key: cord-0708234-d9y31nlj authors: He, Junyu; Chen, Guangwei; Jiang, Yutong; Jin, Runjie; He, Mingjun; Shortridge, Ashton; Wu, Jiaping; Christakos, George title: Comparative Analysis of COVID-19 Transmission Patterns in Three Chinese Regions vs. South Korea,Italy and Iran date: 2020-04-14 journal: nan DOI: 10.1101/2020.04.09.20053223 sha: 8018f692b7ec7a86ead09e4fd42c09becd5e3364 doc_id: 708234 cord_uid: d9y31nlj Background: The outbreak of Coronavirus 2019 (COVID-19) began in January 2020 in the city of Wuhan (Hubei province, China). It took about 2 months for China to get this infectious disease under control in its epicenter at Wuhan. Since February 2020, COVID-19 has been spreading around the world, becoming widespread in a number of countries. The timing and nature of government actions in response to the pandemic has varied from country to country, and their role in affecting the spread of the disease has been debated. Method: The present study proposed a modified susceptible-exposed-infected-removed model (SEIR) model to perform a comparative analysis of the temporal progress of disease spread in six regions worldwide: three Chinese regions (Zhejiang, Guangdong and Xinjiang) vs. three countries (South Korea, Italy and Iran). For each region we developed detailed timelines of reported infections and outcomes, along with government-implemented measures to enforce social distancing. Simulations of the imposition of strong social distancing measures were used to evaluate the impact that these measures might have had on the duration and severity of COVID-19 outbreaks in the three countries. Results: The main results of this study are as follows: (a) an empirical COVID-19 growth law provides an excellent fit to the disease data in all study regions and potentially could be of more general validity; (b) significant differences exist in the spread characteristics of the disease among the three regions of China and between the three regions of China and the three countries; (c) under the control measures implemented in the Chinese regions (including the immediate quarantine of infected patients and their close contacts, and considerable restrictions on social contacts), the transmission rate of COVID-19 followed a modified normal distribution function, and it reached its peak after 1 to 2 days and then was reduced to zero 11, 11 and 18 days after a 1st-Level Response to Major Public Health Emergency was declared in Zhejiang, Guangdong and Xinjiang, respectively; moreover, the epidemic control times in Zhejiang, Guangdong and Xinjiang showed that the epidemic reached an "inflection point" after 9, 12 and 17 days, respectively, after a 1st-Level Response was issued; (d) an empirical COVID-19 law provided an excellent fit to the disease data in the six study regions, and the law can be potentially of more general validity; and (e) the curves of infected cases in South Korea, Italy and Iran would had been significantly flattened and shrunken at a relatively earlier stage of the epidemic if similar preventive measures as in the Chinese regions had been also taken in the above three countries on February 25th, February 25th and March 8th, respectively: the simulated maximum number of infected individuals in South Korea, Italy and Iran would had been 4480 cases (March 9th, 2020), 2335 cases (March 10th) and 6969 cases (March 20th), instead of the actual (reported) numbers of 7212 cases (March 9th), 8514 cases (March 10th, 2020) and 11466 cases (March 20th), respectively; moreover, up to March 29th, the simulated reduction in the accumulated number of infected cases would be 1585 for South Korea, 93490 for Italy and 23213 for Iran, respectively, accounting for 16.41% (South Korea), 95.70% (Italy) and 60.59% (Iran) of the accumulated number of actual reported infected cases. Conclusions: The implemented measures in China were very effective for controlling the spread of COVID-19. These measures should be taken as early as possible, including the early identification of all infection sources and eliminating transmission pathways. Subsequently, the number of infected cases can be controlled at a low level, and existing medical resources could be sufficient for maintaining higher cure rates and lower mortality rate compared to the current situations in these countries. The proposed model can account for these prevention and control measures by properly adjusting its parameters, it computes the corresponding variations in disease transmission rate during the outbreak period, and it can provide valuable information for public health decision-making purposes. comparisons are made between them, and the effects of prevention and control measures are 92 investigated. The results could valuably inform future strategies of epidemic prevention and control. The methodology of the present study consisted of two parts: (a) the disease spread was represented 118 mathematically by an epidemic model, and (b) the effects of the measures taken by the various 119 regions to control the disease were compared in terms of the model parameters. Many models linking the evolution of susceptible, infected and removed cases have been used in 121 the study of infectious disease distributions [22] [23] [24] [25] [26] . A modified