研究表明：美国新冠肺炎疫情较大概率于2019年9月前后已开始流行

中科院预印本平台（ChinaXiv）22日发布的一项基于大数据建模分析的新冠肺炎疫情起源时间研究结果表明，美国新冠肺炎疫情较大概率于2019年9月前后已开始流行。

研究人员依据传染病传播模型和大数据分析的方法，建立优化模型，基于已公开数据对美国东北部12州和中国武汉市、浙江省等地的疫情起源时间进行了推断。

研究结果显示，对于美国东北部这12州，新冠肺炎疫情首例感染发生概率50%的日期多数位于2019年8月到10月，最早是罗德岛州的2019年4月26日，最晚是特拉华州的2019年11月30日，均早于美国官方公布的全美首例确诊日期2020年1月20日。计算结果表明，美国新冠肺炎疫情较大概率于2019年9月前后已开始流行。

计算显示，中国武汉市首例感染发生概率50%的日期为2019年12月20日，中国浙江省首例感染发生概率50%的日期为2019年12月23日。据此推断，中国新冠肺炎疫情较大概率于2019年12月下旬已开始流行。这一结论与流行病学调查结果基本相符，证明该计算方法准确可靠。

A study published on the pre-print platform ChinaXiv, a Chinese open repository for scientific researchers, indicated that COVID-19 had started spreading in the U.S. around September 2019 "with a high confidence probability."

The study, conducted by researchers from the University of Science and Technology of China and the University of Chinese Academy of Sciences, is based on a data and hybrid model-driven method.

By combining a mathematical model and artificial intelligence technology, qualitative and quantitative analysis of the case data of an infectious disease can reveal its epidemic law and detect its origin and development trend, they said.

"By selecting 12 representative regions in the U.S. for analysis, the dates of the first infected case with 50 percent confidence probability are mostly between August and October 2019, which are earlier than the officially announced date of the first confirmed case in the U.S. on January 20, 2020," the authors wrote.

Using the same methodology, the study concluded that the spread of the COVID-19 in China likely began in late December 2019.

Origin tracing is crucial for infectious disease prevention and control, and a series of previous studies showed that countries including the U.S., Spain, France, Italy and Brazil had been hit by the coronavirus before its outbreak in China, the authors wrote. 

推荐阅读：