Prevalence in Sichuan Province (non-foreign imported cases)
Geographical distribution On March 4, Sichuan Province ended the increase in confirmed cases on the mainland (the first wave). Each city of Sichuan Province (21 in total) reported confirmed cases, with a total mainland case number of 539 and an incidence rate of 0.6462/100,000. Analysis of confirmed cases was based on their current address, with 48.33% in urban and 51.67% in rural areas. The top five cities with confirmed case numbers were Chengdu, Ganzi, Dazhou, Nanchong, and Guang'an. The cities with top five incidence rate were Ganzi, Panzhihua, Guang'an, Chengdu, and Dazhou (Table 1). Moreover, the confirmed cases were distributed in 103 counties, accounting for 56.28% of Sichuan Province (183 in total). Among them, 3 cities (out of 21 cities) found cases reported in all their counties. Geographically, the regions with the highest number of cases and counties were located in the eastern part of Sichuan Province, which is directly adjacent to Chongqing and closer to Hubei Province. Although the number of confirmed cases in the eastern counties was relatively high, a small number of counties in the western part also had a high incidence (Table 1, Figure 1, Figure 2).
Time distribution The time of the first confirmed case report in the 21 cities of Sichuan Province mainly came one week after the Wuhan lockdown (January 24-January 31).13 At this stage, the current addresses of the first confirmed cases were distributed in 70 counties, accounting for 68% of the reported counties (103 in total) in Sichuan Province on March 4. The maximum number of cases was on the third and fourth days after the Wuhan lockdown (January 26-27). During those two days, 28 counties (27%) reported their first case. The rapid increase in the confirmed number was concentrated in the four weeks after the closure of Wuhan (January 24-February 20), and the increase was particularly obvious in the first two weeks (Figure 3).
Prevalence in neighbouring provinces of Sichuan and other parts of China
Spatiotemporal distribution: On December 31, 2019, 27 cases were first reported on the official website of Hubei Province, all in Wuhan; on January 23, 2020, Hubei announced that it had entered a Grade II response and the Wuhan lockdown began.14 Of the 835 cases reported nationwide (including 5 in Hong Kong, Macao, and Taiwan), 549 were in Hubei (495 of these in Wuhan) and 267 were in all provinces outside Hubei. A total of 15 cases in Sichuan Province were distributed in 7 cities (the first imported case was diagnosed on January 21), ranking 9th in China. The number of cases in Chongqing, adjacent in eastern Sichuan, was the only neighbouring province with more cases than Sichuan (27 cumulative cases). The epidemic situation in Yunnan and Guizhou adjacent in southern Sichuan was relatively light. Twenty-eight days later, on February 20, a total of 75,993 cases (including Hong Kong, Macao, and Taiwan, which accounted for 102 cases, the total number of which were corrected based on Hubei's revision) were reported, of which 63,088 were in Hubei (45,346 of these were in Wuhan). The revision was originally made by the Health Commission of Hubei Province website on February 21, raising the total to 12,905 cases in provinces outside Hubei (including in Hong Kong, Macao, and Taiwan). A total of 525 cases in Sichuan Province were distributed in all 21 cities, ranking 11th in China, next to Chongqing. Fourteen days later, on March 5, a total of 80,710 cases (including 158 cases in Hong Kong, Macao, and Taiwan) were reported nationwide, including 67,592 cases in Hubei (49,797 of these in Wuhan) and a total of 13,118 cases in provinces outside Hubei (including Hong Kong, Macau, and Taiwan). A total of 539 cases in Sichuan were distributed in all 21 cities, still ranking 11th in China. The number of new cases nationwide fell below one hundred the next day.
The concentrated outbreak time of confirmed cases was the first 4 weeks after the implementation of the lockdown measures (Figure 4). On December 31, cases were only reported in Wuhan. Later, on January 19, the first confirmed case was reported outside Wuhan in Shenzhen of Guangdong Province. From December 31 to January 23 (i.e., 118 cities reported cases on the day of Wuhan lockdown), the number of confirmed cases increased by a factor of 31. In the next four weeks after the Wuhan lockdown, 90% of the cities nationwide were affected, and the number of cases increased by a factor of 91, compared with January 23. In the fifth and sixth weeks, the number of newly confirmed cases and the number of reported areas slowed down significantly, with an increase of only a factor of 0.06 (63 cities except Wuhan increased with a total of 473 cases, and the average daily increase was 0.53 cases/city). In the 7th week (March 5-March 12), there was only a small increase of 62 cases in 10 cities of China except for Wuhan.