susceptible-exposed-infected-122 removed model (SEIR) was used in this study to simulate the COVID-19 spread in the six study infected, cured and dead individuals at time t. The exposed and infected individuals constitute the 129 total number of affected individuals A(t)  E(t)  I (t) . The cured and dead individuals constitute the 130 total number of removed individuals R(t)  C(t)  D(t). The time-varying parameters  1 (t) and  2 (t) 131 denote the rate of COVID-19 transmission when an individual comes in contact with infected and 132 exposed individuals, respectively, the constant  represents the probability that the exposed 133 individuals become infected, and the time-varying parameters (t) and (t) denote the 134 cure rate and death rate, respectively. The SEIR model was developed specifically for this study 135 because the COVID-19 incubation period is about 7 days, and early in this period no symptoms are 136 detected, which means the infected but asymptomatic people will unknowingly infect others before 137 they develop any symptoms. According to the proposed SEIR model, the infection contact rate q(t)  (2) Alternatively, the R 0 (t)  q(t)S 0 represents the number of secondary infections at any time t in the 144 population caused by an initial primary infection. If one person has COVID-19, R 0 (t) determines 145 how many infections on average that person may cause. Then, the epidemic would be considered 146 under control at the time t ECT (epidemic control time, ECT) after which the R 0 (t) is consistently 147 smaller than L 0 (t), i.e., the inequality 149 holds for all t  t ECD . The SEIR model parameters above are very important because they determine (to varying extents 151 and within different contexts) the epidemic spread and its severity. Therefore, identifying good 152 estimates is critical for realistic simulation of infectious disease. In this work, we classified the SEIR regions and countries of interest were assumed to be large enough that they can be assumed to remain 170 All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.04.09.20053223 doi: medRxiv preprint constant during the epidemic. Some interesting inequalities hold, like S(t)  S 0 , i.e., the susceptible 171 population S(t) at any time is smaller than the initial number of susceptible cases ( S 0 ). (ii) In the SEIR equations, the exposure rate includes two parts: one part is proportional to the 173 contacts between susceptible and infected individuals (assessed by  1 ), whereas the other part is 174 proportional to the contacts between susceptible and exposed individuals (assessed by  2 ). Since 175 individuals that are confirmed as infected are immediately sent to hospitals for isolation and medical 176 treatment in the three regions of China, it is assumed that the infection transmission rate was always 177 smaller than the exposure transmission rate, i.e.,  1   2 . Specifically, following the current COVID-178 19 literature [27] , it is assumed that  2  5 1 , and that the probability that an exposed individual 179 becomes infected is   1 7 . The cured and dead individuals are removed at the rate (t)  (t) . In the 180 regions of interest, the values of  and  were zero at the beginning ( t  0), i.e.,  0   0  0 . 181 (iii) Given that the COVID-19 disease spread has been controlled in the three regions of China, 182 only data from the first 28-days were taken into account in the present study. For South Korea, Italy 183 and Iran, complete datasets covering the entire countries were considered, including areas in which 184 epidemic spread is still underway. (iv) Transmission, cure, and mortality rates are unknown and dynamic. We estimate these 186 parameters for each region using particle swarm optimization (PSO), a nonlinear computational fitting 187 procedure [28, 29] . A 6-days long moving window was introduced for computational model parameter 188 fitting (i.e., the infection transmission rate,  1 , the cure rate,  , and the mortality rate,  ). The 189 computational modeling procedure included the steps listed in Table 1 . The actual (empirical) values 190 of the three rates varied during the 6-days window period, but in Table 1 Step Description 1  The date when a COVID-19 case was first reported in a region was set as the time instance t  1.  The number of infected individuals in hospitals together with the accumulated numbers of cured and dead individuals during the period t to t  5 were inserted into the SEIR model.  The initial number of susceptible individuals was equal to the total number of people in the region, and the initial number of exposed individuals was set according to the reported number of people in close contact with infected individuals. 2  The particle swarm optimization (PSO) technique was used to simulate the SEIR model and obtain the PSO-fitted values of  1 ,  and  .  Given that these three parameters represent average transmission, cure and morality rates during the 6 day-window period, they were regarded as the real values during the period t 1 to t  2. 3  The parameters  1 ,  and  were inserted into the SEIR model to calculate the possible numbers of susceptible and exposed individuals at time t  2.  