[Skip to Content]
[Skip to Content Landing]
Figure 1.  The Epidemic Curve, Key Events and Features, and Public Health Interventions Across the 5 Periods During the COVID-19 Outbreak in Wuhan, China
The Epidemic Curve, Key Events and Features, and Public Health Interventions Across the 5 Periods During the COVID-19 Outbreak in Wuhan, China

The epidemic curve is shown as the number of incident cases each day by the symptom onset date. Details of the key events, features of the situation, and public health interventions across the 5 periods are further described in the eMethods in the Supplement. Chunyun is a period of significant travel in China with extremely high traffic load around the Chinese Lunar New Year. Cordons sanitaire restrict movement of people outside of a defined area.

Figure 2.  The Geographic Distribution of Daily Rates of COVID-19 Cases Across the 5 Periods in Wuhan, China
The Geographic Distribution of Daily Rates of COVID-19 Cases Across the 5 Periods in Wuhan, China

The daily rate of cases is expressed as number of laboratory-confirmed cases per day per million people, grouped by each of the 13 districts of the city of Wuhan. COVID-19 indicates coronavirus disease 2019.

Figure 3.  Daily Rates of Cases in Different Groups and Proportion of Severity Categories Across the 5 Periods in Wuhan, China
Daily Rates of Cases in Different Groups and Proportion of Severity Categories Across the 5 Periods in Wuhan, China

The exact values for the daily rates of cases in panels A-C in different groups across the 5 periods are shown in eTable 2 in the Supplement. The clinical severity in panel D was defined according to the 7 editions of the Interim Diagnosis and Treatment of 2019 Novel Coronavirus Pneumonia,14 and details are shown in the eMethods in the Supplement. Error bars indicate 95% CIs.

Figure 4.  The Effective Reproduction Number (Rt) Estimates Based on Laboratory-Confirmed Coronavirus Disease 2019 (COVID-19) Cases in Wuhan, China
The Effective Reproduction Number (Rt) Estimates Based on Laboratory-Confirmed Coronavirus Disease 2019 (COVID-19) Cases in Wuhan, China

The effective reproduction number Rt is defined as the mean number of secondary cases generated by a typical primary case at time t in a population, calculated for the whole period over a 5-day moving average. Results are shown since January 1, 2020, given the limited number of diagnosed cases and limited diagnosis capacity in December 2019. The darkened horizontal line indicates Rt = 1, below which sustained transmission is unlikely so long as antitransmission measures are sustained, indicating that the outbreak is under control. The 95% credible intervals (CrIs) are presented as gray shading. Daily estimates of Rt with 95% CrIs are shown in eTable 3 in the Supplement.

