key: cord-0963078-y4nocgxk authors: Lei, Hao; Wu, Xifeng; Wang, Xiao; Xu, Modi; Xie, Yu; Du, Xiangjun; Cowling, Benjamin J; Li, Yuguo; Shu, Yuelong title: Different transmission dynamics of COVID-19 and influenza suggest the relative efficiency of isolation/quarantine and social distancing against COVID-19 in China date: 2020-10-20 journal: Clin Infect Dis DOI: 10.1093/cid/ciaa1584 sha: a9b38a14792178a739bb06a2d304300e7ba9e2a1 doc_id: 963078 cord_uid: y4nocgxk BACKGROUND: Non-pharmaceutical interventions (NPIs) against Coronavirus Disease 2019 (COVID-19) are vital to reducing the transmission risks. However, the relative efficiency of social distancing against COVID-19 remains controversial, since social distancing and isolation/quarantine were implemented almost at the same time in China. METHODS: In this study, surveillance data of COVID-19 and seasonal influenza in the year 2018-2020 were used to quantify the relative efficiency of NPIs against COVID-19 in China, since isolation/quarantine was not used for the influenza epidemics. Given that the relative age-dependent susceptibility to influenza and COVID-19 may vary, an age-structured Susceptible-Infected-Recovered model was built to explore the efficiency of social distancing against COVID-19 under different population susceptibility scenarios. RESULTS: The mean effective reproductive number, R(t), of COVID-19 before NPIs was 2.12 (95% confidential interval (CI): 2.02-2.21). By March 11, 2020, the overall reduction in R(t) of COVID-19 was 66.1% (95% CI: 60.1%-71.2%). In the epidemiological year 2019/20, influenza transmissibility reduced by 34.6% (95% CI: 31.3%-38.2%) compared with that in the epidemiological year 2018/19. Under the observed contact patterns changes in China, social distancing had similar efficiency against COVID-19 in three different scenarios. By assuming same efficiency of social distancing against seasonal influenza and COVID-19 transmission, isolation/quarantine and social distancing could lead to a 48.1% (95% CI: 35.4%-58.1%) and 34.6% (95% CI: 31.3%-38.2%) reduction of the transmissibility of COVID-19. CONCLUSIONS: Though isolation/quarantine is more effective than social distancing, given that typical basic reproductive number of COVID-19 is 2-3, isolation/quarantine alone could not contain the COVID-19 pandemic effectively in China. Implementing the most effective interventions has become a top public health priority in the ongoing Coronavirus Disease 2019 (COVID- 19) pandemic. Before a highly efficacious vaccine that is accepted by the majority of the population or an antiviral drug is available, only non-pharmaceutical interventions (NPIs) could substantially control infection transmission, based on lessons learned from previous outbreaks, including severe acute respiratory syndrome (SARS), Ebola and the pandemic H1N1 2009. 0-3 However, the relative efficiency of different NPIs remains controversial. To suppress the COVID-19 pandemic in China, the Chinese government implemented a set of NPIs, including not only the classical isolation of the confirmed/suspected cases and quarantine of their close contacts in special facilities, but also unprecedented measures like strict community containments with social distancing. 4 To date, the COVID-19 pandemic has been under control in China. Since different NPIs were implemented almost at the same time, it is difficult to directly quantify the relative efficiency of each individual NPI. Nonetheless, efforts are being made to address this critical question about the relative efficiency of particular NPIs for COVID-19. This is particularly important and time-sensitive as countries around the world devise strategies to rapidly respond to the pandemic. One potential way to assess the relative efficiency of NPIs for COVID-19 is to use data and knowledge about a concurrent and similarly transmitted disease: influenza. Both influenza and COVID-19 are highly contagious respiratory infection diseases, and in China, the seasonal influenza epidemics always peak in January-February, 5 the COVID-19 pandemic which peaked in January 2020, also covered a period that overlapped with the flu season in China. On the one hand, some NPIs that were implemented against COVID-19, such as school and workplace closures, have already been proven to be effective to control influenza transmission. 