The potential value of crowdsourced surveillance systems in supplementing sentinel influenza networks: the case of France.
The potential value of crowdsourced surveillance systems in supplementing sentinel influenza networks: the case of France.
复制标题
DOI：
10.2807/1560-7917.es.2018.23.25.1700337
复制
发表时间：
2018-06
期刊：
影响因子：
--
通讯作者：
Colizza V
中科院分区：
文献类型：
--
作者：
Guerrisi C;Turbelin C;Souty C;Poletto C;Blanchon T;Hanslik T;Bonmarin I;Levy-Bruhl D;Colizza V
预防性监测系统提供丰富的众包数据，对个人及其特定时间的健康状况进行剖析。我们探讨了来自GrippeNet.fr（一个参与式监测系统）的数据在估计法国流感相关疾病发病率方面的有用性。方法：GrippeNet.fr是一个自2012年以来的在线队列，平均约为1000人。每周有5，000名参与者报告提示流感的体征/症状。GrippeNet.fr有灵活的标准来定义流感相关疾病。根据报告的体征/症状和纳入个人报告和参与标准的不同病例定义用于产生流感相关疾病发病率估计值，并与哨点网络的结果进行比较。我们重点关注2012/13和2013/14季节，当时法国存在两个监测流感样疾病（ILI）和急性呼吸道感染（ARI）的哨兵网络。结果：GrippeNet.fr的发病率估计值与官方时间趋势一致，ARI的准确性高于ILI。流感流行高峰期通常预计在一周后，尽管个人不定期参加。欧洲疾病预防和控制中心ILI定义，通常用于欧洲的参与性监测，在跟踪ARI比ILI更好地应用于GrippeNet.fr数据。结论：从众包数据中评估流行强度需要从连续几个季节中估计流行和强度阈值。该研究为众包监测提供了一个标准化的分析框架，在检测人群中流感相关变化方面显示出高灵敏度。它有助于提高各季节流行病的可比性和与哨点系统的可比性。在法国，GrippeNet.fr可以在2014年ARI监测停止后补充ILI哨兵网络。
Participatory surveillance systems provide rich crowdsourced data, profiling individuals and their health status at a given time. We explored the usefulness of data from GrippeNet.fr, a participatory surveillance system, to estimate influenza-related illness incidence in France. Methods: GrippeNet.fr is an online cohort since 2012 averaging ca. 5,000 weekly participants reporting signs/symptoms suggestive of influenza. GrippeNet.fr has flexible criteria to define influenza-related illness. Different case definitions based on reported signs/symptoms and inclusions of criteria accounting for individuals’ reporting and participation were used to produce influenza-related illness incidence estimates, which were compared to those from sentinel networks. We focused on the 2012/13 and 2013/14 seasons when two sentinel networks, monitoring influenza-like-illness (ILI) and acute respiratory infections (ARI) existed in France. Results: GrippeNet.fr incidence estimates agreed well with official temporal trends, with a higher accuracy for ARI than ILI. The influenza epidemic peak was often anticipated by one week, despite irregular participation of individuals. The European Centre for Disease Prevention and Control ILI definition, commonly used by participatory surveillance in Europe, performed better in tracking ARI than ILI when applied to GrippeNet.fr data. Conclusion: Evaluation of the epidemic intensity from crowdsourced data requires epidemic and intensity threshold estimations from several consecutive seasons. The study provides a standardised analytical framework for crowdsourced surveillance showing high sensitivity in detecting influenza-related changes in the population. It contributes to improve the comparability of epidemics across seasons and with sentinel systems. In France, GrippeNet.fr may supplement the ILI sentinel network after ARI surveillance discontinuation in 2014.