EPIDEMIC TRENDS ANALYSIS AND RISK ESTIMATION OF 2019-NCOV OUTBREAK
Qinghe Liu, Deqiang Li, Zhicheng Liu, Zefei Gao, Junkai Zhu, Junyan Yang, Qiao Wang
doi: https://doi.org/10.1101/2020.02.09.20021444
Abstract
Backgroud: Since the outbreak of coronaviruses on December 8, 2019, the spread of new coronaviruses is increasing. With no specific medicine, analysis of epidemic trends is critical for disease control and epidemic risk management. Methods: We collected information on demographic characteristics, exposure history, and illness timelines of laboratory-confirmed cases of NCIP that had been reported by Feburary 6, 2020. Considering the huge population migration during the spring festival of China, a Flow-SEIR model is proposed to perform empirical analysis. Assuming that most of people will go back to work from February 10 to the end of February, the risk of the coming migration from home to workplace is defined and estimated, based on the confirmed cases and the Spring Festival migration data. We use data from Baidu Migration to estimate the coming migration for returning to the workplace after vacation from February 10 to 28, 2020. The risk of migration for returning was then estimated. Main findings: (1). The time when the number of daily new cases has reached a peak level, i.e. the national epidemic inflection point, is expected to be on February 7 - February 9. After the inflection point, the number of infected people will grow slowly and the epidemic will be gradually controlled. We define the temporal number of infected cases as the net cumulative number of cases. And the peak value represents the maximum of the net cumulative number, after this peak point, the current number of cases will gradually decrease and the epidemic will gradually subside. The predicted peak value of confirmed cases in Hubei can reach 62800 (56900 - 70300, 0.95 CI). The peak time in Hubei is on February 29 (Feb. 25 - Mar. 08, 0.95 CI). The peak arrival time of areas except Hubei will be in the interval of Mar.8 - Mar. 15, 2020. It will take about 1.5-2 months from the peak to the end of the epidemic. Most provinces have the predicted peak values less than 1000. The potential epidemic situation will be stable in March 2020- June 2020. (2). Rather than larger outbreak, the estimated number of infective patients in Hubei (in which Wuhan is located) presents a front and steep feature, manifesting the overall trend that the epidemic tends to be consistent. (3). The necessary consciousness of self-protection and isolation is very effective for epidemic prevention and control. It can reduce the number of patients by nearly 90%. The effect of self-protection and isolation is stronger than the policy of traffic blockage. Provincial level traffic blockage can only alleviate 21.06% - 22.38% of the peak number of patients. (4). Early warning of epidemic situation is very important. In the experimental environment, the results show that if there is 1 day in advance of the timely urban blockage, about 3600 people will eventually be reduced in the country. If there is 1 day of delay, about 1800 people will finally be at risk. (5). Assuming that the false negative rate is 50% and the potential transmission rate is 1-4 times higher than that of the confirmed patients, the real peak number of patients is likely to be between 214400-472500. Suppose that the transmission rate of the false negative patient is twice that of the confirmed patient, if the false negative rate decreased by 5% in one week, the peak number of patients will slow down by 11.62%. If the proportion decreased by 10% in one week, the peak number of patients will decreased by 21.91%, which shows timely detection of false negative population is more effective for epidemic control. (6). Under the natural condition, the return of Spring Festival transportation will aggravate the epidemic situation, especially for Guangdong, Zhejiang, Jiangsu, Hunan, Henan, Shanghai, Fujian and Beijing. We did not perform the global propagation of 2019-nCoV because that the available initial data is not sufficient to make valuable conclusions.
*注，本文为预印本论文手稿，是未经同行评审的初步报告，其观点仅供科研同行交流，并不是结论性内容，请使用者谨慎使用.