The moving the window was forwarded to the period t 1 to t  6, and the step 2 was repeated to obtain the new fitted values of  1 ,  and  . South Korea had a relatively low number of confirmed COVID-19 cases (Fig 1) . However, because of 220 high density gathering involving infected people, the number of confirmed cases increased a lot after February 24 th . Since February 28 th , the local authorities focused on curing two classes of patients 238 (severe and critically ill individuals), and on reducing the mortality rate by dividing the hospitals into 239 two groups, according to the above two patient classes. Since March 5 th , the main focus of the local 240 authorities has been on imported cases from abroad. All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.04.09.20053223 doi: medRxiv preprint Italy. In Italy (Fig 1) , the first COVID-19 infected individual was confirmed on January 31 st , 246 2020, i.e., 10 days later than in Guangdong. In the following two days, the Italian government Xinjiang. The total number of confirmed COVID-19 cases during the study period was low (76), 253 but with a relatively high mortality of 3.95% (Fig 1) . On January 25 th , the 1 st -level Response to Major Public Health Emergency was issued in the region. The disease control and prevention measures were 255 shifted from the city to the county level and then to the rural areas. Iran. The first COVID-19 case was reported on February 19 th , 2020 (Fig 1) . One day later, the It was found that during the critical growth period of the epidemic (Fig 2b) , the COVID-19 278 variation in all six regions obeyed the log-linear relationship All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.04.09.20053223 doi: medRxiv preprint the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.04.09.20053223 doi: medRxiv preprint together with the corresponding fitted lines and the 95% confidence interval (shadows); the bottom figures show 323 the reported number of infected cases (red dots), accumulated numbers of cured cases (green dots) and dead 324 cases (gray dots) together with the corresponding SEIR-produced values (lines); graphs for Italy also presents the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is The numerical results are shown in Fig 6. As can be seen in Fig 6, the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.04.09.20053223 doi: medRxiv preprint Emergency was declared at Zhejiang, Guangdong and Xinjiang, respectively, and 3, 5 and 3 days 445 before the number of infected cases reached the peak values. This means, that the COVID-19 should 446 have been under control in these regions after theses t ECT dates (the epidemic reached an "inflection 447 point" at the corresponding dates). 448 449 450 Figure 6 : The trends of R 0 (t) and L 0 (t) variations in each of the six regions (for better visualization, we set 451 R 0 (t)  1 whenever the R 0 (t) value was computed to be greater than 1). Regarding South Korea, Italy and Iran, both the actual and simulated R 0 (t) and L 0 (t) curves are 454 plotted in Fig 6. The actual R 0 (t) and L 0 (t) curves are more irregular than in the three Chinese the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.04.09.20053223 doi: medRxiv preprint point). In Italy and Iran the R 0 (t) and L 0 (t) curves continued fluctuating around each other since 462 March 20 th , 2020, and no definite t ECT dates could be determined at least until the end of March, 2020. Overall, there is a profound difference between the variation patterns of the R 0 (t) and L 0 (t) curves 464 observed in the three Chinese regions and those observed in the three other countries: in the Chinese 465 regions there are well defined t ECT dates, which is not the case in South Korea, Italy and Iran (i.e., the 466 R 0 (t) and L 0 (t) curves fluctuate around each other). If, on the other hand, it is assumed that the measures in Zhejiang, Guangdong and Xinjiang were 468 implemented in South Korea, Italy and Iran, respectively, the following simulated results are obtained: in South Korea the simulated epidemic control date will be t ECT  March 2 rd , 2020, i.e., the same as the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.04.09.20053223 doi: medRxiv preprint Furthermore, on February 1 st Xinjiang intensified its implementation of the prevention and control in 508 rural areas. All these actions may be reflected in the fluctuations of the transmission curve (Fig 5) . The disease transmission rates in South Korea, Italy and Iran during the study period experienced 512 much higher peak values (0.278, 0.262 and 0.142, respectively) than in Zhejiang, Guangdong and 513 Xinjiang (0.096, 0.075 and 0.053, respectively), suggesting that some special events took place that 514 caused these higher peaks in the three countries. Specifically, the Daegu Church gathering in Daegu 515 city, and the infections in Daenam hospital (Gyeongsangbuk-do province) were found to be two major 516 events that produced large infected populations, which comprised 82% of the accumulated infected 517 cases on March 29 th , 2020. In the same day, it was reported that 84.1% of the accumulated infected Specifically, the mortality rates during the first 28 days of the COVID-19 outbreak in Zhejiang, Guangdong and Xinjiang were 0%, 0.15%, 1.32%, respectively. On the other hand, the mortality rates 545 in South Korea, Italy