6, 7 On the other hand, some other NPIs were specifically implemented for COVID-19, including immediate isolation of the COVID-19 confirmed cases, and quarantine of their close contacts in dedicated quarantine facilities. This situation provides a good opportunity to explore the relative efficacy of different NPIs against the COVID-19 pandemic, by simply assuming a similar transmission mode and efficiency of NPIs against influenza and COVID-19. In this study, the transmission dynamics of seasonal influenza and COVID-19 were studies and compared to evaluate the relative efficiency of NPIs against COVID-19 in China. The weekly reports of influenza surveillance data from April 2018 to March 2020, originally provided by the Chinese National Influenza Center (CNIC), were downloaded from the WHO Flunet. More detailed information about the dataset is in the Supplementary 1. In accordance with earlier studies, 8, 9 a count more precisely representing the influenza infections was used in this study, the weekly incidence rate (henceforth, "incidence rate") of influenza, The weekly incidence rate was calculated by multiplying the influenza-like illness (ILI) rate among patients visiting sentinel hospitals by the viral detection positive rate. The weekly incidence rate was then interpolated to daily incidence rate using splines. 7 Though from January 1, 2020, the Huanan seafood wholesale market, where the COVID-19 was firstly discovered, was closed, community-wide implementation of NPIs against COVID-19 began on January 23, 2020, when the first three provinces began a 1-level response, and within the following two days, 27 additional provinces began a 1-level response ( Figure 3 ). In this study, it was assumed that NPIs against COVID-19 began on January 23, 2020. The NPIs used in China to contain COVID-19 transmission included isolation and quarantine, social distancing, and community containment measures. 10 Isolation/ quarantine is the isolation of the confirmed COVID-19 cases and/or all suspected cases, and quarantine of their close contacts during the incubation period in special quarantine facilities. Social distancing is designed to reduce interactions between people in a broader community, while community containment is also designed to reduce personal interaction, but applied to an entire community. 10 Intervention 2: Social distancing. The aim of social distancing was to reduce interactions between people. As disease transmitted by respiratory droplets required a certain proximity of people, so in theory, social distancing of people will reduce influenza and COVID-19 transmission. The efficiency of NPIs against influenza and COVID-19 was evaluated based on changes of the effective reproductive number, R t , which represents the mean number of secondary cases that were infected by a primary case of infection at time t. Time-varying estimates of the effective reproductive number were made using the R package EpiEstim, assuming a mean serial interval of 4.7 days and a standard deviation of 2.9 days for COVID-19, 11 and a mean serial interval of 2.85 days and a standard deviation of 0.93 days for influenza. 12 Changes in transmissibility were evaluated by comparing the R t values before and after NPIs. All analyses were conducted with R version 3.6.3. to COVID-19, it was assumed that =1. By using the log-linear multivariable regression model, the parameter could be estimated. Denoting the efficiency of social distancing and isolation/quarantine on reducing the infection transmissibility to be and respectively. Assuming that social distancing and isolation/quarantine are independent to containing infection transmission, then the overall efficiency of social distancing and isolation/quarantine could be 1-(1-)(1-). In addition, since population susceptibility to influenza and COVID-19 varies, the efficiency of social distancing against influenza and COVID-19 might also vary. In this study, an agestructured Susceptible-Infected-Recovered (SIR) model was built, in which population were divided into three groups: (1) the young, with age less than 18 years old, (2) adults, age between 18 and 60, and (3) the old, with age more than 60 years old. The percentage of the young, adults and the old are 16.9%, 71.2% and 11.9% respectively in China. 14 More detailed information about the age-structured SIR model was in the Supplementary 3. The contact rates before and after NPIs in China were from Zhang et al. 15 Generally, daily contacts were reduced 7-8-fold during the COVID-19 intervention period. 15 The relative susceptibility ( ) to influenza of the children, adult and elderly was assumed to be 2:1:2, 16 since the children and elderly are at high risk from influenza. For COVID-19, since some study reported that the susceptibility increases with age, 15 some study showed that the susceptibility did not change with age. 17 We considered the following three scenarios. Scenario 1: The relative susceptibility to COVID-19 did not change with age, and of three age groups were assumed to be 1:1:1. Scenario 2: The relative susceptibility to COVID-19 increased with age, and of three age groups were assumed to be 1:2:4. Scenario 3: The relative susceptibility to COVID-19 of three age groups were assumed to be 2:1:2, same with seasonal influenza. Though This seems unexpected, there were following reasons. Following NPIs implementation, the total number of hospitalizations declined sharply (Figure 1a) . Furthermore, since some of