Table.  Characteristics of Patients With Laboratory-Confirmed COVID-19 Across the 5 Periods in Wuhan, Chinaa
Characteristics of Patients With Laboratory-Confirmed COVID-19 Across the 5 Periods in Wuhan, Chinaa
1.
World Health Organization. Coronavirus disease 2019 (COVID-19) situation report-76. Accessed April 6, 2020. https://www.who.int/docs/default-source/coronaviruse/situation-reports/20200405-sitrep-76-covid-19.pdf?sfvrsn=6ecf0977_2
2.
Guan  WJ, Ni  ZY, Hu  Y,  et al; China Medical Treatment Expert Group for Covid-19.  Clinical characteristics of coronavirus disease 2019 in China.   N Engl J Med. Published online February 28, 2020. doi:10.1056/NEJMoa2002032 PubMedGoogle Scholar
3.
Wang  D, Hu  B, Hu  C,  et al.  Clinical characteristics of 138 hospitalized patients with 2019 novel coronavirus-infected pneumonia in Wuhan, China.   JAMA. 2020;323(11):1061-1069. doi:10.1001/jama.2020.1585 PubMedGoogle ScholarCrossref
4.
Chen  N, Zhou  M, Dong  X,  et al.  Epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in Wuhan, China: a descriptive study.   Lancet. 2020;395(10223):507-513. doi:10.1016/S0140-6736(20)30211-7 PubMedGoogle ScholarCrossref
5.
Huang  C, Wang  Y, Li  X,  et al.  Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China.   Lancet. 2020;395(10223):497-506. doi:10.1016/S0140-6736(20)30183-5 PubMedGoogle ScholarCrossref
6.
Chang  D, Lin  M, Wei  L,  et al.  Epidemiologic and clinical characteristics of novel coronavirus infections involving 13 patients outside Wuhan, China.   JAMA. 2020;323(11):1092-1093. doi:10.1001/jama.2020.1623 PubMedGoogle ScholarCrossref
7.
Xu  XW, Wu  XX, Jiang  XG,  et al.  Clinical findings in a group of patients infected with the 2019 novel coronavirus (SARS-Cov-2) outside of Wuhan, China: retrospective case series.   BMJ. 2020;368:m606. doi:10.1136/bmj.m606 PubMedGoogle ScholarCrossref
8.
Li  Q, Guan  X, Wu  P,  et al.  Early transmission dynamics in Wuhan, China, of novel coronavirus-infected pneumonia.   N Engl J Med. 2020;382(13):1199-1207. doi:10.1056/NEJMoa2001316 PubMedGoogle ScholarCrossref
9.
Wu  JT, Leung  K, Leung  GM.  Nowcasting and forecasting the potential domestic and international spread of the 2019-nCoV outbreak originating in Wuhan, China: a modelling study.   Lancet. 2020;395(10225):689-697. doi:10.1016/S0140-6736(20)30260-9 PubMedGoogle ScholarCrossref
10.
Kucharski  AJ, Russell  TW, Diamond  C,  et al; Centre for Mathematical Modelling of Infectious Diseases COVID-19 working group.  Early dynamics of transmission and control of COVID-19: a mathematical modelling study.   Lancet Infect Dis. Published online March 11, 2020. doi:10.1016/S1473-3099(20)30144-4 PubMedGoogle Scholar
11.
Li  R, Pei  S, Chen  B,  et al.  Substantial undocumented infection facilitates the rapid dissemination of novel coronavirus (SARS-CoV2).   Science. Published online March 16, 2020. doi:10.1126/science.abb3221 PubMedGoogle Scholar
12.
Chinazzi  M, Davis  JT, Ajelli  M,  et al.  The effect of travel restrictions on the spread of the 2019 novel coronavirus (COVID-19) outbreak.   Science. 2020;eaba9757. Published online March 6, 2020. doi:10.1126/science.aba9757 PubMedGoogle Scholar
13.
Tian  H, Liu  Y, Li  Y,  et al.  An investigation of transmission control measures during the first 50 days of the COVID-19 epidemic in China.   Science. Published online March 31, 2020. doi:10.1126/science.abb6105 PubMedGoogle Scholar
14.
National Health Commission of the People’s Republic of China. Interim diagnosis and treatment of 2019 novel coronavirus pneumonia. 7th ed. March 3, 2020. Accessed March 4, 2020. http://www.nhc.gov.cn/yzygj/s7653p/202003/46c9294a7dfe4cef80dc7f5912eb1989.shtml
15.
Phelan  AL, Katz  R, Gostin  LO.  The novel coronavirus originating in Wuhan, China: challenges for global health governance.   JAMA. 2020;323(8):709-710. doi:10.1001/jama.2020.1097 PubMedGoogle ScholarCrossref
16.
Spiegelman  D, Hertzmark  E.  Easy SAS calculations for risk or prevalence ratios and differences.   Am J Epidemiol. 2005;162(3):199-200. doi:10.1093/aje/kwi188 PubMedGoogle ScholarCrossref
17.
Cori  A, Ferguson  NM, Fraser  C, Cauchemez  S.  A new framework and software to estimate time-varying reproduction numbers during epidemics.   Am J Epidemiol. 2013;178(9):1505-1512. doi:10.1093/aje/kwt133 PubMedGoogle ScholarCrossref
18.
The Novel Coronavirus Pneumonia Emergency Response Epidemiology Team.  Vital surveillances: the epidemiological characteristics of an outbreak of 2019 novel coronavirus diseases (COVID-19)—China, 2020.   China CDC Weekly. 2020;2:113-122.Google Scholar
19.
Wu  Z, McGoogan  JM.  Characteristics of and important lessons from the coronavirus disease 2019 (COVID-19) outbreak in China: summary of a report of 72 314 cases from the Chinese Center for Disease Control and Prevention.   JAMA. 2020;323(13):1239-1242. doi:10.1001/jama.2020.2648 PubMedGoogle ScholarCrossref
20.
Yang  X, Yu  Y, Xu  J,  et al.  Clinical course and outcomes of critically ill patients with SARS-CoV-2 pneumonia in Wuhan, China: a single-centered, retrospective, observational study.   Lancet Respir Med. S2213-2600(20)30079-5. Published online February 24, 2020. doi:10.1016/S2213-2600(20)30079-5 PubMedGoogle Scholar
21.
Vardavas  CI, Nikitara  K.  COVID-19 and smoking: a systematic review of the evidence.   Tob Induc Dis. 2020;18:20. doi:10.18332/tid/119324 PubMedGoogle ScholarCrossref
22.
Report of the WHO-China Joint Mission on Coronavirus Disease 2019 (COVID-19). Accessed March 2, 2020. https://www.who.int/docs/default-source/coronaviruse/who-china-joint-mission-on-covid-19-final-report.pdf
23.
Wang  G, Zhang  Y, Zhao  J, Zhang  J, Jiang  F.  Mitigate the effects of home confinement on children during the COVID-19 outbreak.   Lancet. 2020;395(10228):945-947. doi:10.1016/S0140-6736(20)30547-X PubMedGoogle ScholarCrossref
24.
Adams  JG, Walls  RM.  Supporting the health care workforce during the COVID-19 global epidemic.   JAMA. Published online March 12, 2020. doi:10.1001/jama.2020.3972 PubMedGoogle Scholar
25.
Liu  Y, Gayle  AA, Wilder-Smith  A, Rocklöv  J.  The reproductive number of COVID-19 is higher compared to SARS coronavirus.   J Travel Med. 2020;27(2):taaa021. doi:10.1093/jtm/taaa021 PubMedGoogle Scholar
26.
Chen  TM, Rui  J, Wang  QP, Zhao  ZY, Cui  JA, Yin  L.  A mathematical model for simulating the phase-based transmissibility of a novel coronavirus.   Infect Dis Poverty. 2020;9(1):24. doi:10.1186/s40249-020-00640-3 PubMedGoogle ScholarCrossref
27.
Fong  MW, Gao  H, Wong  JY,  et al.  Nonpharmaceutical measures for pandemic influenza in nonhealthcare settings-social distancing measures.   Emerg Infect Dis. 2020;26(5). Published online February 6, 2020. doi:10.3201/eid2605.190995PubMedGoogle Scholar
28.
Bai  Y, Yao  L, Wei  T,  et al.  Presumed asymptomatic carrier transmission of COVID-19.   JAMA. Published online February 21, 2020. doi:10.1001/jama.2020.2565PubMedGoogle Scholar
29.
Kam  KQ, Yung  CF, Cui  L,  et al.  A well infant with coronavirus disease 2019 (COVID-19) with high viral load.   Clin Infect Dis. 2020;ciaa201. Published online February 28, 2020. doi:10.1093/cid/ciaa201 PubMedGoogle Scholar
30.
Tong  ZD, Tang  A, Li  KF,  et al.  Potential presymptomatic transmission of SARS-CoV-2, Zhejiang Province, China, 2020.   Emerg Infect Dis. 2020;26(5). Published online February 6, 2020. doi:10.3201/eid2605.200198 PubMedGoogle Scholar
31.
Lipsitch  M, Swerdlow  DL, Finelli  L.  Defining the epidemiology of Covid-19: studies needed.   N Engl J Med. 2020;382(13):1194-1196. doi:10.1056/NEJMp2002125 PubMedGoogle ScholarCrossref
32.
Morens  DM, Daszak  P, Taubenberger  JK.  Escaping pandora’s box—another novel coronavirus.   N Engl J Med. 2020;382(14):1293-1295. doi:10.1056/NEJMp2002106 PubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    3 Comments for this article
    EXPAND ALL
    Use of More Appropriate R(t) Method Changes Conclusions About When R(t) Dips Below 1
    Marc Lipsitch, DPhil | Center for Communicable Disease Dynamics, Harvard Chan School of Public Health
    Pan et al. find that the daily reproduction number R(t) remained >1 throughout the first 3 periods of control (through Feb. 2) and fell below 1 only during the 4th period (on Feb. 6), during which involuntary centralized quarantine and isolation were imposed. They argue that control measures short of involuntary quarantine were inadequate. This finding is questionable on its face (confirmed cases trended smoothly downward from onset date January 26 onward apart from an unexplained spike on Feb. 1 and is questioned by the accompanying editorial: “it is difficult to assert that additional interventions in periods 4 and 5 were necessary in driving Rt below 1.0.” Nonetheless, leading voices have cited the findings as strong support for a policy even in the United States of involuntary family separation to enforce such quarantine and isolation (1). 

    The claim that infection rates continued to grow in period three (R(t)>1) and shrink only in period four (R(t)<1) depends strongly on Pan et al.’s use of an estimator for R(t) that was not designed for this purpose. The approach of Cori et al. (2) is by design backward-looking: it uses only data from up to the date on which the estimate is centered to calculate a reproduction number. Thus the estimates of R(t) during the third period are completely determined by the past trend, and if (as an extreme example), case numbers had abruptly dropped to zero on February 2, indicating that control measures prior to February 2 were completely effective, the estimates of R(t) up to February 2 would have been unaffected. An alternative approach, sometimes referred to as the “case reproduction number,” estimates reproduction numbers for each day from past, present, and future cases (3).

