Predicting SARS-CoV-2 variant spread in a completely seropositive population using semi-quantitative antibody measurements in blood donors (preprint)

Background SARS-CoV-2 serologic surveys estimate the proportion of the population with antibodies against historical variants which nears 100% in many settings. New analytic approaches are required to exploit the full information in serosurvey data. Method Using a SARS-CoV-2 anti-Spike (S) protein chemiluminescent microparticle assay, we attained a semi-quantitative measurement of population IgG titres in serial cross-sectional monthly samples of routine blood donations across seven Brazilian state capitals (March 2021-November 2021). In an ecological analysis (unit of analysis: age-city-calendar month) we assessed the relative contributions of prior attack rate and vaccination to antibody titre in blood donors. We compared blood donor anti-S titre across the seven cities during the growth phase of the Delta variant of concern (VOC) and use this to predict the resulting age-standardized incidence of severe COVID-19 cases. Results On average we tested 780 samples per month in each location. Seroprevalence rose to >95% across all seven capitals by November 2021. Driven proximally by vaccination, mean antibody titre increased 16-fold over the study. The extent of prior natural infection shaped this process, with the greatest increases in antibody titres occurring in cities with the highest prior attack rates. Mean anti-S IgG was a strong predictor (adjusted R2 =0.89) of the number of severe cases caused by the Delta VOC in the seven cities. Conclusions Semi-quantitative anti-S antibody titres are informative about prior exposure and vaccination coverage and can inform on the potential impact of future SARS-CoV-2 variants. Summary In the face of near 100% SARS-CoV-2 seroprevalence, we show that average semi-quantitative anti-S titre predicted the extent of the Delta variant’s spread in Brazil. This is a valuable metric for future seroprevalence studies.

SARS-CoV-2 antibody dynamics in blood donors and COVID-19 epidemiology in eight Brazilian state capitals (preprint)

There are large differences in the shape and size of regional SARS-CoV-2 epidemics in Brazil. Here we tested monthly blood donation samples for IgG antibodies from March 2020 to March 2021 in eight of Brazil’s most populous cities. There was large variation in the inferred attack rate adjusted for seroreversion across cities, and seroprevalence was consistently smaller in women and donors older than 55 years. The age-specific infection fatality rate differed between cities and consistently increased with age. The infection hospitalisation rate (IHR) increased significantly during the gamma-dominated second wave in Manaus, suggesting increased morbidity of the Gamma VOC compared to previous variants circulating in Manaus. The higher disease penetrance associated with the health system’s collapse increased the overall IFR by a minimum factor of 2.91 (95% CrI 2.43–3.53). These results demonstrate large heterogeneity in epidemic spread and highlight the utility of blood donor serosurveillance to monitor SARS-CoV-2 epidemics.

Report 46: Factors driving extensive spatial and temporal fluctuations in COVID-19 fatality rates in Brazilian hospitals (preprint)

The SARS-CoV-2 Gamma variant spread rapidly across Brazil, causing substantial infection and death waves. We use individual-level patient records following hospitalisation with suspected or confirmed COVID-19 to document the extensive shocks in hospital fatality rates that followed Gamma’s spread across 14 state capitals, and in which more than half of hospitalised patients died over sustained time periods. We show that extensive fluctuations in COVID-19 in-hospital fatality rates also existed prior to Gamma’s detection, and were largely transient after Gamma’s detection, subsiding with hospital demand. Using a Bayesian fatality rate model, we find that the geographic and temporal fluctuations in Brazil’s COVID-19 in-hospital fatality rates are primarily associated with geographic inequities and shortages in healthcare capacity. We project that approximately half of Brazil’s COVID-19 deaths in hospitals could have been avoided without pre-pandemic geographic inequities and without pandemic healthcare pressure. Our results suggest that investments in healthcare resources, healthcare optimization, and pandemic preparedness are critical to minimize population wide mortality and morbidity caused by highly transmissible and deadly pathogens such as SARS-CoV-2, especially in low- and middle-income countries. Note The following manuscript has appeared as ‘Report 46 - Factors driving extensive spatial and temporal fluctuations in COVID-19 fatality rates in Brazilian hospitals’ at https://spiral.imperial.ac.uk:8443/handle/10044/1/91875 . One sentence summary COVID-19 in-hospital fatality rates fluctuate dramatically in Brazil, and these fluctuations are primarily associated with geographic inequities and shortages in healthcare capacity.