and Iran were 1.64%, 11.03% and 6.89%, respectively, during the period since 546 the first reported case date to March 29 th . In addition, the difference between the simulated and the the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is Table S6 . Lastly, the best way for an individual to contribute is 575 to be self-isolated and pay more attention to personal hygiene. Some basic guidelines are as follows: 576 wash hands frequently, wear a facemask when going outside, and stay at home for a 14-day self-577 isolation period. can lead to a larger cure rate and a smaller mortality rate. The temporal moving window scheme helps characterize the dynamics of the COVID-19 589 transmission rates in these regions during the disease outbreak, which can also benefit the public Korea, Italy and Iran, respectively, accounting for 16.41%, 95.70% and 60.59% of the accumulated 601 number of infected cases), including the flattening and shrinking of the infected case curves, which 602 would also had peaked at an earlier time than the observed curves in these countries. This means that 603 the disease could have been controlled faster and more efficiently in these countries than it actually 604 happened. Beyond being used to quantitatively describe the current transmission conditions in the six study 606 regions worldwide, the proposed approach has the potential to monitor disease transmission rates and 607 predict disease case numbers in future situations. Health authorities could assimilate this valuable 608 information into their disease prevention and control decision-making process. Future work would 609 also focus on employing the dynamic transmission rate to forecast any trends in the numbers of Text S1. Description of the six study areas. Zhejiang province is located in eastern China (27°06′N-31°11′N and 118°01′E-123°10′E) , and has a 705 territory of approximately 0.10 million km 2 and a population of 54 million [4] . Residents between 15-706 59 years old and older than 60 years account, respectively for about 72.90% and 13.89% of the the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is More specifically, the mean temperature ranges from 2-10℃ during January and 23-26℃ during July. In 2018, about 60 million people lived in Italy, among whom 22.6% were older than 65 years, and where a 1 , a 2 and a 3 denote the scale, shape and location parameters, respectively. Additionally, the 751 cured and the mortality rates were fitted to quadratic equations as 752 where b 1 and c 1 are constants, and the b 2 , b 3 , c 2 and c 3 are parameters. All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.04.09.20053223 doi: medRxiv preprint Using the "nlinfit.m" function in the MATLAB package, the parameters of the modified normal 756 distribution equation and quadratic models were calculated as shown in Table S1 . the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.04.09.20053223 doi: medRxiv preprint transportation. Then, we estimated that, generally, the drop in passenger travel by the bus system 788 could be even more than that by rail or air. Hospital arrangements  The medical authorities assigned specific hospitals for treating and quarantining infected individuals.  It arranged for hospital re-construction, the organization of building sections for increasing the number of isolation areas and beds (e.g., Zhejiang used 95 hospitals to accommodate infected individuals, and 335 outpatient clinics for patients who have fever symptoms; after the most severe disease period had passed, the number of hospitals decreased to 30, and all remaining infected cases were put there for better disease management and control). Patients' histories  The medical society carried out systematic epidemiologic investigations.  It archived the individual histories and potential close contacts of the infected patients.  Confirmed infected and suspected individuals were continuously reported, including their locations and historical movement, so that disease control could be implemented accordingly. Online assistance  The medical society created an online diagnostic system for normal patients, which can prevent cross-infection, decrease the health risk, and reduce the turnover.  This system could also be used to collect information on locate infected individuals and their traveling histories.  The system operated 24 hours per day lines for emergency issues. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.04.09.20053223 doi: medRxiv preprint Immediate restrictions  Public places (schools, parks, restaurants, malls, cinemas, libraries etc.) were closed.  Public and family gatherings (in-situ meetings, cultural and sports activities, family feasts etc.) were cancelled.  Public transportation (buses, ships, planes) between provinces, cities and counties were suspended.  Places closely associated with people's everyday lives were strictly monitored (e.g., food markets were disinfected twice a day and the body temperatures of all people entering them were measured).  Check-points were setup at all city and county entrances (e.g., at various points of highway exits) and at all train station and airport exits.  