Spatial stratified heterogeneity q statistics of the incidences among the provinces in China (excluding Hong Kong, Macao and Taiwan) on January 23, February 20, March 5 and March 12 were 0.16 (p>0.05), 0.37 (p<0.05), 0.35 (p<0.05) and 0.35 (p<0.05). Spatial autocorrelation analysis of China showed positive spatial correlation. The Global Moran's Index on January 23, February 20, March 5 and March 12 were 0.02, 0.20, 0.18 and 0.18 (all the p value <0.05). The results showed that from January 23 to February 20, the epidemic situation was rapidly gathering in space. Besides, the spread of the epidemic slowed down after February 20, and the spatial concentration decreased slightly. Furthermore, hot spot analysis showed that the hot spots did not expand after February 20, and the spread of the epidemic was effectively controlled (Figure 5).
Distribution of the epidemic length in each region: In the seventh week after the closure of Wuhan, on March 12, the growth of cases ended in all cities in China, except Wuhan, and in Taiwan and Hong Kong . Based on statistics at the city level (the districts of municipalities were analysed as a city unit according to the high population density of municipalities), the length calculated by the date of occurrence and the end of local cases in a total of 413 cities were used for analysis. The study shows that the longest date length from the first case confirmed to zero growth in each city was 53 days, and the average and median lengths were 22 and 23 days, respectively. The 1st quartile and 3rd quartile were 15 and 28 days, respectively. When the length was 21 days, the frequency of the cities was highest at approximately 26 cities (6.3% in 413 cities). Further analysis of the epidemic length at the county level (regional unit) in Sichuan Province included a total of 103 counties that had epidemics. The results showed that there was no local case increase after March 4. The largest date length from the first confirmed case to zero growth was 33 days, with a mean and median of 9 days. The 1st quartile and 3rd quartile were 1 and 16 days, respectively. The highest city frequency of the date length was 1 day for approximately 31 cities (30% in 103 cities) (Figure 6). Taking the migration from Hubei during the two weeks before its lockdown as a possible high-risk factor, this study compared the provinces for which the migration rank was close to Sichuan Province and found that the peak time in Sichuan Province was earlier than elsewhere and the peak value was lower (Figure 7).
Global situation during the same period
The earliest confirmed cases outside China were reported in Thailand on January 13, followed by Japan (January 15), the Republic of Korea (January 20), the USA (January 21), Singapore and Vietnam (January 23). On February 21, four weeks after the first reported case in Korea, the number of confirmed cases began to rise rapidly. In addition, cases in Italy began to rise rapidly two weeks after the first case was reported there, and cases in Iran increased quickly in the second week after the first case report. Furthermore, the USA (5 weeks after the first case) and other European countries such as Spain (3 weeks after the first case), France (5 weeks after the first case), Germany (4 weeks after the first case), and Switzerland (1 week after the first case) began to see a rapid rise in infections. Overall, many European countries experienced that rise earlier than the USA. The length of time from the first case reported to the period of rapid rise in different countries varied. Most of them entered the rapid growth phase within 1 month (Figure 8). The confirmed cases reported in the USA increased sharply in weeks 8 to 9.
From the scatter plot of confirmed cases (top 10 countries), it was found that when the daily increase exceeded 20-30 cases, most countries entered an obvious increase stage (Figure 9). From the time point of the first daily increase of 20 cases, Germany, Spain, and Italy reached the peak of daily incremental rising after 28, 29, and 34 days, respectively, and then the daily increment showed a downward trend. The average daily increment of a 7-day rolling timeframe also showed an inflection point in Italy. The countries of France, the UK, and the USA appeared to peak at 36 days, 32 days, and 44 days, respectively. According to the regional division in the WHO report, many countries in the African region and Southeast Asia region were closer to the equator with slow growth or only at the early stage of infection spread. This study also analysed the top 10 countries of confirmed cases in each region. A change coefficients for each country was calculated by comparing the average daily increment of the final week to the previous week. The average change coefficient was 1.3 in the Eastern Mediterranean region and 1.25 in the region of the Americas. Besides, the change coefficient was 1.3 in Russian. Most European countries presented an inflection point or were at a plateau period in the fourth to fifth weeks but were falling slowly. The curve of cases in China dropped rapidly after the fifth week (Figure 10). On February 13-15, Hubei Province in China reported the clinical diagnosis. The number of clinical diagnosis cases was counted in the WHO COVID-19 Situation Report-28 with a statement that this was only applicable to Hubei Province. If the clinical diagnosis cases were not taken into account, China reached the peak of the daily increment of the epidemic in 18 days, and if such cases were taken into account, China peaked in 27 days. China has been subjected to imported cases since the eighth week, with basically a disappearance of local cases and the rapid rise of epidemics in other countries abroad (Figure 11). The overseas imported cases were first reported in Ningxia Province in China on February 26 and cumulatively included 1610 as of April 21, 21 of which were reported by Sichuan Province.