Close contact management mode and epidemiological characteristics analysis of COVID-19 in Chengdu, China


 As a potential source of infection, the risk of close contacts has gradually increased. Close contact management mode and epidemiological characteristics for 20,254 close contacts from January 22, 2020 to April 1, 2021 in Chengdu, China were analyzed. The relationship with index cases was mainly co-passengers (73.52%) and relatives (13.64%), and the frequency of contact was mainly occasional contact (68.31%). 277 (1.37%) close contacts who were converted into cases were mainly found in the first and second nucleic acid tests (58.27%), and the contact mode was mainly by sharing transportation (37.55%). In terms of close contact management time, both the supposed ((13.00±3.23) vs. (12.84 ±6.91)) and actual ((14.13±24.02) vs. (14.00±12.98)) isolation time in Chengdu were longer than that of outer cities (P<0.001). Tracing and management of close contacts is a two-way management measure, which requires the cooperation among departments. Enhancing existing monitoring and response capabilities can control the spread of the epidemic to a certain extent.


Introduction
The novel coronavirus disease 2019(COVID -19) has been spreading contrary to expectations since the rst report in December 2019, and its outbreak has posed a considerable threat to people around the world 1,2 . As of April 1, 2021, more than 129 million people in 223 countries have been infected, and 5 million have died. Due to the high incidence and mortality rate of initial COVID-19 cases and with the discovery and reporting of new cases, a large number of close contacts have been generated. It is urgent to establish a set of scienti c management systems for close contacts. The WHO and China's prevention and control measures have both cited the tracking and management of COVID-19 close contact to prevent and control the rapid spread of the disease 3,4 .
In contrast to SARS CoV, the transmission of COVID-19 occurs in the precursor period of mild illness and daily activities, which contributes to the transmission of infection 5,6 . If the potentially infected person does not timely controlled, they will go to work and travel, potentially spreading the virus to their contacts and even internationally 7,8 . Drawing on the experience of control measures for in uenza A(H 1 N 1 ) in 2009 9 , since the beginning outbreak of COVID-19, the local Centers for Disease Control and Prevention (CDC) has tracked all close contacts of con rmed COVID-19 cases, investigated contact history, tested nucleic acids, and tracked health status, so as to prevent new cases and local transmission 10 . Jing and colleagues 11 shared the prevention and control strategy of China, namely, actively discovering and identifying close contact, and then quarantining and monitoring them, which effectively controlled the epidemic, and enabled Wuhan to be lifted after 76 days of lockdown. This has also laid the foundation for important achievements in China's phased epidemic prevention and control.
Our previous research has shown that the main ways of contact between close contacts and index cases were domesticity and transportation 12 . Although the infection risk for individuals who share public transportation with infected cases was low, the number of people on public transportation was large and it was di cult to identify all close contacts, which may cause a large number of people being infected in this situation 13,14 . In addition, previous studies have shown that family members were also an important source of close contacts in the epidemiology of infectious disease. SARS-CoV-2 was more likely to spread among family members through coughing, sneezing, or direct contact with surfaces contaminated by the virus. They spend more time at home, resulting in a higher frequency and longer duration of unprotected exposure 13 , which leads to a higher risk of secondary infection through family contact 15,16 .
For instance, several studies have analyzed the epidemiological characteristics of close contact. However, these studies limited to the analysis of their regions, few studies analyzed the tracking and management of cross-regional personnel. The investigation presented here expect to analyze the relevant information and data of COVID-19 close contact in Chengdu, China, and those transferred to other regions. Through the description of contact route, distribution and isolation status of close contacts, we hope to understand the effectiveness and feasibility of existing close contact management measures in Chengdu, so as to provide reference for further prevention and control of the epidemic.

