key: cord-0845110-ragcpbl6
authors: Zhang, Juanjuan; Litvinova, Maria; Wang, Wei; Wang, Yan; Deng, Xiaowei; Chen, Xinghui; Li, Mei; Zheng, Wen; Yi, Lan; Chen, Xinhua; Wu, Qianhui; Liang, Yuxia; Wang, Xiling; Yang, Juan; Sun, Kaiyuan; Longini, Ira M.; Halloran, M. Elizabeth; Wu, Peng; Cowling, Benjamin J.; Merler, Stefano; Viboud, Cecile; Vespignani, Alessandro; Ajelli, Marco; Yu, Hongjie
title: Evolving epidemiology of novel coronavirus diseases 2019 and possible interruption of local transmission outside Hubei Province in China: a descriptive and modeling study
date: 2020-02-23
journal: medRxiv : the preprint server for health sciences
DOI: 10.1101/2020.02.21.20026328
sha: 87b89b038b28d2ba0bcfd2203845bf15f922c1b5
doc_id: 845110
cord_uid: ragcpbl6

Background The COVID-19 epidemic originated in Wuhan City of Hubei Province in December 2019 and has spread throughout China. Understanding the fast evolving epidemiology and transmission dynamics of the outbreak beyond Hubei would provide timely information to guide intervention policy. Methods We collected individual information on 8,579 laboratory-confirmed cases from official publically sources reported outside Hubei in mainland China, as of February 17, 2020. We estimated the temporal variation of the demographic characteristics of cases and key time-to-event intervals. We used a Bayesian approach to estimate the dynamics of the net reproduction number (Rt) at the provincial level. Results The median age of the cases was 44 years, with an increasing of cases in younger age groups and the elderly as the epidemic progressed. The delay from symptom onset to hospital admission decreased from 4.4 days (95%CI: 0.0-14.0) until January 27 to 2.6 days (0.0-9.0) from January 28 to February 17. The mean incubation period was estimated at 5.2 days (1.8-12.4) and the mean serial interval at 5.1 days (1.3-11.6). The epidemic dynamics in provinces outside Hubei was highly variable, but consistently included a mix of case importations and local transmission. We estimate that the epidemic was self-sustained for less than three weeks with Rt reaching peaks between 1.40 (1.04-1.85) in Shenzhen City of Guangdong Province and 2.17 (1.69-2.76) in Shandong Province. In all the analyzed locations (n=10) Rt was estimated to be below the epidemic threshold since the end of January. Conclusion Our findings suggest that the strict containment measures and movement restrictions in place may contribute to the interruption of local COVID-19 transmission outside Hubei Province. The shorter serial interval estimated here implies that transmissibility is not as high as initial estimates suggested.

Since December 2019, an increasing number of atypical pneumonia cases caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) have been reported in Wuhan City of Hubei Province, China 1 In the present study, we aim to describe the epidemiological characteristics of COVID-19 outbreak after 50 days of detection in the provinces outside Hubei. We also estimate changes in key time-to-event distributions and reproduction numbers to assess whether the strict control measures have been able to slow transmission.

Since the outbreak of atypical pneumonia cases was detected in Wuhan at the end of December 2019, the Chinese Center for Disease Control and Prevention (China CDC) has launched a new surveillance system, first in Wuhan, then extended to the entire country, to record information on COVID-19 cases. Case definitions for suspected cases and laboratory-confirmed cases, and the description of the surveillance system have been published 1 and reported elsewhere 5 . Details are summarized in Appendix (page 2).

In the first version of "Guideline on diagnosis and treatment of novel coronavirus infected pneumonia (NCIP)" issued by China CDC on January 15, 2020, a suspected NCIP case was defined as pneumonia that fulfilled clinical criteria (fever; radiographic findings of pneumonia; or normal or reduced white blood 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/10.1101/2020.02. 21.20026328 doi: medRxiv preprint cell count, or reduced lymphocyte count at early onset of symptoms) and had an epidemiologic link to the Huanan Seafood Wholesale Market in Wuhan or travel to Wuhan within 14 days before symptom onset. The subsequent two versions of case definitions issued on January 18 and 22, separately, removed one of the clinical criteria (i.e. no reduction in symptoms after antimicrobial treatment for 3 days following standard clinical guidelines) to accelerate identification of cases and revised the epidemiological link (e.g. a travel history to Wuhan or direct contact with patients from Wuhan who had fever or respiratory symptoms within 14 days of symptom onset, or to be a potential case in a cluster). In the fourth version issued on January 27, the clinical criteria was loosened to meet any two of the three prior clinical criteria (i.e. fever; radiographic findings of pneumonia; normal or reduced white blood cell count, or reduced lymphocyte count at early stage of illness) while the epidemiological link added one more criteria (e.g. epidemiological link with confirmed COVID-19 case). In the fifth version issued on February 4, clinically-diagnosed cases were defined as suspected cases with radiographic findings of pneumonia, to be used exclusively in Hubei 7 .