    Raw counts were not available when requested from a corresponding author. We therefore digitized the figure using WebPlotDigitizer version 4.2 (https://automeris.io/WebPlotDigitizer), replicated the paper’s analysis using the method of Cori et al. (2), and compared the results against those from the alternative method of 5. Using the Cori method as Pan et al. did, our curve closely follows Fig. 4 of 1 when using the same method. As anticipated, the alternative approach, taking into account the evident decline of the epidemic curve that began during Period 3, finds that Rt dropped below 1 during Period 3, specifically on 28-29 Jan. (Fig. 1, available https://github.com/keyajoshi/Pan_response)

    There is no ideal method for estimating R(t) during a period when it is changing on a time scale equal to or faster than the serial interval of the infection. In this particular case, a few days’ difference in the timing of when Rt dropped below 1 leads to divergent policy interpretations. We believe that a purely backward-looking method like that used by the authors inevitably finds too late a date in such a situation because it cannot account for drops in future case numbers, while the alternative method of Wallinga & Teunis may be influenced to some degree by future changes in transmission. We are currently studying the performance of these different approaches to detect timing of R(t)<1. Meanwhile, caution in interpretation is warranted. In particular it is not clear that the extreme and invasive measures begun after February 2 were necessary to control transmission in Wuhan, or how this generalizes.

    with: Keya Joshi (HSPH), Sarah E. Cobey (U Chicago)

    REFERENCES

    1. https://www.nytimes.com/2020/04/07/opinion/coronavirus-smart-quarantine.html
    2. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3816335/
    3. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7110200/
    CONFLICT OF INTEREST: None Reported
    READ MORE
    Multifaceted public health interventions not only flattened but also crushed the epidemic curve in Wuhan
    An Pan, PhD | Huazhong University of Science and Technology
    We appreciate Dr. Marc Lipsitch’s comment on our paper (1). We respectfully disagree with his interpretation of our results and implications.

    As Dr. Lipsitch mentioned, there are different methods to estimate the reproduction number R, each with its own strengths and limitations, and there is “no ideal method for estimating Rt during a period when it is changing on a time scale equal to or faster than the serial interval of the infection.” The Wallinga and Teunis method (2) has its forward-looking advantages and also some drawbacks; as Cori et al (3) pointed out, “estimates are right-censored, because the
    estimate of R at time t requires incidence data from times later than t.” Therefore, this method relies on counting the incident cases in the future, which could result in a biased estimate. It might not provide real time estimates.

    To investigate the robustness of different estimates, we also calculated the Rt based on a recently published method (4). We found that the Rt using this method declined to below 1.0 on Feb 5 (https://github.com/panan-hust/Rt), similar to our findings.

    We fully agree with Dr. Lipsitch’s comment that “caution in interpretation is warranted”. In our paper, we did not state that the centralized quarantine and treatment was the policy that drove the Rt to below 1.0; instead, we emphasized “multiple interventions at the same time or in a short timeframe to control the outbreak, and thus individual strategies could not be evaluated. In addition, the observational study design precludes causal inference (1)." We provided the changes in Rt over time but did not link the decrease directly to one, specific intervention alone.

    It would be useful to compare Wuhan’s multifaceted measures with other outbreak areas that use different measures. However, countries and regions are not only different in intervention measures and scales, but also differ in many other ways, including laws and legislations, social environment, culture, population density, people’s lifestyle and living environment; therefore, the interventions in Wuhan may not be exactly applicable to other countries and regions, and countries need to develop tailored responses to COVID-19 based on their own situation. In addition, how and to what degree various interventions were implemented in practice would be extremely important in determining their effects.

    Dr. Lipsitch’s calculation showed that R quickly dropped since Jan 15 when there were almost no interventions before the city lockdown on Jan 23, and fell below 1.0 on Jan 28-29, shortly after complete traffic restriction and social distancing measures. If so, there is no basis for relating the observed reduction in R in Wuhan to policy changes. Wuhan’s experience, however, differs from that in a number of European countries and many US states (https://rt.live) where the estimated Rt is stuck around 1.0, even after several weeks of social distancing policy and stay-at-home orders.

    In contending with the COVID-19 pandemic, we believe every country would be advantaged by learning from the experience of others. What distinguished Wuhan was the adoption of these multifaceted public health interventions, and while the effects of any one component cannot be fully disentangled from the others, the cumulative result was striking.

    With Prof Tangchun Wu and Prof Sheng We (HUST) and Prof Xihong Lin (Harvard).

    REFERENCES

    1. JAMA. doi:10.1001/jama.2020.6130.
    2. Am J Epidemiol. 2004;160(6):509-16.
    3. Am J Epidemiol. 2013;178(9):1505-12.
    4. Lancet Infect Dis 2020;20: 30230-9.
    CONFLICT OF INTEREST: None Reported
    READ MORE
    Using Symptom Onset Instead of Incidence Results in Time-Lagged Estimate of R(t)
    Qingyuan Zhao, PhD | Statistical Laboratory, University of Cambridge
    The instantaneous reproduction number R(t) is defined as the average number of secondary cases that would be generated by a primary case infected on day t if conditions remained the same after that day [1]. R(t) can be estimated by the number of new infections on day t divided by the effective infectivity on day t of the individuals already infected [1, 2].

    Although R(t) is defined in terms of and ideally should be estimated using new infections, times of infection are rarely observed in practice. Instead, Cori et al. [2] suggested that their method can also be applied
    with the infection incidence curve replaced by the more commonly observed symptom onset curve. But this comes with an important caveat: the estimated R(t) would have a time lag equal to the incubation period. They suggested that a possible strategy to correct the time lag is to use the incubation period distribution to back-calculate the curve of incidences from the observed curve of symptom onsets.

    We followed this suggestion and reanalyzed the (digitized) epidemic curve in Figure 1 of Pan et al. To capture the uncertainty of the imputed incidences, we used 1000 Monte Carlo samples of the incidence curve instead of the point estimator suggested by Fraser [1]. To simulate the incidence, we used a Gamma-distributed incubation period with a median of 4.5 days and 95% quantile of 13.4 days, estimated in a previous study of ours [4]. For each simulated incidence curve, we obtained one posterior sample of R(t) using the method of Cori et al. [2]. We then pooled the posterior samples across the 1000 simulations to assess the uncertainty of the estimated R(t). To ease comparison, we replicated the analyses in Pan et al. and the comment by Lipsitch et al. which used the method of Wallinga and Teunis [3]. Code for the analysis and results can be found in https://github.com/phyllisju/rt.

    Not surprisingly, back-calculating the incidence leads to an R(t) curve quite different from the one in Pan et al. The new curve is visually smoother before Period 2 and close to the curve obtained by Lipsitch et al. The new R(t) curve starts to decrease around January 15 and dips below 1 on January 31, before the implementation of some more aggressive public health measures on February 2. The earlier comment by Lipsitch et al. and response by Pan et al. attributed their difference in the estimated curves of R(t) to the “forward-looking” and “backward-looking” nature of the statistical methods they used. Our reanalysis shows that another contributing factor is the time lag of the estimated R(t), occurred when the method of Cori et al. [2] is applied with symptom onsets instead of (back-calculated) incidences.