Police and medical staff monitored the body temperatures of all people passing through check-points, and their origins were registered (e.g., people who lived in epidemicaffected areas or whose temperature was above normal were sent to hospitals for further testing, diagnosing or quarantining).  During the most severe epidemic period, only people with local identification cards and certified local workers could pass the check-points at the county and city entrances. Resources  Special funds were timely allocated to control local disease spread, prepare medical facilities, purchase protective equipment, and further respond to disease transmissions (e.g., by January 27 th , Zhejiang authorities had provided more than 1.2 billion yuan to support disease control actions).  Sufficient staff was timely transferred to support local communities and medical systems in order to prevent disease spread (e.g., by January 26 th , i.e. 3 days after Zhejiang announced a 1 st -Level Response to Major Public Health Emergency, more than 330 thousand officers contributed to disease control together with community managers).  If the number of infected people was too large and there were not enough hospital beds available, public places (e.g., gymnasiums, theatres and music-halls) were restructured to serve as shelter hospitals.  Additional specific guidelines were setup to separate household waste from medical facilities waste and other relevant places hosting patients. Health-codes  A 3-color health code was used to manage the local flow of people by identifying the condition of people entering a local county or city (Zhejiang was the first province to use this code).  People applied for health code in specific cities based on their location, current health condition and the places they had been during the past two weeks.  A red health code meant that an individual was infected, or suspected, or had closely contacted an infected (or suspected) individual, or had possibly contacted an infected (or suspected) individual, or came from a severely epidemic-affected region. In all cases, people were quarantined for 14 days, reporting their health condition every day. After that, an individual's code would turn into green.  A yellow health code meant that one came from a severe disease region or had potentially contacted someone from that region. In this case, one needed to be quarantined for 7 days, having to report one's health condition during each day. After that, an individual's code would turn to green.  A green health code meant that the individual had very low risk of getting the disease and was free to travel.  Schools and companies also used this code to monitoring the health condition of students, teachers and staff.  Local governments, schools and companies may have more restricted codes for allowing their staff to return to these places and maintain their basic functions. Asymptotic treatment  A sufficient number of places were arranged for asymptomatic people who had, though, been in close contact with infected individuals, and the necessary health staff was assigned to these places.  Hotels, public houses and school dormitories were requisitioned for these people reside in.  Health-care teams (each consisting of 1 officer, 1 doctor and 2 nurses) were put together to take care of 50-people groups. The doctor and nurses monitored people's temperature and overall condition 2-3 times a day, and those whose body temperature exceeded a threshold or exhibited other disease symptoms were sent to hospitals for further treatment. Scientific indexes  A scientific index was used to assess each county's or city's health risk, and subsequent control strategies were implemented at each county or city. All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.04.09.20053223 doi: medRxiv preprint  Zhejiang used 5-color maps to assess each county's health risk by considering the accumulated number of infected individual, ratio of the number of local patients over total number of patients, clustered cases, the number of continuous days that no new infected individuals were detected (e.g., Zhejiang announced that the disease control focus switched from preventing the import of infected individuals to avoiding local disease spread, and its efforts switched from disease control to both disease control and ensuring economic development at the same time). Legal punishment  People whose actions threatened public health were legally punished (e.g., people who hide their health condition, violated local disease control measures, or concealed that they were coming from or traveling to areas at a severe disease state). Community health duties  Among the community managers' duties were to make sure that everyone who was in a public place was wearing a face-mask.  They were also instructed to continuously assess their community's status, including the daily health condition and travel history of every community member.  Managers classified their community members into four groups: of low concern were members who completed the isolation period without any symptoms, of current focus were member still within the isolation period, newly arrived (i.e., today) members from other places should be body temperature tested and closely monitored, and early prevention measures should be taken for the member who arrived recently (i.e., yesterday or earlier) from other places.  