Scale-free dynamics of Covid-19 in a Brazilian city (preprint)

Mathematical models can provide insights into the control of pandemic COVID-19, which remains a global priority. The dynamics of directly-transmitted infectious diseases, such as COVID-19, are usually described by compartmental models where individuals are classified as susceptible, infected and removed. These SIR models typically assume homogenous transmission of infection, even in large populations, a simplification that is convenient but inconsistent with observations. Here we use original data on the dynamics of COVID-19 spread in a Brazilian city to investigate the structure of the transmission network. We find that transmission can be described by a network in which each infectious individual has a small number of susceptible contacts, of the order of 2-5, which is independent of total population size. Compared with standard models of homogenous mixing, this scale-free, fractal infection process gives a better description of COVID-19 dynamics through time. In addition, the contact process explains the geographically localized clusters of disease seen in this Brazilian city. Our scale-free model can help refine criteria for physical and social distancing in order to more effectively mitigate the spread of COVID-19. We propose that scale-free COVID-19 dynamics could be a widespread phenomenon, a topic for further investigation.

Asymptomatic SARS-CoV-2 infection and the demography of COVID-19 (preprint)

Asymptomatic individuals carrying SARS-CoV-2 can transmit the virus and contribute to outbreaks of COVID-19, but it is not yet clear how the proportion of asymptomatic infections varies by age and geographic location. Here we use detailed surveillance data gathered during COVID-19 resurgences in six cities of China at the beginning of 2021 to investigate this question. Data were collected by multiple rounds of city-wide PCR test with detailed contact tracing, where each patient was monitored for symptoms through the whole course of infection. We find that the proportion of asymptomatic infections declines with age (coefficient =-0.006, P<0.01), falling from 56% in age group 0-9 years to 12% in age group >60 years. Using an age-stratified compartment model, we show that this age-dependent asymptomatic pattern together with the age distribution of overall cases can explain most of the geographic differences in reported asymptomatic proportions. Combined with demography and contact matrices from other countries worldwide, we estimate that a maximum of 22%-55% of SARS-CoV-2 infections would come from asymptomatic cases in an uncontrolled epidemic based on asymptomatic proportions in China. Our analysis suggests that flare-ups of COVID-19 are likely if only adults are vaccinated and that surveillance and possibly control measures among children will be still needed in the future to contain epidemic resurgence.

Global disparities in SARS-CoV-2 genomic surveillance (preprint)

Genomic sequencing provides critical information to track the evolution and spread of SARS-CoV-2, optimize molecular tests, treatments and vaccines, and guide public health responses. To investigate the spatiotemporal heterogeneity in the global SARS-CoV-2 genomic surveillance, we estimated the impact of sequencing intensity and turnaround times (TAT) on variant detection in 167 countries. Most countries submit genomes >21 days after sample collection, and 77% of low and middle income countries sequenced <0.5% of their cases. We found that sequencing at least 0.5% of the cases, with a TAT <21 days, could be a benchmark for SARS-CoV-2 genomic surveillance efforts. Socioeconomic inequalities substantially impact our ability to quickly detect SARS-CoV-2 variants, and undermine the global pandemic preparedness. One-Sentence SummarySocioeconomic inequalities impacted the SARS-CoV-2 genomic surveillance, and undermined the global pandemic preparedness.

Reinfection by the SARS-CoV-2 P.1 variant in blood donors in Manaus, Brazil (preprint)

The city of Manaus, north Brazil, was stricken by a severe epidemic of SARS-Cov-2 in March 2020, reaching a seroprevalence of 76% by October 2020. Nevertheless, in late November an abrupt increase in hospitalizations and deaths hit Manaus, causing higher number of deaths compared to the first epidemic wave. It has been hypothesized that virus lineages circulating in the second wave, namely the P.1 variant of concern first detected in early December in Manaus, could be better at evading immunity generated in response to previous infection with other lineages. In order to estimate the reinfection rate during the resurgence of SARS-CoV-2 in Manaus, we tested serial samples from 238 unvaccinated repeat blood donors using a SARS-CoV-2 anti-N IgG chemiluminescence microparticle assay. Blood donors were divided into six groups that reflected the inferred sequence of infection and reinfection with non-P.1 and P.1 variants. We assumed that reinfections induce a recrudescence (or boosting) of plasma anti-N IgG antibody levels, yielding a V-shaped time series of antibody reactivity levels. We infer that 16.9% (95% CI [9.48%, 28.5%]) of all presumed P.1 infections that were observed in 2021 were reinfections. If we also include cases of probable or possible reinfections (defined by considering the time period when the antibody levels are expected to grow after recovery and the range of half-lives for antibody waning after seroconversion), these percentages increase respectively to 25.8% (95% CI [16.7%, 37.4%]), and 31.0% (95% CI [21.4%, 42.5%]). Our data suggest that reinfection due to P.1 is common and more frequent than what has been detected by traditional epidemiologic, molecular and genomic surveillance of clinical cases.

