School of Mathematics and Statistics, Beijing Technology and Business University, Beijing, 100048, China.
School of Science, Beijing University of Posts and Telecommunications, Beijing, 100876, China.
Environ Sci Pollut Res Int. 2021 Apr;28(16):20240-20246. doi: 10.1007/s11356-020-11859-w. Epub 2021 Jan 6.
The outbreak of COVID-19 has become a global public health event. Many researchers have proposed many epidemiological models to predict the outbreak trend of COVID-19, but all use confirmed cases to predict "onset cases." In this article, a total of 5434 cases were collected from National Health Commission and other provincial Health Commission in China, spanning from 1 December 2019 to 23 February 2020. We studied the delayed distribution of patients from onset to be confirmed. The delay is divided into two stages, which takes about 15 days or even longer. Therefore, considering the right truncation of the data, we proposed a "predict-in-advance" method, used the number of "visiting hospital cases" to predict the number of "onset cases." The results not only show that our prediction shortens the delay of the second stage, but also the predicted value of onset cases is quite close to the real value of onset cases, which can effectively predict the epidemic trend of sudden infectious diseases, and provide an important reference for the government to formulate control measures in advance.
新型冠状病毒肺炎疫情的爆发已成为全球公共卫生事件。许多研究人员提出了许多流行病学模型来预测 COVID-19 的爆发趋势，但都使用确诊病例来预测“发病病例”。本文从国家卫健委和其他省级卫健委共收集了 5434 例病例，时间跨度为 2019 年 12 月 1 日至 2020 年 2 月 23 日。我们研究了患者从发病到确诊的延迟分布。延迟分为两个阶段，大约需要 15 天甚至更长时间。因此，考虑到数据的右截断，我们提出了一种“提前预测”方法，使用“就诊病例”数来预测“发病病例”数。结果不仅表明我们的预测缩短了第二阶段的延迟，而且发病病例的预测值与发病病例的实际值非常接近，这可以有效预测突发传染病的流行趋势，为政府提前制定防控措施提供重要参考。