    We would like to stress three further points. First, instead of the instantaneous reproduction number R(t), the method of Wallinga and Teunis [3] in fact targets the case reproduction number which can be viewed as a weighted average of R(t) after t [1]. Second, the method of Wallinga and Teunis [3] is also commonly applied with symptom onsets instead of incidences, resulting in a similar time lag with the actual case reproduction number. Finally, Cori et al. [2] did not adopt a fully Bayesian approach and instead used a sliding window method to smoothen R(t). Caution is thus needed to interpret its credible intervals, especially in time periods where R(t) changes sharply.

    With Nianqiao Ju (Harvard University) and Sergio Bacallado (University of Cambridge).

    [1] DOI: 10.1371/journal.pone.0000758
    [2] DOI: 10.1093/aje/kwt133
    [3] DOI: 10.1093/aje/kwh255
    [4] arXiv: 2004.07743
    CONFLICT OF INTEREST: None Reported
    READ MORE
    Views 130,086
    Citations 0
    Original Investigation
    April 10, 2020

    Association of Public Health Interventions With the Epidemiology of the COVID-19 Outbreak in Wuhan, China

    Author Affiliations
    • 1Department of Epidemiology and Biostatistics, School of Public Health, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China
    • 2Department of Occupational and Environmental Health, Ministry of Education Key Laboratory of Environment and Health, and State Key Laboratory of Environmental Health (Incubating), School of Public Health, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China
    • 3School of Public Health, Ministry of Education Key Laboratory of Public Health Safety, Fudan University, Shanghai, China
    • 4Department of Biostatistics, Harvard T.H. Chan School of Public Health, Boston, Massachusetts
    • 5Department of Statistics, Harvard University, Boston, Massachusetts
    JAMA. 2020;323(19):1915-1923. doi:10.1001/jama.2020.6130
    Key Points

    Question  Was there an association of public health interventions with improved control of the COVID-19 outbreak in Wuhan, China?

    Findings  In this cohort study that included 32 583 patients with laboratory-confirmed COVID-19 in Wuhan from December 8, 2019, through March 8, 2020, the institution of interventions including cordons sanitaire, traffic restriction, social distancing, home quarantine, centralized quarantine, and universal symptom survey was temporally associated with reduced effective reproduction number of SARS-CoV-2 (secondary transmission) and the number of confirmed cases per day across age groups, sex, and geographic regions.

    Meaning  A series of multifaceted public health interventions was temporally associated with improved control of the COVID-19 outbreak in Wuhan and may inform public health policy in other countries and regions.

    Abstract

    Importance  Coronavirus disease 2019 (COVID-19) has become a pandemic, and it is unknown whether a combination of public health interventions can improve control of the outbreak.

    Objective  To evaluate the association of public health interventions with the epidemiological features of the COVID-19 outbreak in Wuhan by 5 periods according to key events and interventions.

    Design, Setting, and Participants  In this cohort study, individual-level data on 32 583 laboratory-confirmed COVID-19 cases reported between December 8, 2019, and March 8, 2020, were extracted from the municipal Notifiable Disease Report System, including patients’ age, sex, residential location, occupation, and severity classification.

    Exposures  Nonpharmaceutical public health interventions including cordons sanitaire, traffic restriction, social distancing, home confinement, centralized quarantine, and universal symptom survey.

    Main Outcomes and Measures  Rates of laboratory-confirmed COVID-19 infections (defined as the number of cases per day per million people), across age, sex, and geographic locations were calculated across 5 periods: December 8 to January 9 (no intervention), January 10 to 22 (massive human movement due to the Chinese New Year holiday), January 23 to February 1 (cordons sanitaire, traffic restriction and home quarantine), February 2 to 16 (centralized quarantine and treatment), and February 17 to March 8 (universal symptom survey). The effective reproduction number of SARS-CoV-2 (an indicator of secondary transmission) was also calculated over the periods.

    Results  Among 32 583 laboratory-confirmed COVID-19 cases, the median patient age was 56.7 years (range, 0-103; interquartile range, 43.4-66.8) and 16 817 (51.6%) were women. The daily confirmed case rate peaked in the third period and declined afterward across geographic regions and sex and age groups, except for children and adolescents, whose rate of confirmed cases continued to increase. The daily confirmed case rate over the whole period in local health care workers (130.5 per million people [95% CI, 123.9-137.2]) was higher than that in the general population (41.5 per million people [95% CI, 41.0-41.9]). The proportion of severe and critical cases decreased from 53.1% to 10.3% over the 5 periods. The severity risk increased with age: compared with those aged 20 to 39 years (proportion of severe and critical cases, 12.1%), elderly people (≥80 years) had a higher risk of having severe or critical disease (proportion, 41.3%; risk ratio, 3.61 [95% CI, 3.31-3.95]) while younger people (<20 years) had a lower risk (proportion, 4.1%; risk ratio, 0.47 [95% CI, 0.31-0.70]). The effective reproduction number fluctuated above 3.0 before January 26, decreased to below 1.0 after February 6, and decreased further to less than 0.3 after March 1.

    Conclusions and Relevance  A series of multifaceted public health interventions was temporally associated with improved control of the COVID-19 outbreak in Wuhan, China. These findings may inform public health policy in other countries and regions.

    Introduction

    Coronavirus disease 2019 (COVID-19) is an emerging respiratory infectious disease caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), which was first detected in early December 2019 in Wuhan, China. As of April 6, 2020, COVID-19 had quickly spread to the majority of countries worldwide, affected more than 1.1 million individuals, and caused nearly 63 000 deaths.1 Although studies have described the clinical characteristics of patients with COVID-19,2-7 and a previous study has reported the early transmission dynamics of the first 425 confirmed cases in Wuhan,8 more recent data are required to illustrate the full spectrum of the epidemiological characteristics of the outbreak in the epicenter.

    Several modeling studies have used the international cases exported from Wuhan to extrapolate the severity of the epidemic,9-13 estimating that the Wuhan travel ban delayed the epidemic progression by 3 to 5 days in mainland China,12,13 while reducing case importations to other countries by nearly 80% through mid-February.12 However, to our knowledge, no study has yet comprehensively evaluated the association of various public health interventions implemented by the Chinese government (including but not limited to intensive intracity and intercity traffic restriction, social distancing measures, home isolation and centralized quarantine, and improvement of medical resources; Figure 1) with outbreak control within Wuhan city.

    In this study, the epidemiological characteristics of patients with COVID-19 in Wuhan through March 8, 2020, were described, and the rate of confirmed cases and effective reproduction number in different periods according to key events and interventions were compared to evaluate the temporal associations of multiple public health interventions with control of the COVID-19 outbreak in Wuhan.

