key: cord-0981234-3xgkenje authors: Gu, Jia; Yan, Han; Huang, Yaxuan; Zhu, Yuru; Sun, Haoxuan; Qiu, Yumou; Chen, Song Xi title: Comparing containment measures among nations by Epidemiological Effects of COVID-19 date: 2020-09-19 journal: Natl Sci Rev DOI: 10.1093/nsr/nwaa243 sha: 94f86266611892abad0105fc47d52ebc5750373d doc_id: 981234 cord_uid: 3xgkenje nan Our study is focused on 25 countries which had experienced the COVID-19 epidemic earlier in the pandemic such that they have experienced at least four weeks of established community infections by April 20th, 2020. It is conducted by evaluating and comparing the effective reproduction number R t curves of the countries and associating them with the timing and the extend of the control measures taken by those countries. The study is based on an extended SEIR model (11) with time varying coefficients (vSEIdR model). Unlike the conventional SEIR model (12) , the vSEIdR model allows (i) infections both before and after diagnosis to reflect the clinical reality that many infections are made before being diagnosed in the latent period (13) , and (ii) the infection and removal rates being time varying to accommodate changing dynamics of the epidemics. There are three categories of actions countries can employ as part of the containment strategies: (i) reduce human contact and quarantine the confirmed infected cases to reduce the infectious rate; (ii) increase population virus screening and diagnosis; (iii) better medical treatments that shorten the recovery time from the disease. The three actions' epidemiological effects are well reflected in the expression of the effective reproduction number under the vSEIdR model (11): (1) where and are the infection rates in the pre-diagnosis Exposure state and the Infected state after diagnosis respectively, is the removal rate, is the diagnostic rate, and s(t) is the proportion of the susceptible population. The daily counts of infected, dead and recovered cases are obtained from data platforms of Johns Hopkins University (14) , World Health Organization (WHO) and Dingxiang Doctor. We did not consider data from China's Hubei province due to the incomplete observation before January 16, 2020. This actually makes the epidemics of the 25 countries more comparable as they are all started with imported cases. The start date for community transmission (DCT) of a country, reported in Table 1 , is determined by the first maximum of the estimated infection rate after the WHO local infection date to avoid the early epidemic period caused by imported cases. The study period is from DCT of each country up to April 20th. The COVID-19 related action date information of the countries is provided in Table S1 in the supplementary data (SD) based on both governmental and credible media sources. When a series of measures are implemented over a time window, we take the average date of the start and the ending dates of the time window as the action time. See Table S1 in SD for specifics. Ten countries have taken actions to counter COVID-19 in less than 13 days from their start dates of local transmission, which are considered as quick action countries, and the other 15 countries as slow action countries. Table 1 reports the estimated R t (see (11) for the estimation procedure) on the start (t = 0), which can be viewed as the basic reproduction number R 0 , and the average R t in Weeks 1-4 and Week 4 since the start date. The R t s measure the underlying reproduction dynamics of the infection beyond the more intuitive statistics, but are dependent with those statistics. Figure S1 in SD presents a scattered plot of the average R t in Weeks 2-4 and the cases per 100,000 population on April 20th, which shows significant correlation between the two variables. Taking quicker containment measures is shown to be effective in reducing the reproduction. Table 1 and China (non-Hubei) and Republic of Korea (South Korea) were the two nations that responded to the COVID-19 emergency the quickest among the 25 countries (see SD), and are found to be the most effective in bringing down the reproduction of COVID-19 in the first four weeks as shown in Table 1 and Behind South Korea and China's rapid declines in their R t s were two similar but not the same strategies. China's strategy was largely to suppress human contacts by limiting population movement, sealing off cities, enforcing high levels of self-isolation at homes and quarantining confirmed cases in newly built hospitals, which led to rapid decline in the contact rates and then in the infection rates and . In addition to limiting population contacts and a quick blockade of Daegu, South Korean conducted more active testing for potential infections in the population