Genomics and epidemiology of a novel SARS-CoV-2 lineage in Manaus, Brazil (preprint)

Cases of SARS-CoV-2 infection in Manaus, Brazil, resurged in late 2020, despite high levels of previous infection there. Through genome sequencing of viruses sampled in Manaus between November 2020 and January 2021, we identified the emergence and circulation of a novel SARS-CoV-2 variant of concern, lineage P.1, that acquired 17 mutations, including a trio in the spike protein (K417T, E484K and N501Y) associated with increased binding to the human ACE2 receptor. Molecular clock analysis shows that P.1 emergence occurred around early November 2020 and was preceded by a period of faster molecular evolution. Using a two-category dynamical model that integrates genomic and mortality data, we estimate that P.1 may be 1.4–2.2 times more transmissible and 25-61% more likely to evade protective immunity elicited by previous infection with non-P.1 lineages. Enhanced global genomic surveillance of variants of concern, which may exhibit increased transmissibility and/or immune evasion, is critical to accelerate pandemic responsiveness. One-Sentence Summary We report the evolution and emergence of a SARS-CoV-2 lineage of concern associated with rapid transmission in Manaus.

Levels of SARS-CoV-2 Lineage P.1 Neutralization by Antibodies Elicited after Natural Infection and Vaccination (preprint)

Background: A new SARS-CoV-2 lineage, named P.1 (20J/501Y.V3), has recently been detected in Brazil. Mutations accrued by the P.1 lineage include amino acid changes in the receptor-binding domain of the spike protein that also are reported in variants of concern in the United Kingdom (B.1.1.7) and South Africa (B.1.325).Methods: We isolated two P.1-containing specimens from nasopharyngeal and bronchoalveolar lavage samples of patients of Manaus, Brazil. We measured neutralization of the P.1 virus after incubation with the plasma of 19 COVID-19 convalescent blood donors and recipients of the chemically-inactivated CoronaVac vaccine and compared these results to neutralization of a SARS-CoV-2 B-lineage previously circulating in Brazil.Findings: The immune plasma of COVID-19 convalescent blood donors had 6-fold less neutralizing capacity against the P.1 than against the B-lineage. Moreover, five months after booster immunization with CoronaVac, plasma from vaccinated individuals failed to efficiently neutralize P.1 lineage isolates.Interpretation: These data indicate that the P.1 lineage may escape from neutralizing antibodies generated in response to polyclonal stimulation against previously circulating variants of SARS-CoV-2.Funding: São Paulo Research Foundation, MCTI/FINEP, Medical Research Council, National Council for Scientific and Technological Development, National Institutes of Health.Conflict of Interest: M.S.D. is a consultant for Inbios, Vir Biotechnology, NGMBiopharmaceuticals, and Carnival Corporation, and on the Scientific Advisory Boards of Moderna and Immunome. The Diamond laboratory has received funding support in sponsored research agreements from Moderna, Vir Biotechnology, and Emergent BioSolutions.Ethical Approval: All procedures followed the ethical standards of the responsible committee on humanexperimentation and approved by the ethics committees from the University of Campinas, Brazil (Approval number CONEP 4.021.484 for plasma collection of blood donors, CAEE32078620.4.0000.5404 and 30227920.9.0000.5404 for the sampling of vaccinated and viral genome sequencing, respectively). All patient data were anonymized before study inclusion.Informed consent was obtained from all subjects for being included in the study.

MODELLING PRESYMPTOMATIC INFECTIOUSNESS IN COVID-19 (preprint)

This paper considers SEPIR, the extension of an existing parametric SEIR continuous simulation compartment model. Both models can be fitted to real data as they include parameters that can simply be estimated from the data. However SEPIR deploys an additional presymptomatic (also called asymptomatic) infectious stage that is not included in SEIR but which is known to exist in COVID-19. This stage is also parametrised and so can be fitted to data. Both SEPIR and the existing SEIR model assume a homogeneous mixing population, an idealisation that is unrealistic in practice when dynamically varying control strategies are deployed against virus. This means that if either model is to represent more than just a single period in the behaviour of the epidemic, then the parameters of the model will have to be time dependent. This issue is also discussed in this paper. A numerical example using Swiss observational data is given.