Daily aggregated data on the number of cumulative cases in mainland China were extracted from the official websites of national, provincial, and municipal Health Commissions (see Tab. S1 in Appendix).

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/10.1101/2020.02. 21.20026328 doi: medRxiv preprint Individual records on laboratory-confirmed COVID-19 cases were collected from two official publicly available sources, including: (i) websites of national, provincial, and municipal Health Commission; (ii) websites of national and local government affiliated medias. Individual information was extracted and entered into a structured database comprising demography, exposure and travel history, timeline from exposure, symptom onset, hospital admission, and date of official announcement (reporting date). Each individual record was extracted and entered by three coauthors and was cross-checked to ensure data accuracy.

Conflicting information was resolved based on the data source (i). In this study, we used information on age, sex, location of detection, exposure history, dates of symptom onset, hospital admission, and official announcement. Details on the collection of individual data and assessment of completeness of variables used in the study are provided in Tab. S1-S2 in Appendix.

We also validated our individual records against the official line lists obtained from the websites of Shandong Provincial Health Commission, Shenzhen Municipal Health Commission, and Hunan Provincial Health Commission for the key variables used in this study (see Tab. S3 in Appendix).

We restricted analyses to the provinces other than Hubei where the majority of 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/10.1101/2020.02. 21.20026328 doi: medRxiv preprint our individual records are available (98%, 8,579/8,738) . We used the date of one key change in case definition to divide the epidemic into two time periods. The first period runs from the emergence of COVID-19 in Wuhan through to January 27, when the definition of suspected cases in the fourth version issued by China CDC was loosened to capture milder cases (radiographic evidence of pneumonia was no longer necessary, and epidemiological links extended to contact with confirmed cases). The second period runs from January 28 through to February 17. We performed statistical analyses of demographic and epidemiological characteristics of confirmed cases stratified by the two epidemic periods.

We estimated key epidemiologic parameters on time-to-event distributions for COVID-19 cases, including symptom onset to first healthcare consultation, hospital admission, and official announcement. We estimated the time from infection to symptom onset (incubation period) by analyzing COVID-19 cases with asserted epidemiological links (clusters) identified by prospective contact tracing. The date of presumed infection was estimated from cases' history of exposure, excluding cases with exposure to Wuhan. When multiple exposures were reported, we considered the interval between the first and last recorded dates of exposure. We fitted three parametric distributions (Weibull, gamma, and lognormal) to time-to-event data and selected the best fit based on the minimum Akaike information criterion (AIC). 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/10.1101/2020.02. 21.20026328 doi: medRxiv preprint We analyzed clusters of COVID-19 cases with epidemiological links asserted by prospective contact tracing to estimate the serial interval, which is defined as the interval between onset of symptoms in a primary case and the onset of symptoms in secondary cases generated by that primary case. The serial interval is then estimated by fitting a gamma distribution to the lag between the dates of symptom onset of the primary and secondary cases (with no travel history to/from Wuhan/Hubei) across all clusters.

By leveraging the estimated distribution of the serial interval, we provide estimates of the net reproduction number (Rt), which is the average number of secondary cases generated by a typical primary case at time t. Consecutive generations of cases arise after a period measured by the serial interval or by the generation time. We use a Bayesian approach to estimate Rt from the time series of symptom onset dates and the distribution of the serial interval 8, 9 . For this analysis, the last 9 days of the dataset were not considered to deal with the possible incompleteness of the dataset due to reporting delays. Details on the methodology are reported in the Appendix (page 18).

Statistical analyses were performed with R (version 3.6.0). The reproduction number was estimated from a code in C language written by the authors and available upon request. 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/10.1101/2020.02. 21.20026328 doi: medRxiv preprint

The study was approved by the Institutional review board from School of Public Health, Fudan University (IRB#2020-02-0802). All data were collected from publicly available sources and did not contain any personal information. The median age of the cases was 44 years (range, 1 month to 97 years), with an increasing proportion of cases in age groups below 18 years old (p<0.001) and above 64 years old (p<0.001) observed in the more recent time period. However, 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/10.1101/2020.02. 21.20026328 doi: medRxiv preprint the proportion of cases among individuals aged <18 years remains low (5%), see Tab. 1. The proportion of male cases decreased between the two epidemic periods (p<0.001), from 54% to 49%. As of February 17, 51% of cases were male (see Fig. S2 in Appendix).

