In this study, we mainly applied Thompsona’s method to estimate the Rt of COVID-19 in different areas to assess the spread of the epidemic in China and oversea. We found out that the Rt in China had almost been falling below one, which mean the non-pharmaceutical intervention in China were effective. But the situation of overseas is not optimistic as Rt >1, which indicated the epidemic was still in the stage of gradual spread.
As of Feb 29, almost 80,000 confirmed cases of COVID-19 in China have been reported and nearly 3,000 people have died. As we all know, Wuhan as the origin outbreak is also the most severely affected area accounting nearly 60 percent of cumulative cases nationwide in China. And the fatality rate in Wuhan is much higher than in other regions. Although existing confirmed cases was still increasing the rate of increase was steadily declining. It meant the series of non-pharmaceutical interventions implemented by the Chinese Government have achieved remarkable achievement. In this study, we found daily incidence of three areas were at different levels in China. Most of the new cases still came from Wuhan. The overall daily incidence has been gradually decreasing since February 4 and the daily incidence in Hubei Province excluded Wuhan and China excluded Hubei Province was gradually approaching zero, while Wuhan was still at a relatively high level and the decline of incidence was relatively slow.
As of Feb 11, the cumulative number of reported cases in China showed a bifurcation. According to the update of “The diagnosis and treatment program of novel coronavirus pneumonia (Trial Fifth Edition)”, only Hubei province added the classification of “clinical diagnosis”, that the suspected cases with clinical features of pneumonia were clinically diagnosed cases. And it was cancelled in the sixth edition a few days. That’s because the Chinese authorities considered the limited medical resources and huge number of suspected patients. During Feb 4th and Feb 10th, the daily incidences in Wuhan fluctuated around 2,000, indicating that due to limited medical resources, many suspicious patients were still waiting for diagnosis, which might result in aggravated conditions and the risk of potential outward transmission. To speed up controlling the epidemic, Chinese authorities had adopted the standard of adding clinical diagnosis. When most suspected patients are diagnosed and treated effectively, the chain of disease transmission will become clearer.
In order to find out the trend of the epidemic, we calculated the Rt in different regions in China. And we also further calculated the Rt abroad to understand the international development of the epidemic. Based on official data and published literature, we analyzed the interval time between the onset dates of 126 pairs of infected and infected cases. By fitting the Gamma distribution, we got the median serial time of COVID-19 is 4.18 days. It’s longer than Hiroshi Nishiura’s result (2.6 days), which estimated by 26 infector-infected pairs. The intergenerational time obtained in this study was smaller than the estimated value of 5 clustered outbreaks in 425 cases at beginning of the epidemic. And the time was also smaller than the average incubation period of 5.2 days, which was also estimated by 425 cases at beginning. We supposed that the infected person was infected before the onset of symptoms.
Then, by inputting the serial time distribution and the latest (as of February 16) the incidence of laboratory confirmed cases with 7 days as the moving time window to estimate the value of Rt.
At first, we compared the Rt among three different regions in China. We could see all Rt in three areas had downward trends, as well as had fallen below one. It meant the series non-pharmaceutical interventions implemented in China had achieved some achievement.
And Wuhan as the origin of the epidemic, which also having the majority of patients, had the different change trend of Rt. Compared with other two regions, the average Rt of Wuhan was higher and token longer time to drop below one. In addition, as we all know, there are still over thousands accumulated cases in Wuhan, so it still has a long way to control the epidemic.
Optimistically, the contagiousness of COVID-19 in China has been well controlled. As long as strict management is maintained, the number of incidences will drop to zero soon. As of February 25, the incidence in China had dropped from thousands to hundreds, meanwhile the number of cures was also increasing. Zeliang Chen et al predicted the peak incidence to be in early or middle February.
Secondly, we assessed the development of the epidemic overseas. Considering the extent of the epidemic, we listed top five countries with relatively severe outbreaks, Japan, Korea, Singapore, Iran and Italy, and combined other overseas countries for Rt calculations. As of February 26, except Singapore, other four countries’ Rt were still upper than one, which meant the epidemic might be growing abroad. On February 7, the Singapore government officially decided to upgrade the response level of the outbreak response system from yellow to orange in view of the discovery of several local cases in Singapore, which were unrelated to previous cases and had no history of travel to China. In our study, the result showed the Rt in Singapore had fallen below to one on February 20, which meant the outbreak in Singapore had been under control. Although the other overseas countries’ Rt also had fallen below to one, it didn’t mean the outbreak around the world was in control. On 30 January 2020, the outbreak was declared a Public Health Emergency of International Concern by WHO. And WHO and countries have been engaged in massive preparedness activities. And according to the growth rate of the worldwide new confirmed cases, there were still a trend of rapid growth, which liked the early stage of epidemic in Wuhan. We hope it should be taken more attention and learn the experience from China or Singapore.
Our study used the latest generation time and new confirmed case data at the same time, trying to estimate the Rt of each region in China and oversea, which could provide a basis for grasping the development trend of the epidemic and evaluate prevention and the interventions. But there were some limitations in our study. First, we collected the personal history of epidemiological investigation data from different publish websites to estimate the serial interval, which made bias. We hope the government might build the database to collect the history epidemiology data, what could provide more reliable information.[21-22] Second, failure to exclude imported cases in the calculation of Rt might make bias, but it could still reflect a certain epidemic situation and the effectiveness of prevention and intervention. Another limitation is that we used the time-series data of confirmed cases to calculate according to the official date of diagnosis. So, the date we estimated that the Rt went below one should be delayed by about 10~17 days (the incubation period of 7 days added days from onset to diagnosis of 10 days). Although we estimated the time when the epidemic was under control had a delay than the actual time, it could still provide information of the trend of the epidemic.