COVID-19 herd immunity in the Brazilian Amazon (preprint)

The herd immunity threshold is the proportion of a population that must be immune to an infectious disease, either by natural infection or vaccination such that, in the absence of additional preventative measures, new cases decline and the effective reproduction number falls below unity. This fundamental epidemiological parameter is still unknown for the recently-emerged COVID-19, and mathematical models have predicted very divergent results. Population studies using antibody testing to infer total cumulative infections can provide empirical evidence of the level of population immunity in severely affected areas. Here we show that the transmission of SARS-CoV-2 in Manaus, located in the Brazilian Amazon, increased quickly during March and April and declined more slowly from May to September. In June, one month following the epidemic peak, 44% of the population was seropositive for SARS-CoV-2, equating to a cumulative incidence of 52%, after correcting for the false-negative rate of the antibody test. The seroprevalence fell in July and August due to antibody waning. After correcting for this, we estimate a final epidemic size of 66%. Although non-pharmaceutical interventions, plus a change in population behavior, may have helped to limit SARS-CoV-2 transmission in Manaus, the unusually high infection rate suggests that herd immunity played a significant role in determining the size of the epidemic.

Baseline T cell immune phenotypes predict virologic and disease control upon SARS-CoV infection (preprint)

The COVID-19 pandemic has revealed that infection with SARS-CoV-2 can result in a wide range of clinical outcomes in humans, from asymptomatic or mild disease to severe disease that can require mechanical ventilation. An incomplete understanding of immune correlates of protection represents a major barrier to the design of vaccines and therapeutic approaches to prevent infection or limit disease. This deficit is largely due to the lack of prospectively collected, pre-infection samples from indiviuals that go on to become infected with SARS-CoV-2. Here, we utilized data from a screen of genetically diverse mice from the Collaborative Cross (CC) infected with SARS-CoV to determine whether circulating baseline T cell signatures are associated with a lack of viral control and severe disease upon infection. SARS-CoV infection of CC mice results in a variety of viral load trajectories and disease outcomes. Further, early control of virus in the lung correlates with an increased abundance of activated CD4 and CD8 T cells and regulatory T cells prior to infections across strains. A basal propensity of T cells to express IFNg and IL17 over TNFa also correlated with early viral control. Overall, a dysregulated, pro-inflammatory signature of circulating T cells at baseline was associated with severe disease upon infection. While future studies of human samples prior to infection with SARS-CoV-2 are required, our studies in mice with SARS-CoV serve as proof of concept that circulating T cell signatures at baseline can predict clinical and virologic outcomes upon SARS-CoV infection. Identification of basal immune predictors in humans could allow for identification of individuals at highest risk of severe clinical and virologic outcomes upon infection, who may thus most benefit from available clinical interventions to restrict infection and disease.

Comparative analysis reveals the species-specific genetic determinants of ACE2 required for SARS-CoV-2 entry (preprint)

Coronavirus interaction with viral receptor is a primary genetic determinant of host range and tissue tropism. SARS-CoV-2 utilizes ACE2 as the receptor to enter the host cell in a species-specific manner. We and others have previously shown that ACE2 orthologs from New World monkeys, koala and mouse cannot interact with SARS-CoV-2 to mediate viral entry, and this defect can be restored by humanization of the restrictive residues in New World monkey ACE2. To better understand the genetic determinants of susceptibility of ACE2 orthologs to viral entry, we compared koala and mouse ACE2 sequences with human ortholog, and identified the key residues in koala or mouse ACE2 that restrict its viral receptor activity. Humanization of these critical residues could render the capabilities of koala and mouse ACE2 to bind viral spike protein and facilitate the viral entry. Our work identifies the genetic determinant of ACE2 for SARS-CoV-2 susceptibility, and a single mutation could restore the mouse ACE2 receptor activity, providing a potential avenue for the development of mouse model of SARS-CoV-2.

