Comparative Analysis of Early Dynamic Trends in Novel Coronavirus Outbreak: A Modeling Framework

Background: The 2019 coronavirus disease (COVID-19) represents a significant public health threat globally. Here we describe efforts to compare the dynamic transmissibility trends and epidemic size in China, as well as 9 countries in Asia, Europe, North America in the early epidemic phase. Methods: Using the time series of cases reported from January 20, 2020 to February 13, 2020 and transportation data from December 1, 2019 to January 23, 2020, we have built a novel time-varying growth model to predict the epidemic trend in China. We extended our method, using cases reported from January 26, 2020 - or the date of the first case reported, to March 15, 2020 to predict future epidemic trend and size in Japan, South Korea, Singapore, Iran, Italy, Germany, Spain, France, and the United States from January 26, 2020 to March 9, 2020. We compared the real-time effective reproduction number Rt of COVID-19 across countries. Findings: Here we predicted that the final epidemic size will be millions in Germany, Spain, and the United States, with more than 100,000 cases in Italy and France, and around 50,000 cases in Japan. The estimated epidemic would peak around April to July 2020. The duration will last on average 5-6 months in Europe and Asia, but longer in the United States at 10 months. The initial Rt at the onset of the epidemic is similar across all the countries (median 1.90 to 3.81) except the lowest in Singapore (median 0.53) and highest in Iran (10.96). Rt declined with time across all the countries, whereas it was reduced most rapidly in China with the lowest Rt in the fourth week (median 0.26 to 0.27). Effective intervention measures by reducing the infectious period in 2 days would result in a 4.0 to 49.3% reduction (median 43%) in the final epidemic size. Conclusions: A large-scale pandemic in Europe and the United States is inevitable. The heterogeneity in Rt reveals the impact of intervention measures in different countries. The control measures in South Korea, Singapore and China could lend further insight to other countries. Funding Statement: The research were partially supported by National Natural Science Foundation of China (Nos. 11931014 and 11829101) and Fundamental Research Funds for the Central Universities (No. JBK1806002) of China. Declaration of Interests: The authors declare no competing interests.