Congregate air sampling for population-based detection of tuberculosis

项目摘要 结核病是全世界发病率和死亡率的主要原因之一。 研究结核病传播和早期病例的传统方法 检测，一直基于个人层面的筛选。由于时间长， 结核病的亚临床阶段，在此期间，传染性个体有许多偶然的 接触，了解传播发生在社区环境中一直是一个 棘手的挑战社会网络和分子流行病学研究表明， 80%的传播与密切接触者或家庭成员无关。 此外，在寻求医疗保健之前的这种长时间传播使得 结核病难以控制。积极的病例发现方法，包括筛查 个人在寻求护理之前，由于需要大量的资源， 在大多数高负担国家，这种做法是不可持续的。在这里，我们建议 一种全新的结核病检测方法， 限制：识别M.肺结核在公共场所的空气中传播。使用 定制的空气采样装置和高灵敏度和特异性的分子诊断 技术，我们将严格调查这种方法与三个基本的，真实的- 世界应用。第一个是评估聚集空气采样作为一个群体筛选 结核病早期病例检测和确定最佳抽样的方法 通过经验数据收集和基于模型的分析，第二 应用是通过以下方式确定结核病传播的高风险环境， 在高度流行的南非，在公共场所进行系统的空气采样 乡三是高效生成人口级结核病患病率 通过结合空气采样和社会混合数据，并应用统计 基于合并诊断抽样的推理模型。该项目将产生 关键数据对这种新的人口为基础的方法，以M。结核病检测，并将 制定流行病学和统计框架，将其转化为工具 用于监测和早期发现病例。

Project Summary Tuberculosis is one of the leading causes of morbidity and mortality worldwide. Conventional approaches to studying tuberculosis transmission, and to early case detection, have been based on individual-level screening. Because of the long, subclinical phase of tuberculosis, during which infectious individuals have many casual contacts, understanding where transmission occurs in community settings has been a vexing challenge. Social network and molecular epidemiologic studies indicate that over 80% of transmission cannot be linked to close contacts or household members. Additionally, this prolonged period of transmission prior to healthcare-seeking makes tuberculosis difficult to control. Active case finding approaches, which involve screening individuals prior to care-seeking, are resource-intensive due to the high number-needed- to-test and have not been sustainable in most high burden countries. Here, we propose a fundamentally new approach to tuberculosis detection that would overcome these limitations: identifying M. tuberculosis in the shared air of congregate settings. Using custom-built air sampling devices and highly sensitive and specific molecular diagnostic techniques, we will rigorously investigate this approach with three fundamental, real- world applications. The first is to evaluate congregate air sampling as a group screening method for early case detection of tuberculosis, and to identify optimal sampling approaches through empirical data collection and model-based analysis. The second application is to identify high-risk environments for tuberculosis transmission, through systematic air sampling in public settings throughout a highly endemic South African township. The third is to efficiently generate population-level tuberculosis prevalence estimates by combining air sampling and social mixing data, and applying statistical inferential models based upon pooled diagnostic sampling. This project will generate critical data on this new population-based approach to M. tuberculosis detection, and will develop an epidemiologic and statistical framework with which to translate this into a tool for surveillance and early case detection.