The presence of at least one known exposure was reported by 77% of cases. In A similar decreasing trend over time was observed for the interval from symptom onset to first healthcare consultation as well (Tab. 2).

We analyzed the time interval from exposure to illness onset for 49 cases (with no travel history to/from Wuhan/Hubei ) identified by prospective contact tracing in 37 clusters. We estimated a mean incubation period of 5.2 days (95%CI:

1.8-12.4), with the 95th percentile of the distribution at 10.5 days. We find that the distribution of the incubation period is well approximated by a lognormal distribution (see Fig. S3 and Tab. S9 in Appendix).

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/10.1101/2020.02. 21.20026328 doi: medRxiv preprint We analyzed the time interval between symptom onset in 35 secondary cases and symptom onset in 28 corresponding primary cases. These secondary cases were identified by prospective contact tracing of 28 clusters and had no travel history to Wuhan/Hubei (see Appendix, Fig. S4 ). One case who reported the onset of symptoms on the same day as the index case was dropped from the analysis (see Appendix, Tab. S10 and Fig. S5 for a sensitivity analysis) . Moreover, as we cannot exclude that a fraction of these secondary cases had a previous exposure to an unidentified infection source, we performed a sensitivity analysis by different levels of data censoring (see Appendix, pages 16 and 17). We estimate the serial interval to follow a gamma distribution with a mean of 5.1 days (95%CI: 1.3-11.6). A comparison between the distribution of the incubation period and of the serial interval is reported in Fig. 2 .

The transmission dynamics of COVID-19 are extremely different among the provinces outside Hubei. Here we report the results for a city (Shenzhen -a major city with more than 12 million inhabitants in Guangdong Province) and two Provinces (Hunan and Shandong) for which we have validated our individual records against the full official line list compiled by the respective local health authorities. Results for other seven locations are included in Appendix (Fig. S6 and S7). Despite the selected locations are all among the provinces reporting the largest number of COVID-19 cases as of February 17, 2020 3 , they show highly 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 to January 24) with a maximum value of 1.40 (95%CI: 1.04-1.85) when considering a serial interval of 5.9 days on average. The outbreak was mostly sustained by cases with travel history to/from Wuhan/Hubei (Fig. 3A) . In Hunan Province, we estimated Rt to be above the epidemic threshold for about two weeks with a peak value of 1.58 (95%CI: 1.29-1.92) considering an average serial interval of 5.9 days (Fig. 3B) . Shandong Province showed an longer period (more than two weeks) characterized by sustained local transmission and a larger peak value of Rt: 2.17 (95%CI: 1.69-2.76) considering an average serial interval of 5.9 days (Fig. 3C ). In these three locations, Rt remained steadily below the epidemic threshold since the end of January 2020 (Fig. 3A-C) . In general, we found that in all the analyzed areas (8 out of 9 most affected provinces outside Hubei and one additional location), Rt was below the epidemic threshold as of February 8.

We have provided an assessment of the rapidly changing epidemiology and transmission dynamics of the COVID-19 outbreak beyond Hubei Province in mainland China. We found significant differences in the epidemiology of COVID-19 as the epidemic continues to spread across China. Most importantly, as of February 8, 2020, we estimate the net reproduction number to be below the epidemic threshold in the provinces that have reported the largest number of 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 At the beginning of the epidemic, COVID-19 cases were mostly observed among the elderly 10 . As the epidemic progressed, we observed a shift towards younger (<18 years) and older (65+ years) cases. Since January 28, 2020, however, the proportion of confirmed cases aged <18 years is still only about 5%, although this age group represents approximately 20% of the Chinese population. From the data available here, it is not possible to distinguish whether younger individuals have a reduced risk of infection or a propensity for milder clinical outcome given infection (thus resulting in a lower risk of detection). It should also be considered that schools in China were closed for most of the epidemic due to the 2020 Chinese New Year holidays 11 . It is unclear whether nationwide school breaks could have contributed to the low proportion of confirmed COVID-19 cases among school-age individuals, and whether schools reopening will lead to a change in the transmission patterns of COVID-19.