Cryo-EM structure of S-Trimer, a subunit vaccine candidate for COVID-19 (preprint)

Less than a year after its emergence, the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has infected over 22 million people worldwide with a death toll approaching 1 million. Vaccination remains the best hope to ultimately put this pandemic to an end. Here, using Trimer-Tag technology, we produced both wild-type (WT) and furin site mutant (MT) S-Trimers for COVID-19 vaccine studies. Cryo-EM structures of the WT and MT S-Trimers, determined at 3.2 Angstrom and 2.6 Angstrom respectively, revealed that both antigens adopt a tightly closed conformation and their structures are essentially identical to that of the previously solved full-length WT S protein in detergent. These results validate Trimer-Tag as a platform technology in production of metastable WT S-Trimer as a candidate for COVID-19 subunit vaccine.

SARS-CoV-2 Nsp1 suppresses host but not viral translation through a bipartite mechanism (preprint)

The Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) is a highly contagious virus that underlies the current COVID-19 pandemic. SARS-CoV-2 is thought to disable various features of host immunity and cellular defense. The SARS-CoV-2 nonstructural protein 1 (Nsp1) is known to inhibit host protein translation and could be a target for antiviral therapy against COVID-19. However, how SARS-CoV-2 circumvents this translational blockage for the production of its own proteins is an open question. Here, we report a bipartite mechanism of SARS-CoV-2 Nsp1 which operates by: (1) hijacking the host ribosome via direct interaction of its C-terminal domain (CT) with the 40S ribosomal subunit and (2) specifically lifting this inhibition for SARS-CoV-2 via a direct interaction of its N-terminal domain (NT) with the 5 untranslated region (5 UTR) of SARS-CoV-2 mRNA. We show that while Nsp1-CT is sufficient for binding to 40S and inhibition of host protein translation, the 5 UTR of SARS-CoV-2 mRNA removes this inhibition by binding to Nsp1-NT, suggesting that the Nsp1-NT-UTR interaction is incompatible with the Nsp1-CT-40S interaction. Indeed, lengthening the linker between Nsp1-NT and Nsp1-CT of Nsp1 progressively reduced the ability of SARS-CoV-2 5 UTR to escape the translational inhibition, supporting that the incompatibility is likely steric in nature. The short SL1 region of the 5 UTR is required for viral mRNA translation in the presence of Nsp1. Thus, our data provide a comprehensive view on how Nsp1 switches infected cells from host mRNA translation to SARS-CoV-2 mRNA translation, and that Nsp1 and 5 UTR may be targeted for anti-COVID-19 therapeutics.

An immunodominance hierarchy exists in CD8+ T cell responses to HLA-A*02:01-restricted epitopes identified from the non-structural polyprotein 1a of SARS-CoV-2. (preprint)

COVID-19 vaccines are being rapidly developed and human trials are underway. Almost all of these vaccines have been designed to induce antibodies targeting spike protein of SARS-CoV-2 in expectation of neutralizing activities. However, non-neutralizing antibodies are at risk of causing antibody-dependent enhancement. Further, the longevity of SARS-CoV-2-specific antibodies is very short. Therefore, in addition to antibody-induced vaccines, novel vaccines on the basis of SARS-CoV-2-specific cytotoxic T lymphocytes (CTLs) should be considered in the vaccine development. Here, we attempted to identify HLA-A*02:01-restricted CTL epitopes derived from the non-structural polyprotein 1a of SARS-CoV-2. Eighty-two peptides were firstly predicted as epitope candidates on bioinformatics. Fifty-four in 82 peptides showed high or medium binding affinities to HLA-A*02:01. HLA-A*02:01 transgenic mice were then immunized with each of the 54 peptides encapsulated into liposomes. The intracellular cytokine staining assay revealed that 18 out of 54 peptides were CTL epitopes because of the induction of IFN-{gamma}-producing CD8+ T cells. In the 18 peptides, 10 peptides were chosen for the following analyses because of their high responses. To identify dominant CTL epitopes, mice were immunized with liposomes containing the mixture of the 10 peptides. Some peptides were shown to be statistically predominant over the other peptides. Surprisingly, all mice immunized with the liposomal 10 peptide mixture did not show the same reaction pattern to the 10 peptides. There were three pattern types that varied sequentially, suggesting the existence of an immunodominance hierarchy, which may provide us more variations in the epitope selection for designing CTL-based COVID-19 vaccines. ImportanceFor the development of vaccines based on SARS-CoV-2-specific cytotoxic T lymphocytes (CTLs), we attempted to identify HLA-A*02:01-restricted CTL epitopes derived from the non-structural polyprotein 1a of SARS-CoV-2. Out of 82 peptides predicted on bioinformatics, 54 peptides showed good binding affinities to HLA-A*02:01. Using HLA-A*02:01 transgenic mice, 18 in 54 peptides were found to be CTL epitopes in the intracellular cytokine staining assay. Out of 18 peptides, 10 peptides were chosen for the following analyses because of their high responses. To identify dominant epitopes, mice were immunized with liposomes containing the mixture of the 10 peptides. Some peptides were shown to be statistically predominant. Surprisingly, all immunized mice did not show the same reaction pattern to the 10 peptides. There were three pattern types that varied sequentially, suggesting the existence of an immunodominance hierarchy, which may provide us more variations in the epitope selection for designing CTL-based COVID-19 vaccines.