with more than half million tests being carried out in the first four weeks (18, 19) , which increased the diagnosis rate and hence reduced R t as implied by (1). Spain and Brazil were two early action countries. Their implementations were not effective as reflected by their rather high four-week average R t (Table 1) . Among the three Scandinavian countries Denmark, Norway and Sweden, Norway and Denmark took containment measures in 9 and 11.5 days, respectively, with the effects reflected in the quick declines of R t of the two countries ( Table 1 ). The Week 4 average R t were 1.07 and 1.68 for Denmark and Norway, representing 75% and 86% decline from their R 0 . In contrast, it took Sweden 26 days to put forward an action and its R t was much larger and slowly declined, with its average R t s in Weeks 2, 3 and 4 hovering near 2. The slow and ineffective actions made Sweden incur larger infection and death rates, which were 478 and 47 per 100,000 populations, respectively, as June 12, 2020 from WHO. In a sharp contrast, Denmark which has more than 5 times population density than Sweden had recorded just over 208 cases The nations have put forward a set of control measures as summarized in Table 1 supplemented by (Table S1 ). Although the average weekly reduction rates of R t were higher in the High-Level control group, no significant differences were detected between the two groups as shown in Table S2 in SD for details. Our study has two limitations. While the vSEIdR model is more realistic than the SIR and SEIR models, the asymptomatic cases and imported cases are not explicitly accounted for due to lack of data, which may cause bias in the estimation. While allowing infection in the latent stage reduces the bias caused by the asymptomatic cases, deaths and recoveries from asymptomatic cases are still unaccounted for. The bias caused by the imported cases is limited as we choose the DCT to avoid the very early stage of the epidemic largely caused by the imported cases, which is further helped by the fact that cross-country travel has been much dis-encouraged as the first set of counter-measures by countries. There are several critical lessons one can learn from the 25 countries' COVID-19 experiences. The first one is to take action as early as possible with vigorous enforcement to reduce the contact rates so as to reduce the infection rates and the R t . Acting early vigorously can hugely impact the infection size and thus lessen demands on medical resources down the track, and eventually improve the removal processes for those infected. The second lesson is to maintain a high level of the diagnostic rate to Association of Public Health Interventions With the Epidemiology of the COVID-19 Outbreak in Wuhan, China Evolving epidemiology and transmission dynamics of coronavirus disease 2019 outside Hubei province, China: a descriptive and modelling study The effect of control strategies to reduce social mixing on outcomes of the COVID-19 epidemic in Wuhan, China: a modelling study First-wave COVID-19 transmissibility and severity in China outside Hubei after control measures, and second-wave scenario planning: a modelling impact assessment An investigation of transmission control measures during the first 50 days of the COVID-19 epidemic in China The effect of travel restrictions on the spread of the 2019 novel coronavirus (COVID-19) outbreak Effect of Alert Level 4 on effective reproduction number: review of international COVID-19 cases Inferring the effective start dates of non-pharmaceutical interventions during COVID-19 outbreaks Effective Reproductive Number estimation for initial stage of COVID-19 pandemic in Latin American Countries Outbreak dynamics of COVID-19 in Europe and the effect of travel restrictions Better Strategies for Containing COVID-19 Epidemics -A Study of 25 Countries via an Extended Varying Coefficient SEIR Model The mathematics of infectious diseases Clinical characteristics of coronavirus disease 2019 in China An interactive web-based dashboard to track COVID-19 in real time Early transmission dynamics in Wuhan, China, of novel coronavirus-infected pneumonia Effective containment explains subexponential growth in recent confirmed COVID-19 cases in China The effect of human mobility and control measures on the COVID-19 epidemic in China COVID-19 test case total number South Korea Coronavirus cases have dropped sharply in South Korea. What's the secret to its success? Tracking Changes in SARS-CoV-2 Spike: Evidence that D614G Increases Infectivity of the COVID-19 Virus Acknowledgement We thank the Section Editor for insightful suggestions that have enriched the content of the work, and three reviewers for beneficial comments and suggestions.