Use of Twitter data to improve Zika virus surveillance in the United States during the 2016 epidemic
Use of Twitter data to improve Zika virus surveillance in the United States during the 2016 epidemic
复制标题
DOI：
10.1186/s12889-019-7103-8
复制
发表时间：
2019-06-14
影响因子：
4.5
通讯作者：
Wu, Jun
中科院分区：
文献类型：
--
作者：
Masri, Shahir;Jia, Jianfeng;Wu, Jun
背景寨卡病毒（ZIKV）是一种新出现的蚊媒虫媒病毒，可造成严重的公共卫生后果。2016 年，ZIKV 在包括美国在内的世界许多国家引起流行。ZIKV 监测和病媒控制对于应对未来的流行病至关重要。然而，与及时公布病例报告有关的挑战极大地限制了当前监测方法的有效性。在许多基础设施薄弱的国家，往往没有既定的病例报告系统。以前对 H1N1 大流行、普通流感和最近埃博拉疫情的调查研究表明，可以利用即时可用的带有时间和地理标记的 Twitter 数据来克服这些限制。方法在这项研究中，我们使用了最近开发的名为 Cloudberry 的系统来过滤 Twitter 数据的随机样本，以调查在国家和州（佛罗里达州）层面使用此类数据进行 ZIKV 流行病跟踪的可行性。结果虽然模型在病例数较低时往往预测过度，而在病例数极高时则预测不足，但在模型校准后，将预测的每周 ZIKV 病例数与观察到的病例数进行比较，结果表明预测准确性总体合理，佛罗里达州模型的 R-2 为 0.74，美国模型的 R-2 为 0.70。内部交叉验证后对预测和观测到的 ZIKV 病例进行的时间序列分析显示出非常相似的模式，证明了模型的合理性能。在空间上，美国所有 50 个州的累积 ZIKV 病例数（本地和旅行相关）和寨卡推特的分布在根据人口进行调整后显示出高度的相关性（r=0.73）。这对流行病学家和负责在未来疫情爆发时保护公众的公共卫生官员具有很高的价值。
BackgroundZika virus (ZIKV) is an emerging mosquito-borne arbovirus that can produce serious public health consequences. In 2016, ZIKV caused an epidemic in many countries around the world, including the United States. ZIKV surveillance and vector control is essential to combating future epidemics. However, challenges relating to the timely publication of case reports significantly limit the effectiveness of current surveillance methods. In many countries with poor infrastructure, established systems for case reporting often do not exist. Previous studies investigating the H1N1 pandemic, general influenza and the recent Ebola outbreak have demonstrated that time- and geo-tagged Twitter data, which is immediately available, can be utilized to overcome these limitations.MethodsIn this study, we employed a recently developed system called Cloudberry to filter a random sample of Twitter data to investigate the feasibility of using such data for ZIKV epidemic tracking on a national and state (Florida) level. Two auto-regressive models were calibrated using weekly ZIKV case counts and zika tweets in order to estimate weekly ZIKV cases 1 week in advance.ResultsWhile models tended to over-predict at low case counts and under-predict at extreme high counts, a comparison of predicted versus observed weekly ZIKV case counts following model calibration demonstrated overall reasonable predictive accuracy, with an R-2 of 0.74 for the Florida model and 0.70 for the U.S. model. Time-series analysis of predicted and observed ZIKV cases following internal cross-validation exhibited very similar patterns, demonstrating reasonable model performance. Spatially, the distribution of cumulative ZIKV case counts (local- & travel-related) and zika tweets across all 50U.S. states showed a high correlation (r=0.73) after adjusting for population.ConclusionsThis study demonstrates the value of utilizing Twitter data for the purposes of disease surveillance. This is of high value to epidemiologist and public health officials charged with protecting the public during future outbreaks.