the symptoms of COVID-19 confirmed/suspected cases were similar to those seen in influenza cases, such as fever and cough, the ILI rate during the first two weeks after NPIs would be expected to be much higher than the real situation ( Figure 1b) . Therefore, the incidence rate of influenza, which was calculated by multiplying the ILI rate among patients visiting sentinel hospitals by the viral detection positive rate during this period, would also be expected to be higher than the real situation. After the first two weeks following NPIs implementation, when the ILI rate was close to that in the year 2018/19 (Figure 1b) , the estimated R t declined quickly then (Figure 4a ). This study estimated the basic reproductive number for COVID-19 as 2.12 (95% CI: 2.02-2.21), which is close to 2.2-2.68 reported in the literature. 18, 19 This reproductive number is much higher than that of seasonal influenza, which is 1.28 (interquartile range: 1.19-1.37), 20 so it is more difficult to control COVID-19 than influenza. The estimated reduction in the reproductive number of seasonal influenza epidemic during the Chinese school winter holiday alone was 12.5% (95% CI: 9.1%-16.5%, Phase 2 VS Phase 1) in the years 2018/19, which coincided with studies in Hong Kong SAR, China and South Korea. 6, 7, 13, 21 In Social distancing interventions have been shown to be effective against COVID-19. 22 In three scenarios, social distancing has similar efficiency against COVID-19 transmission. The estimated 48.1% reduction of transmissibility due to isolation/quarantine intervention coincides with the results from a retrospective cohort study in Guangzhou, China, in which it was estimated that prompt isolation/quarantine only reduced 20-50% of the effective reproductive number of COVID-19 in Guangzhou. 23 Given that the typical basic reproductive number of COVID-19 was 2-3, isolation/quarantine alone is not enough to contain COVID-19 transmission in China. In contrast, isolation and quarantine was effective for SARS in 2003. 10 The reason could be that viral shedding peaks at 6-11 days after onset of illness, enabling early isolation before massive transmission. 10 Transmission dynamics of the etiological agent of SARS in Hong Kong: impact of public health interventions Likely effectiveness of pharmaceutical and nonpharmaceutical interventions for mitigating influenza virus transmission in Mongolia The effect of reactive school closure on community influenza-like illness counts in the state of Michigan during the 2009 H1N1 pandemic Report of the WHO-China Joint Mission on Coronavirus Disease 2019 (COVID-19) Characterization of regional influenza seasonality patterns in China and implications for vaccination strategies: spatio-temporal modeling of surveillance data Inferring influenza infection attack rate from seroprevalence data Mitigation of Influenza B Epidemic with School Closures Predicting the epidemic sizes of influenza A/H1N1, A/H3N2, and B: a statistical method Real-time influenza forecasts during the 2012-2013 season Can we contain the COVID-19 outbreak with the same measures as for SARS? Serial interval of novel coronavirus (2019-nCoV) infections How generation intervals shape the relationship between growth rates and reproductive numbers Effects of school holidays on seasonal influenza in South Korea Changes in contact patterns shape the dynamics of the COVID-19 outbreak in China Household transmission of 2009 pandemic influenza A(H1N1) in the United States Epidemiology and Transmission of COVID-19 in Shenzhen China: Analysis of 391 cases and 1,286 of their close contacts Early Transmission Dynamics in Wuhan, China, of Novel Coronavirus-Infected Pneumonia Nowcasting and forecasting the potential domestic and international spread of the 2019-nCoV outbreak originating in Wuhan, China: a modelling study Estimates of the reproduction number for seasonal, pandemic, and zoonotic influenza: a systematic review of the literature Impact assessment of non-pharmaceutical interventions against COVID-19 and influenza in Hong Kong: an observational study An investigation of transmission control measures during the first 50 days of the COVID-19 epidemic in China Household secondary attack rate of COVID-19 and associated determinants in Guangzhou, China: a retrospective cohort study Temporal dynamics in viral shedding and transmissibility of COVID-19 Quantifying SARS-CoV-2 transmission suggests epidemic control with digital contact tracing Non-pharmaceutical interventions used to control COVID-19 reduced seasonal influenza transmission in China Clinical characteristics of coronavirus disease 2019 in China Incubation periods of acute respiratory viral infections: a systematic review Does influenza transmission occur from asymptomatic infection or prior to symptom onset? Reconstruction of the full transmission dynamics of COVID-19 in Wuhan All authors declare no competing interests.