    Methods
    Source of Data

    Characteristics of patients with COVID-19 from December 2019 through March 8, 2020, were extracted on March 9 from the municipal Notifiable Disease Report System, including birth date, sex, occupation, residential district, date of symptom onset (the self-reported date of symptoms such as fever, cough, or other respiratory symptoms), and date of confirmed diagnosis (the laboratory confirmation date of SARS-CoV-2 infection in the biosamples). A case was recorded as a health care worker if the patient reported working in a hospital or clinic. Waiver of informed consent for collection of epidemiological data from patients with COVID-19 was granted by the National Health Commission of China as part of the infectious disease outbreak investigation. All identifiable personal information was removed for privacy protection.

    Case Definitions

    Cases were diagnosed and the severity status was categorized as mild, moderate, severe, or critical according to the Diagnosis and Treatment Scheme for COVID-19 released by the National Health Commission of China (details in the eMethods in the Supplement).14 A laboratory-confirmed case was defined if the patient had a positive result for SARS-CoV-2 virus by real-time reverse transcriptase–polymerase chain reaction (RT-PCR) assay or high-throughput sequencing of nasal and pharyngeal swab specimens. Only laboratory-confirmed cases were included in primary analyses for the consistency of case definition throughout the periods. There were an additional 17 365 cases of clinically diagnosed COVID-19 (ie, by symptom report and chest x-ray but without a positive RT-PCR result) in the data set. However, only clinically diagnosed cases were permitted for case tracking from February 9 through February 19, when testing became more broadly available, and only permitted for presumptive cases in Hubei Province. Furthermore, many of the cases were actually not new, but previous cases for which retrospective diagnoses were made based on medical records. Therefore, we only included laboratory-confirmed cases in our analyses for the consistency of case definition throughout the periods, as well as to be comparable to data in other areas outside Hubei Province. Consequently, the number of cases in the primary analysis was smaller than the officially reported number, although clinically diagnosed cases were included in a sensitivity analysis (outlined below).

    Classification of 5 Time Periods

    To better reflect the dynamics of the COVID-19 epidemic and corresponding interventions, 5 periods were classified based on important dates that could affect the virus transmission in Wuhan (Figure 1). The time before January 10, 2020, the first date of Chunyun (massive migration for the Chinese Lunar New Year), was considered as the first period, when no COVID-19–specific interventions were imposed. The second period was the Chunyun of January 10 to 22, 2020, when massive population movement occurred and was expected to accelerate the spread of COVID-19.15 No strong intervention was imposed during the Chunyun period; the first announcement of human-to-human transmission and infections in health care workers was made on January 20. During this period, hospitals started to be overcrowded with patients with fever or respiratory symptoms.

    During the third period, between January 23 and February 1, the local government first blocked all outbound transportation from the city and subsequently suspended public transit and banned all vehicular traffic within the city. Other social distancing measures were also implemented, including compulsory mask wearing in public places and cancellation of social gatherings. Due to a severe shortage of medical resources in this period, many confirmed or presumptive cases could not receive timely diagnosis and treatment and were self-quarantined at home. On February 2, with improvement in medical resources, the government implemented a policy of centralized quarantine and treatment of all confirmed and presumptive cases, those with fever or respiratory symptoms, and close contacts of confirmed cases in designated hospitals or facilities. Meanwhile, a stay-at-home policy was implemented for all residents in the city. On February 17, the government initiated door-to-door and individual-to-individual symptom screening for all residents with support from thousands of community workers.

    Taken together, the outbreak was divided into 5 periods: no intervention before January 10, massive human migration between January 10 and 22, city lockdown with traffic suspension and home quarantine between January 23 and February 1, intensified measures with centralized quarantine and treatment between February 2 and 16, and community universal symptom survey on and after February 17, 2020. Specific events and features are shown in Figure 1.

    Outcomes

    Daily rate of confirmed cases, defined as the number of laboratory-confirmed cases per day per million people, were estimated by patient age, sex, health care occupation, and residential district across periods. The calculation used the number of cases in each period divided by the number of days in each period (33, 13, 10, 15, and 21 days) and the subtotal population size in each stratum from the Wuhan Statistical Yearbook 2018. The effective reproduction number Rt, defined as the mean number of secondary cases generated by a typical primary case at time t in a population, was calculated as an indicator to measure the transmission of SARS-CoV-2 both before and after the interventions. Clinical severity information was extracted from the data set and was available for 32 325 confirmed cases (258 cases had missing values and were not included in the clinical severity analyses).

    Statistical Analysis

    The distribution of age, sex, and occupation in laboratory-confirmed cases was described, and the epidemic curve by the symptom onset date and important dates of interventions was plotted. A sensitivity analysis was also conducted to include the clinically diagnosed cases in the epidemic curve. The geographical distributions of daily rates of COVID-19 cases across Wuhan city through the 5 periods were presented using ArcGIS software version 10.6 (Environmental Systems Research Institute Inc). A modified Poisson regression with robust variance was used to evaluate the relationship between age, sex, time period, and health care occupation with disease severity (mild and moderate vs severe and critical).16 The 4 variables were included simultaneously and risk ratios (RRs) were reported along with 95% CIs. All statistical analyses were performed using SAS statistical software version 9.3 (SAS Institute Inc), and P values were 2-tailed with statistical significance set at .05.

    The Rt was calculated using the method developed by Cori et al17 in R version 3.6.2 (R Foundation for Statistical Computing). The daily number of reported COVID-19 cases and the serial interval (mean, 7.5 days [SD, 3.4 days]; constant across periods), derived from a previous epidemiological survey of the first 425 cases in Wuhan,8 were used to estimate Rt and its 95% credible interval on each day via a 5-day moving average. The Rt was calculated for the whole period, but results were shown beginning with January 1, 2020, given the limited number of diagnosed cases and limited diagnosis capacity in December 2019.

    Results
    Characteristics of Patients With COVID-19

    The analyses included a total of 32 583 confirmed cases, among whom 15 766 (48.4%) were men and 16 817 (51.6%) were women (Table). The median age of the patients was 56.7 years (range, 0-103; interquartile range, 43.4-66.8), with the majority (n = 24 203, 74.3%) aged 40 to 79 years (Table).

    The epidemic curve according to the symptom onset date and key interventions is shown in Figure 1. Most cases occurred between January 20 and February 6, with a spike on February 1. The epidemic curve with inclusion of clinically diagnosed cases is shown in eFigure 1 in the Supplement, and the epidemic curve limiting the sample to only severe and critical cases is shown in eFigure 2 in the Supplement; both showed similar patterns to the main analysis. There was a substantial delay between symptom onset date and laboratory confirmation date in the early periods, with the lag decreasing over time (median, 26, 15, 10, 6, and 3 days for the 5 periods, respectively; eFigure 3 in the Supplement).

    Geographic Spread and Confirmed Case Rates

    The outbreak started from the urban districts and gradually spread to the suburban and rural areas across the 5 periods. There were strong geographic differences in rates of confirmed cases, with the highest rates in the urban districts (Figure 2).