Methods
Source of close contact. The source of close contact was as described in our previous study 12 . In short, con dential data were mainly con rmed by epidemiological investigation of indexed cases or big-data platforms of public security and transportation. The CDC staff made a preliminary judgment rstly, and then input and maintained the information into the "COVID-19 Close Contact Information Management System of Chengdu". While close contacts outside Chengdu were managed in other cities, and this information was managed through cooperative letters, telephone inquiries, and other channels.
Research object. Epidemiological data of COVID-19 indexed cases were collected from January 22, 2020, to April 1, 2021. Close contacts who have been quarantined due to COVID-19 were selected, which including close contacts of related cases in Chengdu, and those investigated and managed outside Chengdu were also included.

Related de nitions
Close contact. According to the Chinese National Health Commission in the Protocol on Prevention and Control of COVID-19 (Edition 6) 17 , close contact was de ned as people who had not taken effective protection and had close contact with con rmed or suspected case (within 1 meter) within 2 days prior to illness onset or 2 days before sampling taken of asymptomatic carrier.
Personnel classi cations. Close contacts mainly include family members, relatives, friends, colleagues, classmates, medical workers, and service personnel. Relatives were mainly refer to family members who live with the case and other relatives who have close contact through meals and other forms.
Transportation contact refers to people who take the same transportation (airplane, train, car, ship, etc.) and have closed contact (within 1 meter), which include carers and companions (family, colleagues, friends, etc.) on the some transportation. Medical personnel include patients in the same hospital ward,

Results
Close contact control strategy. For the management of close contacts, once after the information was veri ed and con rmed, the main information was pushed to relevant districts or other regions.
Subsequently, the information was exchanged with the community, street and other departments. And medical staff, and hospital staff who had not taken appropriate protective measures. While other personnels include friends and colleagues contact with work or social life, individuals who provide business or life services for patients, and other occasional close contacts.
Close contact management. Close contacts were subject to centralized isolation for medical observation. The medical observation period was 14 days after the last unprotected contact with indexed cases. Close contacts at the centralized observation point should be strictly managed in a single room. Those who were not suitable for centralized isolation due to age, diseases, and other reasons were subject to home quarantine for medical observation.
During the observation period, the medical staff measured their body temperature twice a day and asked their health status, whether they had a fever, chills, dry cough, shortness of breath, dyspnea, chest tightness, conjunctival congestion, diarrhea, or abdominal pain, and lled in the "Medical Observation Registration Form for Close Contacts of COVID-19". On the 1 st and 14 th days of isolation, the medical staff collected nasal and/or throat swab samples from close contacts for nucleic acid testing, and the frequency of testing was adjusted according to the national scheme. If the swab test was positive or any symptoms were found, close contact was sent to a designated hospital for infection evaluation and diagnosis. Individuals whose samples were negative and showed no symptoms were exempted after the quarantine period expired.
Observation index. Close contact information of COVID-19 con rmed or suspected cases and asymptomatic carrier was collected, which including basic information about close contacts, relationship with related cases, observation methods, contact methods, exposure time, outcome, sampling and testing information.
Statistical Analysis. Retrospective data and relevant information was collected and recorded through the "COVID-19 Close Contact Information Management System of Chengdu" from January 22, 2020 to April 1, 2021. Excel was used to conduct preliminary data sorting. The average value was used to supplement the missing data in some indicators. Epidemiological and contact characteristics of close contacts were analyzed by SPSS of version 22.0 software ( IBM Corp, NY, USA). Continuous variables were expressed as mean ± S.D, and the differences between groups were expressed by independent sample t-test. While the categorical variables were expressed by count and percentage, the differences between groups were analyzed by chi-square. In addition, the odds ratio (OR) and its 95% con dence interval (95% CI) were calculated. P<0.05 was considered to be statistical signi cance. The spatial distribution of close contacts was describe by ArcGIS version 10.5 software (Environmental Systems Research Institute, Redlands, CA, USA) nally, the close contacts were transferred to the isolated hotel for centralized management. In this process, coordination with other departments should be maintained, and the implementation of close contact management and release of isolation by the hotel should be tracked. Then, relevant information was entered into the "COVID-19 Close Contact Information Management System of Chengdu". Finally, the system was maintained and managed by relevant staff (Fig. 1).