During the isolation period, the managers visited their community members and measured their body temperature 3-times a day. If a member's temperature was above the threshold or some typical symptoms of illness occurred, the member was sent to the nearest designated hospital, and the closely contacted members should be homequarantined for further monitoring (e.g., in a Hangzhou city community, a group of 339 managers visited 12025 families in 334 buildings, collected detailed information about them, and 122 disease-related clues were detected for the isolated families).  Check-points were setup to restrict community access 24 hours per day. Only community members could pass through the check-points after body temperature testing.  During the most severe disease period, one member per family could leave the community every 2 days to purchase food. Community assistance  Managers assured satisfactory living conditions for their community members.  During the isolation period, the community assisted all its members, especially senior ones, with their daily food purchases.  A 24 hours a day emergency response system was established at each community. Disinfection  Community managers arranged for the disinfection of public areas, physical exercise facilities, garbage centers, elevators, apartment entrances etc. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.04.09.20053223 doi: medRxiv preprint Companies  High tech industry used the latest technology (big data analysis, cloud computing etc.) to locate infected and suspected individuals.  The industry help built online health reporting and monitoring systems, such as the color health code system mentioned above.  It donated funds and supplies for the medical community, the local governments and for emergency health management (e.g., the Alibaba company donated 1 billion yuan).  The industry adopted some of its operations to the environment created by the epidemic (e.g., it used online procedures to sign its contracts).  Some companies switched their production lines to manufacture medical supplies (e.g., before the COVID-19 outbreak in Wenzhou city there was only one company manufacturing face masks, but by February 25 th there were more than 25 companies capable of manufacturing face masks with the production being increased 20 times).  Companies setup a health reporting and monitoring system for their own staff, which is similar to that of local governments and communities (e.g., at a company's entrance the staff's body temperature was measured and their hands disinfected). Society  Young people and healthy adults volunteered for a variety of jobs.  They supported local governments working at highways, train stations, airports etc.  They helped monitor the health condition of community members, supported the medical infrastructure, and cleaned facilities etc. (e.g., by Jan. 26, 2020 more than 10 thousand young volunteers contributed to disease control, and their number increased to nearly 40 thousand by the end of January).  Charity groups, local groups and the red cross society raised funds and medical supplies for disease control and prevention in the country and worldwide. A pneumonia outbreak associated with a new 624 coronavirus of probable bat origin Clinical characteristics of 2019 novel coronavirus infection in China COVID-19: Gastrointestinal manifestations and potential fecal-oral transmission Structural basis for the recognition of the SARS-CoV-2 by full-length human 630 ACE2 Clinical features of patients infected with 2019 novel coronavirus in Wuhan Pathological findings of COVID-19 associated with acute respiratory distress 634 syndrome. The Lancet respiratory medicine National Health Commission of the People's Republic of China On the origin and continuing evolution of SARS-CoV-2 The effect of travel restrictions on the spread of the 2019 novel 641 coronavirus (COVID-19) outbreak. Science High resolution climatic regionalization of typical cash crops in Guangdong Ministry of Foreign Affairs of the People's Republic of China Stochastic amplification in epidemics Infectious Diseases of Humans: Dynamics and Control Spatiotemporal infectious disease modeling: a BME-SIR approach Modeling of space-time infectious 662 disease spread under conditions of uncertainty An online spatiotemporal prediction model 665 for dengue fever epidemic in Kaohsiung (Taiwan) Modified SEIR and AI prediction of the epidemics trend of COVID-19 in 667 China under public health interventions A review of particle swarm optimization. Part I: background 671 and development Interdisciplinary Public Health 677 Reasoning and Epidemic Modelling: The Case of Black Death New space-time perspectives on the propagation characteristics of the 679 Black Death epidemic and its relation to bubonic plague. Stochastic Environmental Research and Risk 680 Assessment Probabilistic logic analysis of 682 the highly heterogeneous spatiotemporal HFRS incidence distribution in Heilongjiang province (China) 683 during 2005-2013 Regional Gap in Socio-economic Development of Xinjiang Discussion on correlation between vegetation cover change and climate and human factor in 820 Xinjiang area High resolution climatic regionalization of typical cash crops in Guangdong People's daily online