    The daily confirmed case rate per million people increased from 2.0 (95% CI, 1.8-2.1) before January 10, to 45.9 (95% CI, 44.6-47.1) between January 10 and 22, and to 162.6 (95% CI, 159.9-165.3) between January 23 and February 1, and then decreased to 77.9 (95% CI, 76.3-79.4) between February 2 and 16, and 17.2 (95% CI, 16.6-17.8) after February 16 (Figure 3A; eTable 1 in the Supplement). Similar patterns were observed for men and women, with a slightly higher rate in women (43.7 [95% CI, 43.0-44.4]) compared with men (39.4 [95% CI, 38.8-40.0]) over the whole period (Figure 3A).

    A total of 1496 local health care workers had confirmed cases, representing 4.6% of all cases (Table). The daily rate of cases in local health care workers (130.5 per million people [95% CI, 123.9-137.2]) was higher than that in the general population (41.5 per million people [95% CI, 41.0-41.9]) over the whole period. The rate among health care workers peaked in the third period (617.4 per million people [95% CI, 576.3-658.4]), but decreased in the last 2 periods when comprehensive personal protective equipment was more widely used (Figure 3A; eTable 1 in the Supplement).

    Rates of confirmed cases and trends also differed by age. It peaked in the third period and declined thereafter for those older than 20 years, while it continued to increase for children and adolescents (age <20 years) (Figure 3B; eTable 1 in the Supplement), particularly for infants younger than 1 year (Figure 3C). The rate over the whole period among infants younger than 1 year was 7.9 per million people (95% CI, 5.8-10.0), while it ranged from 2.0 to 5.4 among other age groups of children and adolescents (Figure 3C; eTable 1 in the Supplement).

    Clinical Severity of Disease

    Confirmed cases with available data (n = 32 325) were classified into mild (n = 15 531, 48.0%), moderate (n = 9655, 29.9%), severe (n = 6169, 19.1%), and critical (n = 970, 3.0%) (Figure 3D). The proportion of severe and critical cases decreased gradually over time, accounting for 53.1%, 35.1%, 23.5%, 15.9%, and 10.3% of the classifiable cases in the 5 periods, respectively (Figure 3D).

    The unadjusted proportion of severe and critical cases increased with age (4.10% in those aged <20 years, 12.1% in those aged 20-39 years, 17.4% in those aged 40-59 years, 29.6% in those aged 60-79 years, and 41.3% in those aged ≥80 years), and the corresponding multivariable-adjusted RRs (95% CIs) were 0.47 (0.31-0.70), 1.00 (reference), 1.41 (1.30-1.53), 2.33 (2.16-2.52), and 3.61 (3.31-3.95), respectively (eTable 2 in the Supplement).

    Females were at lower risk of severe and critical disease than were males (unadjusted proportion, 20.6% vs 23.7%; adjusted RR, 0.90 [95% CI, 0.86-0.93]), while there were no significant differences in clinical severity between health care workers and other occupation groups (unadjusted proportion, 17.4% vs 22.3%; adjusted RR, 1.08 [95% CI, 0.96-1.21]).

    Estimates of Rt

    Estimates of the effective reproduction number Rt varied in the first period (Figure 4), gradually increased in the second period with a peak of 3.82 on January 24, and declined thereafter. The Rt fell below 1.0 on February 6, 2020, and further decreased to below 0.3 on March 1, 2020. The daily estimated values of Rt are shown in eTable 3 in the Supplement.

    Discussion

    In this cohort study, the number of incident COVID-19 cases, rates of confirmed cases, and Rt were reduced and the outbreak was under improved control in Wuhan after implementation of multifaceted public health measures (including but not limited to intensive intracity and intercity traffic restriction, social distancing measures, home confinement and centralized quarantine, and improvement of medical resources).

    Among the 32 583 confirmed COVID-19 cases, females had a higher rate of confirmed cases compared with males, but males were more likely to have severe or critical illness. This is consistent with previous reports from China suggesting a higher crude fatality rate among men compared with women (2.8% vs 1.7%),18,19 and another study in critically ill patients demonstrating that more men were affected (67%) than women (33%).20 Although the reasons for these differences are unknown, it is possible that men were more likely to be current smokers21 and had a higher proportion of comorbid conditions, which might worsen their prognosis. Also consistent with early reports, younger people were less likely to be affected,8,18,22 although the rate continued to increase in children and adolescents over time. Infants younger than 1 year had the highest rate of cases among children and adolescents. These results suggest that vigorous efforts should be made to protect and reduce transmission and symptom progression in vulnerable populations including both elderly people and young children.22,23

    The rate of cases in health care workers was substantially higher than in the general population between January 11 and February 1, indicating a high risk of nosocomial infection. This might be due to lower awareness and inadequate use of personal protective equipment in the early periods in this study, and later a severe shortage of medical resources. The rate of confirmed cases among local health care workers quickly decreased in the later periods, after increasing awareness of and wider use of personal protective equipment, proper training, adequate hospital-level prevention and management, and support from more than 30 000 health care workers from other provinces of China. Furthermore, none of the health care workers from other provinces were infected, supporting the importance of carefully protecting health care workers in the outbreak of a high transmissible infectious disease.24

    The findings of this study may be valuable in the current efforts to combat the global pandemic of COVID-19. Many countries, regions, and communities have or will confront a surge in incident cases,1 similar to what happened in Wuhan in January. In Wuhan, vigorous and multifaceted measures of containment, mitigation, and suppression were temporally associated with improved control of the epidemic when there was neither effective drug nor vaccine. In a city with 10 million residents, mitigation measures, such as traffic restriction, cancellation of social gatherings, and home quarantine, were associated with a reduction in the degree of transmission. However, despite these interventions, the confirmed case rate continued to increase in the third period, perhaps in part due to shortages of pharmaceuticals and medical equipment and delayed diagnosis and access to medical treatment. Without rapid diagnosis, the risk of cross-infection in hospitals was high and patients likely continued to infect family members and close contacts. It has been reported that about 80% of the cluster transmission occurred in families in China.22

    With the improvement of medical resources (designated hospitals and wards, use of personal protective equipment, increased testing capacity and accelerated reporting, and timely medical treatment) and increasing supply of health care workers, the Chinese government issued a centralized quarantine and treatment policy on February 2. All confirmed and presumptive cases, those with fever or respiratory symptoms, as well as close contacts of confirmed cases were allocated to designated hospitals or facilities. Centralized quarantine of patients and close contacts appears to have been associated with a reduction in in-hospital, household, and community transmission. Between February 16 and 18, the government further initiated a door-to-door and individual-to-individual universal symptom survey to single out presumptive cases in the community, which was associated with further reductions in the spread of COVID-19 in Wuhan.

