Sampling Criteria for Monitoring Influenza Emergencies Under Constrained Testing Capabilities

流感病毒几乎每年都会造成突发事件，现有的监测系统对于解决疾病检测、监测和控制方面的不确定性至关重要。监测流感活动的一个重要工具是流感病毒类型确诊病例的趋势。确诊病例趋势有助于估计季节开始的时间和哪些病毒正在传播，因为每种流感病毒针对不同的年龄组，并产生不同类型和严重程度的症状。在流感紧急情况下，来自感染病例的样本大量到达检测实验室，这导致了混乱，随着实验室为了应对压倒性的检测需求而增加了能力，增加了运营成本。此外，在这种混乱环境下产生的确诊病例数据正被决策者和研究界认真利用。该奖项支持基础研究，以了解常规和非常规样本采样方法如何将每天确诊流感病例数量的公布延迟和不确定性降至最低。新的发现和方法不仅将改进抽样和测试实践，还将保证其用户的数据质量。这项研究促进了多样性，因为该项目将招募女性研究生。此外，这项研究还促进了知识传播，因为将准备培训材料并提供给公共卫生政策制定者、工程学和医学研究生。分析人员利用报告的确诊流感病例趋势来确定正在传播的病毒的流行病学特征。然而，目前还不确定确诊病例在多大程度上代表了病毒的特性。为了解决这一问题，已经提出了几种方法，但它们存在严重的局限性。例如，这些方法需要难以估计的信息，它们没有考虑受限的实验室能力，它们没有考虑到疫情演变和样本收集过程中的不确定性。研究小组将在实验室开发一种在疾病紧急情况下操作的方法(模型)。模型是一种实时采样和贝叶斯学习流行病参数的算法，该算法考虑了疾病过程和样本采集过程中的不确定性。期望模型特征对动态抽样理论有一定的贡献。研究工作还将侧重于在流感紧急情况下评估任何类型的样本抽样标准的框架。该框架得出的结果将有助于病毒传染病的管理和监测的理论。

Influenza viruses create emergencies almost every year, and existing surveillance systems are important for addressing uncertainty in the disease detection, monitoring and control. An essential tool for monitoring influenza activity is the trend of cases confirmed with an influenza virus type. Confirmed case trends are useful to estimate when the season starts and which viruses are circulating, as each flu virus targets different age groups and produces symptoms of different types and severity. During influenza emergencies, specimens from infected cases massively arrive at the testing laboratories, which leads to chaos and increased operational costs as the labs increase their capacity in response to the overwhelming testing demand. Moreover, confirmed case data produced under this chaotic environment is being used intently by decision makers and the research community. This award supports fundamental research to understand how conventional and non-conventional methods for specimen sampling minimize the publication delay and uncertainty in the daily number of influenza confirmed cases. The new findings and methods will not only improve sampling and testing practices, but will also guarantee quality of data to its users. The research promotes diversity, as female graduate students will be recruited for the project. In addition, the research promotes knowledge dissemination, as training materials will be prepared and delivered to public health policymakers, and graduate students in engineering and medicine.Analysts use the reported trend of confirmed influenza cases to identify the epidemiological features of a circulating virus. However, there is uncertainty about the extent to which confirmed cases are representative of the properties of the virus. To address this concern, several methods have been suggested but they present significant limitations. For example, the methods require information that is hard to estimate, they do not consider constrained lab capacity, and they do not account for the uncertainty in the outbreak evolution and the specimen collection process. The research team will develop a Method for Operations during Disease Emergencies in the Lab (MODEL). MODEL is an algorithm for real-time sampling and Bayesian learning of epidemic parameters that accounts for the uncertainties in the course of the disease and in the specimen collection process. It is expected that MODEL features will contribute to the theory of dynamic sampling. Research efforts will also focus on a framework to evaluate any type of specimen sampling criteria during an influenza emergency. The results derived from the framework will contribute to the theory of management and surveillance of viral infectious diseases.