接近实时的社区来源的数据已经被认为提供了重要的机会来改善传染病建模工作的及时性和范围，但关于社区感染数据对理解、监测和预测的价值仍然存在根本问题。为此，这里的工作将研究社区和临床生成的数据如何比较疾病发病率的衡量标准、贡献的人口统计数据以及流感动态的时空覆盖范围。向公众公布我们的研究和结果，将有助于在数据生成和预测方面向公众展示和教育他们。这项计划涉及对急性呼吸道感染的同期社区和临床数据进行严格和系统的比较。这项工作的目标将是首先产生一个具有确定的参考人口的多样化的社区来源的数据集。然后，我们将评估不同群体在社区和临床数据中结果的重要性，并考虑人口统计学和流行病学因素。将使用动态建模和贝叶斯推理方法来开发和增强疾病预测。来自社区样本的标准化和市政规模估计将被整合，数据生成和建模工作将一起用于评估社区数据对实时和接近实时的模拟和预测的影响。这项高风险的工作可能会改变我们在疾病和更广泛的社会问题预测方法中收集和使用数据的方式。

This EAGER represents timely and essential exploratory work assessing the value of community-sourced data in infectious disease modeling efforts. Community-generated data can suffer from lack of information about the reference population, which hinders prevalence estimates. In theory, real-time and near real-time community-sourced data has been recognized to offer important opportunity to improve timeliness and scope of infectious disease modeling efforts, but there are still fundamental questions regarding the value of community infection data for understanding, monitoring and forecasting. Towards this, work here will study how community and clinically generated data compare regarding measures of disease incidence, contributing population demographics, and spatio-temporal coverage in influenza dynamics. Public dissemination of our research and findings will help expose and educate the community in data generation and forecasting efforts.This project involves a rigorous and systematic comparison between contemporaneous community and clinical data on acute respiratory infections. The goal of this work will be to first generate a diverse community-sourced data set with a defined reference population. We will then assess significance of outcomes between groups in community and clinical data, accounting for demographic and epidemiological factors. Dynamical modeling and Bayesian inference methods will be used to develop and augment disease forecasts. Normalized and municipal scale estimates from the community samples will be integrated and the data generation and modeling efforts will together be used to assess the impact of community data on real-time and near-real time simulations and forecasts. The high-risk work can potentially be paradigm shifting regarding how we collect and use data in forecasting methods for disease as well as a broader range of societal issues.