Close contact overview. Through the epidemiological investigation of indexed cases and the push of bigdata information, as of April 1, 2021, Chengdu has received a total of 23,462 close contacts. Eventually, a total of 20,254 close contacts were included in the management through the screening of suspected cases, tracking time over 14 days, and incomplete information, of which 17,952 were in Chengdu and 2,302 were outside Chengdu (Fig. 2).
Analysis on the COVID-19 in Chengdu. As of April 1, 2021, a total of 552 con rmed cases (67.20%) and 269 asymptomatic infection patients (32.80%) of COVID-19 were detected in Chengdu. The epidemic situation presented as three peaks (Fig. 3a). After April, 2020, the cases were mainly imported from abroad (79.54%), and close contacts showed the same trend. Close contacts were mainly from con rmed cases and asymptomatic carriers, accounting for 39.20% and 39.54%, respectively (Fig. 3b). The distribution of close contacts outside Chengdu showed dispersion from center to north and south ( Fig. 3c-d), while close contacts inside Chengdu showed a diffusion from center to surroundings (Fig. 3e).
Basic characteristics of close contact. Among the 20,254 close contacts, 68.40%were male, with an average age of (34.92±13.79) years, mainly in the 15-60-year-old age group (84.90%). No signi cant difference was found in the age groups between Chengdu and outside the city. The relationship between close contact and indexed cases was mainly co-passengers (73.52%) and relatives (13.64%), and the frequency of contact was mainly occasional (68.31%). A total of 18,385 (90.77%) close contacts were brought into centralized quarantine, and 19,977 (98.63%) close contacts were released from quarantine. During the centralized isolation period, a total of 277 (1.37%) closed contacts were transferred to con rmed cases or asymptomatic carriers (Table 1).
Close contacts converted to cases. A total of 277 cases were found among the close contacts, including 129 con rmed cases. The average age of the patients was 21.50 years-old (IQR: 5.00-33.75); 72.92% (202/277) were male. The average detection time of cases was 6.52 days, mainly in the rst and second nucleic acid tests (58.27%) (Fig. 4a). 96.75% of the cases were related to contact with con rmed cases or asymptomatic infected persons (Fig. 4c). The frequency of contact was mainly occasional contact (88.16%), which was mainly through sharing transportation (37.55%) (Fig. 4d-e). There was no signi cant difference in the incidence of close contacts between overseas imported patients (206/15,431=1.33%) and local cases (71/4,823=1.47%) (Fig. 4f). were longer, and the difference was statistically signi cant (P<0.001) (Fig. 5b-c).

Discussion
Fast access to cross-regional case and close contact information, and rapid identi cation and management of close contacts are of great signi cance to reduce the risk of infection in the population and control the spread of the epidemic 18 . Since the pandemic of COVID-19, Chengdu has carried out case investigation and disposal, close contact tracking, developed a close contact management system, and carried out the management mode of self-management and co-investigation for close contacts. Through the circulation of indexed cases, as of April 1, 2021, Chengdu tracked and managed 20,254 close contacts. Among these, 17,952 were in Chengdu, while other cities and provinces assisted in the management of 1,171 and 1,131, respectively.
Actually, it is a two-way management measure of close contacts tracking and management. First, it helps to understand the transition status of close contact. Once a close contact converted to case, this may facilitate the rapid management and control of him, especially for close contact transferred to other places. Second, it is helpful to improve the feedback e ciency of the state information of close contact, and provide a reference for close contact release in time. In this study, 277 cases were found in a timely and accurate manner through the implementation of centralized medical management of close contacts, which reduced the possibility of second-generation cases, and fully demonstrated that the current management model in Chengdu is effective and feasible. Since it taken a certain time to judge and analyze the close contact, and the information exchange with other regions different from conventional communication, there was a certain time difference in time feedback, which inevitably lead to the discovery time of close contact different in Chengdu and other regions ((1.00 ± 3.23) day vs. (1.16 ± 6.91) day).