At the very beginning of the epidemic, a disproportionate faction of COVID-19 cases were male 10 . As of February 17, 2020, however, we estimate that about the 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/10.1101/2020.02.21.20026328 doi: medRxiv preprint same number of cases are observed among males and females (51% of cases are male). This suggests either differential exposure by sex occurring at the beginning of the epidemic (most of cases reported possible exposure to live markets 12, 13 ) or possible bias in the detection of the first few cases.

In the provinces outside Hubei we estimate the incubation period to be 5.2 days on average, in agreement with previous studies 5, 14 . The 95 th percentile of the distribution (10.5 days) suggests the importance of contact tracing and medical observations for those with long incubation period.

In the provinces outside Hubei we estimate the serial interval to be on average 5.1 days (95%CI: 1.3-11.6). This estimation is considerably shorter than the preliminary estimate derived from the analysis of six serial intervals in Wuhan 5 .

This may be linked to the short time interval from symptom onset to hospital admission (about 2.6 days on average since January 28) we have estimated for mainland China outside Hubei, as compared to what was measured in Wuhan in the early phase of the outbreak 5 , which could have prevented longer serial intervals from being observed. It should however be stressed that, as suggested by a theoretical study 8 , the serial interval estimated from the analysis of household clusters may be up to 20% shorter than the true value.

We estimate the serial interval to have about the same length as the incubation 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 We estimate the net reproduction number to have followed markedly different 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/10.1101/2020.02. 21.20026328 doi: medRxiv preprint patterns in different Chinese provinces. We found that in the analyzed provinces (selected among those mostly affected by the COVID-19 epidemic outside Hubei), the epidemic was self-sustained only for short periods of time (no more than 3 weeks). Most importantly, we estimated that since the end of January 2020, Rt is below the epidemic threshold in all the analyzed provinces in mainland China other than Hubei. This is confirmed by the gradual decrease in the number of It is important to stress that this study is affected by the usual limitations 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/10.1101/2020.02. 21.20026328 doi: medRxiv preprint pertaining to the data analysis of rapidly evolving infectious disease outbreaks.

The statistical analysis could therefore include biases due to case ascertainment and non-homogenous 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 b. Sample size may be different from the sum of the two periods as it includes also cases having not recorded date of symptom onset (which is used to the classification into temporal periods).

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/10.1101/2020.02. 21.20026328 doi: medRxiv preprint Estimated net reproduction number (Rt) over 4-day moving average. We excluded the last 9 days of data (i.e., data after February 8, 2020) to account for reporting delays. In fact, we estimated the 90 th percentile of the distribution of the time from onset to announcement in mainland China outside Hubei Province during period 2 of the epidemic to be 9.0 days. B Same as A, but for Hunan Province. C Same as A, but for Shandong Province.

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/10.1101/2020.02. 21.20026328 doi: medRxiv preprint A B 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/10.1101/2020.02. 21 

Chinese Center for Disease Control and Prevention. Epidemic update and risk assessment of 2019 Novel Coronavirus

National Health Commission of the People's Republic of China

World Health Organization. Coronavirus disease 2019 (COVID-19) Situation Report

Emergency Committee regarding the outbreak of novel coronavirus (2019-nCoV)

Early Transmission Dynamics in Wuhan, China, of Novel Coronavirus-Infected Pneumonia

The Novel Coronavirus Pneumonia Emergency Response Epidemiology Team. The Epidemiological Characteristics of an Outbreak of 2019 Novel Coronavirus Diseases (COVID-19) -China

Diagnosis and treatment guideline on pneumonia infection with

Measurability of the epidemic reproduction number in data-driven contact networks

Ebola virus disease in West Africa--the first 9 months of the epidemic and forward projections

Epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in Wuhan, China: a descriptive study

Ministry of Education of the People's Republic of China. Notice of the Ministry of Education on the postponement of the spring semester of 2020

Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China

Real-time tentative assessment of the epidemiological characteristics of novel coronavirus infections in Wuhan, China, as at 22

Incubation period of 2019 novel coronavirus (2019-nCoV) infections among travellers from Wuhan, China

Effects of School Closures

Reactive school closure weakens the network of social interactions and reduces the spread of influenza

People's Government of Hubei Province. Notice of the General Office of the People's Government of Hubei Province on Extending the Spring Festival Holiday of 2020

2019-nCoV: new challenges from coronavirus

A familial cluster of infection associated with the 2019 novel coronavirus indicating potential person-to-person transmission during the incubation period

SARS-CoV-2 Viral Load in Upper Respiratory Specimens of Infected Patients

National Health Commission of the People's Republic of China. Inclusion of 2019 novel coronavirus diseases (COVID-19) into statutory infectious disease management

WHO chief praises containment efforts