The scale and dynamics of COVID-19 epidemics across Europe (preprint)

The frequently stated goal of COVID-19 control is to 'flatten the curve'; that is, to slow the epidemic by limiting contacts between infectious and susceptible individuals, and to reduce and delay peak numbers of cases and deaths. Here we investigate how the scale and dynamics of COVID-19 epidemics differ among 26 European countries in which the numbers of reported deaths varied more than 100-fold. Under lockdown, countries reporting fewer deaths in total also had lower peak death rates, as expected, but shorter rather than longer periods of epidemic growth. This empirical analysis highlights the benefits of intervening early to curtail COVID-19 epidemics: the cumulative number of deaths reported in each country by 20 June was 60-200 times the number reported by the date of lockdown.

Risk of COVID-19 is associated with long-term exposure to air pollution (preprint)

High risk of severe disease of COVID-19 has been associated with patients with chronic obstructive pulmonary disease, cardiovascular disease or hypertension, and long-term exposure to PM2.5 has been associated with COVID-19 mortality. We collate individual level data of confirmed COVID-19 cases during the first wave of the epidemic in mainland China by March 6, 2020. We pair these data with a mobile phone dataset, covering human movements from Wuhan before the travel ban and inner-city movements during the time of emergency response from 324 cities in China. Adjusting for socio-economic factors, an increase of 10 g/m3 in NO2 or PM2.5 was found to be associated with a 22.41% (95%CI: 7.28%-39.89%) or 15.35% (95%CI: 5.60%-25.98%) increase in the number of COVID-19 cases, and a 19.20% (95%CI: 4.03%-36.59%) or 9.61% (95%CI: 0.12%-20.01%) increase in severe infection, respectively. Our results highlight the importance of air quality improvements to health benefits.

The effect of human mobility and control measures on the COVID-19 epidemic in China (preprint)

The ongoing COVID-19 outbreak has expanded rapidly throughout China. Major behavioral, clinical, and state interventions are underway currently to mitigate the epidemic and prevent the persistence of the virus in human populations in China and worldwide. It remains unclear how these unprecedented interventions, including travel restrictions, have affected COVID-19 spread in China. We use real-time mobility data from Wuhan and detailed case data including travel history to elucidate the role of case importation on transmission in cities across China and ascertain the impact of control measures. Early on, the spatial distribution of COVID-19 cases in China was well explained by human mobility data. Following the implementation of control measures, this correlation dropped and growth rates became negative in most locations, although shifts in the demographics of reported cases are still indicative of local chains of transmission outside Wuhan. This study shows that the drastic control measures implemented in China have substantially mitigated the spread of COVID-19.

Early evaluation of the Wuhan City travel restrictions in response to the 2019 novel coronavirus outbreak (preprint)

Respiratory illness caused by a novel coronavirus (COVID-19) appeared in China during December 2019. Attempting to contain infection, China banned travel to and from Wuhan city on 23 January and implemented a national emergency response. Here we evaluate the spread and control of the epidemic based on a unique synthesis of data including case reports, human movement and public health interventions. The Wuhan shutdown slowed the dispersal of infection to other cities by an estimated 2.91 days (95%CI: 2.54-3.29), delaying epidemic growth elsewhere in China. Other cities that implemented control measures pre-emptively reported 33.3% (11.1-44.4%) fewer cases in the first week of their outbreaks (13.0; 7.1-18.8) compared with cities that started control later (20.6; 14.5-26.8). Among interventions investigated here, the most effective were suspending intra-city public transport, closing entertainment venues and banning public gatherings. The national emergency response delayed the growth and limited the size of the COVID-19 epidemic and, by 19 February (day 50), had averted hundreds of thousands of cases across China.