Based on the WHO and China's prevention and control management measures, Chengdu strictly controlled the community management time before April 2021. Compared with some domestic region 19,20 , the control time of close contact in Chengdu extended by 7 -14 days, which increased the possibility of nding potential cases during the isolation observation period. At the same time, according to the epidemic situation, China has continuously adjusted the prevention and control plan and increased the frequency of nucleic acid detection. Through this step, we more effectively found the possible cases and prevented the generation of second-generation cases. In this study, the secondary generation rate of close contact was 1.37%, which was lower than that in relevant studies 10,19 . While, different from some studies 13,21 , we found that 37.55% of the second-generation cases were infected by taking vehicles together with indexed cases, which might be due to the different judgment methods of close contact.
Our data present that 73.52% of close contacts had shared transportation experience with indexed cases, and 88.43% (13,168/14,891) were passengers on the same ight. Since the month of April 2020, COVID-19 cases in Chengdu were mainly imported from abroad. Different from local close contacts, these people have a relatively single track of activity and enter strict centralized isolation management after preliminary nucleic acid detection 22 . Moreover, with the assistance of airlines and public security, these people were easier to control. Therefore, the case discovery time was lower than that of local cases ((1.26±3.01) day vs. (6.12±5.20) day). Among the 277 second-generation cases, 54.15%(150/277) cases were on the same ight and these may have been infected before departure. In addition, there were not many family members among the close contacts of immigrants, which may lead to a certain difference in the analysis of family members.
It is worth noting that in most family environments, especially when COVID-19 indicated cases in the incubation period, it is di cult to achieve physical alienation and avoid direct oral communication 23 . Compared with published studies in other countries, once a con rmed case was found, the local CDC immediately isolated family contacts at designated locations, which may reduce the secondary incidence of family contacts 24 . In our study, family contacts accounted for 13.64%, and the secondary generation rate was 1.34% (37/2762), which was lower than that in other cities in China (11.2%) 10 and the United States (10.5%) 25 . Due to the contact frequency between family members, reduction or neglect of personal protective equipment use, shared living and eating environment, and the continuous existence of SARS-CoV-2 on different surfaces, it provides favorable conditions for the mutual spread of family members 26,27 . Therefore, in terms of management, centralized isolation should be adopted as much as possible to avoid further transmission among family members 10 .
This investigation had a number of limitations. As mentioned in our previous research, it is mainly re ect in information offset 12 . The reason lies in objective forgetting, subjective deliberate concealment, and incomplete information recall. At the same time, due to the presence of close contacts for more than 14 days, which suggested the coinvestigation mechanism needs to be improved, so as to further avoid the possibility of infection risk. In addition, we could not rule out the possibility that infected close contacts have another unknown source of infection before isolation, which may affect our conclusion to a certain extent. Finally, with the change of prevention and control period, the determination and control methods of close contacts were also changed, but the overall difference was small. However, due to the occurrence of variant strains, such as the emergence of the Delta and Omicron strain, the range of close contacts has expanded, which may have a certain impact on secondary infection rate.
In summary, our present study demonstrated that close contacts were still dominated by fellow passengers and relatives. Based on the control of imported cases from abroad, the possibility of close contact of local cases still needs to be prevented and controlled. On the basis of increasing vaccination efforts, continuing to maintain the monitoring and response capacity of close contacts can control the further spread of the epidemic to a certain extent.  The process of close contacts screening   Close contacts converted to cases. a, Time detection of con rmed cases or asymptomatic infections. b, Division of close contacts transferred to cases. c, Source of transferred cases. d, Contact frequency with indexed cases. e, Relationship with indexed cases. f, Source distribution of indexed cases. Figure 5