    Previous studies have reported varied R0 (range 1.40-6.49 with a mean of 3.28) for the COVID-19 outbreak due to different data sources, time periods, and statistical methods.25 Using the first 425 patients in Wuhan, an early study reported an R0 of 2.20 based on the growth rate of the epidemic curve and the serial interval,8 while a recent analysis based on a transmission network model reported an R0 of 3.58,26 similar to the current estimates in the first and second period. This study found that the implementation of public health interventions was associated with a reduction in Rt to below 1.0 on February 6 and to below 0.3 on March 1, which may have implications for global efforts to contain the pandemic more broadly.27

    Limitations

    The study has several limitations. First, the Chinese government implemented multiple interventions at the same time or in a short timeframe to control the outbreak, and thus individual strategies could not be evaluated. In addition, the observational study design precludes causal inference. However, clinical trials were not feasible or ethical under such public health emergencies, and there are not yet data available to compare the experience in Wuhan with other outbreak areas pursuing different policies.

    Second, data were extracted from the infectious disease reporting system, and no information was available for other epidemiological variables and clinical characteristics, such as incubation period, time to hospitalization, time to discharge, medical treatment strategies, and vital status. Cases occurring in the early days of a new period included infections that were acquired during the previous period, and thus there were lags for the interventions to take effect.

    Third, there were unexplained peaks in the epidemic curve, particularly a surge on February 1. Fourth, no data were available on the diagnostic testing pattern, ascertainment rate, and proportion of asymptomatic cases. The shortage of testing in the early periods suggests that ascertainment bias may in part explain the initially high proportion of severe and critical cases; testing was widely available in the fourth and fifth periods and these estimates may reflect a more accurate measure of disease severity. There were likely substantial proportions of undetected cases in Wuhan in early periods,9-11 and asymptomatic cases could be infectious to others.28-30 Further field investigations and serologic studies are needed to confirm the ascertainment rate and the effect of unascertained cases (asymptomatic cases or patients with mild symptoms who could recover without seeking medical care) on the epidemic course31,32; this represents an important area for future study.

    Conclusions

    A series of multifaceted public health interventions was temporally associated with improved control of the COVID-19 outbreak in Wuhan, China. These findings may inform public health policy in other countries and regions to combat the global pandemic of COVID-19.

    Back to top
    Article Information

    Corresponding Authors: Tangchun Wu, MD, PhD (wut@tjmu.edu.cn), and Sheng Wei, MD, PhD (shengwei@hust.edu.cn), School of Public Health, Tongji Medical College, Huazhong University of Science and Technology, 13 Hongkong Rd, Wuhan 430030, Hubei, China; and Xihong Lin, PhD, Department of Biostatistics, Harvard T.H. Chan School of Public Health, 665 Huntington Ave, Boston, MA 02115 (xlin@hsph.harvard.edu).

    Accepted for Publication: April 3, 2020.

    Published Online: April 10, 2020. doi:10.1001/jama.2020.6130

    Author Contributions: Drs Wei and Wu had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Joint first authors are Drs Pan, Liu, C. Wang, Guo, and Hao.

    Concept and design: Pan, C. Wang, Wu.

    Acquisition, analysis, or interpretation of data: All authors.

    Drafting of the manuscript: Pan, C. Wang, Lin.

    Critical revision of the manuscript for important intellectual content: Liu, Guo, Hao, Q. Wang, Huang, He, Yu, Lin, Wei, Wu.

    Statistical analysis: Pan, Liu, C. Wang, Guo, Hao, Q. Wang, Huang, Lin, Wei.

    Obtained funding: Wu.

    Administrative, technical, or material support: He, Yu, Wu.

    Supervision: C. Wang, Lin, Wu.

    Conflict of Interest Disclosures: Dr Yu reported receiving grants from the National Natural Science Foundation of China, the Program of Shanghai Academic/Technology Research Leader, and the National Science and Technology Major Project of China during the conduct of the study; grants from Sanofi Pasteur, GlaxoSmithKline, Yichang HEC Changjiang Pharmaceutical Company, and bioMérieux Diagnostic Product (Shanghai) outside the submitted work. No other disclosures were reported.

    Funding/Support: This study was partly supported by the Fundamental Research Funds for the Central Universities (2019kfyXMBZ015), the 111 Project (Drs Pan, Liu, C. Wang, Guo, Hao, Q. Wang, Huang, Wei, and Wu). Dr Lin is supported by Harvard University.

    Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

    Additional Contributions: We are grateful to all staff at the national, provincial, and municipal Center for Disease Control and Prevention for providing the data and all medical staff members and field workers who are working on the frontline of caring for patients and collecting the data.

    References
    1.
    World Health Organization. Coronavirus disease 2019 (COVID-19) situation report-76. Accessed April 6, 2020. https://www.who.int/docs/default-source/coronaviruse/situation-reports/20200405-sitrep-76-covid-19.pdf?sfvrsn=6ecf0977_2
    2.
    Guan  WJ, Ni  ZY, Hu  Y,  et al; China Medical Treatment Expert Group for Covid-19.  Clinical characteristics of coronavirus disease 2019 in China.   N Engl J Med. Published online February 28, 2020. doi:10.1056/NEJMoa2002032 PubMedGoogle Scholar
    3.
    Wang  D, Hu  B, Hu  C,  et al.  Clinical characteristics of 138 hospitalized patients with 2019 novel coronavirus-infected pneumonia in Wuhan, China.   JAMA. 2020;323(11):1061-1069. doi:10.1001/jama.2020.1585 PubMedGoogle ScholarCrossref
    4.
    Chen  N, Zhou  M, Dong  X,  et al.  Epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in Wuhan, China: a descriptive study.   Lancet. 2020;395(10223):507-513. doi:10.1016/S0140-6736(20)30211-7 PubMedGoogle ScholarCrossref
    5.
    Huang  C, Wang  Y, Li  X,  et al.  Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China.   Lancet. 2020;395(10223):497-506. doi:10.1016/S0140-6736(20)30183-5 PubMedGoogle ScholarCrossref
    6.
    Chang  D, Lin  M, Wei  L,  et al.  Epidemiologic and clinical characteristics of novel coronavirus infections involving 13 patients outside Wuhan, China.   JAMA. 2020;323(11):1092-1093. doi:10.1001/jama.2020.1623 PubMedGoogle ScholarCrossref
    7.
    Xu  XW, Wu  XX, Jiang  XG,  et al.  Clinical findings in a group of patients infected with the 2019 novel coronavirus (SARS-Cov-2) outside of Wuhan, China: retrospective case series.   BMJ. 2020;368:m606. doi:10.1136/bmj.m606 PubMedGoogle ScholarCrossref
    8.
    Li  Q, Guan  X, Wu  P,  et al.  Early transmission dynamics in Wuhan, China, of novel coronavirus-infected pneumonia.   N Engl J Med. 2020;382(13):1199-1207. doi:10.1056/NEJMoa2001316 PubMedGoogle ScholarCrossref
    9.
    Wu  JT, Leung  K, Leung  GM.  Nowcasting and forecasting the potential domestic and international spread of the 2019-nCoV outbreak originating in Wuhan, China: a modelling study.   Lancet. 2020;395(10225):689-697. doi:10.1016/S0140-6736(20)30260-9 PubMedGoogle ScholarCrossref
    10.
    Kucharski  AJ, Russell  TW, Diamond  C,  et al; Centre for Mathematical Modelling of Infectious Diseases COVID-19 working group.  Early dynamics of transmission and control of COVID-19: a mathematical modelling study.   Lancet Infect Dis. Published online March 11, 2020. doi:10.1016/S1473-3099(20)30144-4 PubMedGoogle Scholar
    11.
    Li  R, Pei  S, Chen  B,  et al.  Substantial undocumented infection facilitates the rapid dissemination of novel coronavirus (SARS-CoV2).   Science. Published online March 16, 2020. doi:10.1126/science.abb3221 PubMedGoogle Scholar
    12.
    Chinazzi  M, Davis  JT, Ajelli  M,  et al.  The effect of travel restrictions on the spread of the 2019 novel coronavirus (COVID-19) outbreak.   Science. 2020;eaba9757. Published online March 6, 2020. doi:10.1126/science.aba9757 PubMedGoogle Scholar
    13.
    Tian  H, Liu  Y, Li  Y,  et al.  An investigation of transmission control measures during the first 50 days of the COVID-19 epidemic in China.   Science. Published online March 31, 2020. doi:10.1126/science.abb6105 PubMedGoogle Scholar
    14.
    National Health Commission of the People’s Republic of China. Interim diagnosis and treatment of 2019 novel coronavirus pneumonia. 7th ed. March 3, 2020. Accessed March 4, 2020. http://www.nhc.gov.cn/yzygj/s7653p/202003/46c9294a7dfe4cef80dc7f5912eb1989.shtml
    15.
    Phelan  AL, Katz  R, Gostin  LO.  The novel coronavirus originating in Wuhan, China: challenges for global health governance.   JAMA. 2020;323(8):709-710. doi:10.1001/jama.2020.1097 PubMedGoogle ScholarCrossref
    16.
    Spiegelman  D, Hertzmark  E.  Easy SAS calculations for risk or prevalence ratios and differences.   Am J Epidemiol. 2005;162(3):199-200. doi:10.1093/aje/kwi188 PubMedGoogle ScholarCrossref
    17.
    Cori  A, Ferguson  NM, Fraser  C, Cauchemez  S.  A new framework and software to estimate time-varying reproduction numbers during epidemics.   Am J Epidemiol. 2013;178(9):1505-1512. doi:10.1093/aje/kwt133 PubMedGoogle ScholarCrossref
    18.
    The Novel Coronavirus Pneumonia Emergency Response Epidemiology Team.  Vital surveillances: the epidemiological characteristics of an outbreak of 2019 novel coronavirus diseases (COVID-19)—China, 2020.   China CDC Weekly. 2020;2:113-122.Google Scholar
    19.
    Wu  Z, McGoogan  JM.  Characteristics of and important lessons from the coronavirus disease 2019 (COVID-19) outbreak in China: summary of a report of 72 314 cases from the Chinese Center for Disease Control and Prevention.   JAMA. 2020;323(13):1239-1242. doi:10.1001/jama.2020.2648 PubMedGoogle ScholarCrossref
    20.
    Yang  X, Yu  Y, Xu  J,  et al.  Clinical course and outcomes of critically ill patients with SARS-CoV-2 pneumonia in Wuhan, China: a single-centered, retrospective, observational study.   Lancet Respir Med. S2213-2600(20)30079-5. Published online February 24, 2020. doi:10.1016/S2213-2600(20)30079-5 PubMedGoogle Scholar
    21.
    Vardavas  CI, Nikitara  K.  COVID-19 and smoking: a systematic review of the evidence.   Tob Induc Dis. 2020;18:20. doi:10.18332/tid/119324 PubMedGoogle ScholarCrossref
    22.
    Report of the WHO-China Joint Mission on Coronavirus Disease 2019 (COVID-19). Accessed March 2, 2020. https://www.who.int/docs/default-source/coronaviruse/who-china-joint-mission-on-covid-19-final-report.pdf
    23.
    Wang  G, Zhang  Y, Zhao  J, Zhang  J, Jiang  F.  Mitigate the effects of home confinement on children during the COVID-19 outbreak.   Lancet. 2020;395(10228):945-947. doi:10.1016/S0140-6736(20)30547-X PubMedGoogle ScholarCrossref
    24.
    Adams  JG, Walls  RM.  Supporting the health care workforce during the COVID-19 global epidemic.   JAMA. Published online March 12, 2020. doi:10.1001/jama.2020.3972 PubMedGoogle Scholar
    25.
    Liu  Y, Gayle  AA, Wilder-Smith  A, Rocklöv  J.  The reproductive number of COVID-19 is higher compared to SARS coronavirus.   J Travel Med. 2020;27(2):taaa021. doi:10.1093/jtm/taaa021 PubMedGoogle Scholar
    26.
    Chen  TM, Rui  J, Wang  QP, Zhao  ZY, Cui  JA, Yin  L.  A mathematical model for simulating the phase-based transmissibility of a novel coronavirus.   Infect Dis Poverty. 2020;9(1):24. doi:10.1186/s40249-020-00640-3 PubMedGoogle ScholarCrossref
    27.
    Fong  MW, Gao  H, Wong  JY,  et al.  Nonpharmaceutical measures for pandemic influenza in nonhealthcare settings-social distancing measures.   Emerg Infect Dis. 2020;26(5). Published online February 6, 2020. doi:10.3201/eid2605.190995PubMedGoogle Scholar
    28.
    Bai  Y, Yao  L, Wei  T,  et al.  Presumed asymptomatic carrier transmission of COVID-19.   JAMA. Published online February 21, 2020. doi:10.1001/jama.2020.2565PubMedGoogle Scholar
    29.
    Kam  KQ, Yung  CF, Cui  L,  et al.  A well infant with coronavirus disease 2019 (COVID-19) with high viral load.   Clin Infect Dis. 2020;ciaa201. Published online February 28, 2020. doi:10.1093/cid/ciaa201 PubMedGoogle Scholar
    30.
    Tong  ZD, Tang  A, Li  KF,  et al.  Potential presymptomatic transmission of SARS-CoV-2, Zhejiang Province, China, 2020.   Emerg Infect Dis. 2020;26(5). Published online February 6, 2020. doi:10.3201/eid2605.200198 PubMedGoogle Scholar
    31.
    Lipsitch  M, Swerdlow  DL, Finelli  L.  Defining the epidemiology of Covid-19: studies needed.   N Engl J Med. 2020;382(13):1194-1196. doi:10.1056/NEJMp2002125 PubMedGoogle ScholarCrossref
    32.
    Morens  DM, Daszak  P, Taubenberger  JK.  Escaping pandora’s box—another novel coronavirus.   N Engl J Med. 2020;382(14):1293-1295. doi:10.1056/NEJMp2002106 PubMedGoogle ScholarCrossref
    ×