RESUMO
BackgroundReal-time genomic sequencing has played a major role in tracking the global spread and local transmission of SARS-CoV-2, contributing greatly to disease mitigation strategies. After effectively eliminating the virus, New Zealand experienced a second outbreak of SARS-CoV-2 in August 2020. During this August outbreak, New Zealand utilised genomic sequencing in a primary role to support its track and trace efforts for the first time, leading to a second successful elimination of the virus. MethodsWe generated the genomes of 80% of the laboratory-confirmed samples of SARS-CoV-2 from New Zealands August 2020 outbreak and compared these genomes to the available global genomic data. FindingsGenomic sequencing was able to rapidly identify that the new COVID-19 cases in New Zealand belonged to a single cluster and hence resulted from a single introduction. However, successful identification of the origin of this outbreak was impeded by substantial biases and gaps in global sequencing data. InterpretationAccess to a broader and more heterogenous sample of global genomic data would strengthen efforts to locate the source of any new outbreaks. FundingThis work was funded by the Ministry of Health of New Zealand, New Zealand Ministry of Business, Innovation and Employment COVID-19 Innovation Acceleration Fund (CIAF-0470), ESR Strategic Innovation Fund and the New Zealand Health Research Council (20/1018 and 20/1041).
RESUMO
BackgroundNew Zealand, Australia, Iceland, and Taiwan all saw success at controlling the first wave of the COVID-19 pandemic. As islands, they make excellent case studies for exploring the effects of international travel and human movement on the spread of COVID-19. MethodsWe employed a range of robust phylodynamic methods and genome subsampling strategies to infer the epidemiological history of SARS-CoV-2 in these four countries. We compared these results to transmission clusters identified by the New Zealand Ministry of Health by contract tracing strategies. FindingsWe estimated the effective reproduction number of COVID-19 as 1-1.4 during early stages of the pandemic, and show that it declined below 1 as human movement was restricted. We also showed that this disease was introduced many times into each country, and that introductions slowed down markedly following the reduction of international travel in mid March 2020. Finally, we confirmed that New Zealand transmission clusters identified via standard health surveillance strategies largely agree with those defined by genomic data. InterpretationWe have demonstrated how the use of genomic data and computational biology methods can assist health officials in characterising the epidemiology of viral epidemics, and for contact tracing. FundingThis research was funded by the Health Research Council of New Zealand, the Ministry of Business, Innovation, and Employment, the Royal Society of New Zealand, and the New Zealand Ministry of Health. Research in ContextO_ST_ABSEvidence before this studyC_ST_ABSOur study looks at the early months of the COVID-19 pandemic, a period in which the first wave was controlled in four "island" nations - New Zealand, Australia, Taiwan, and Iceland. All prior data used in this study was collected from late 2019 until the end of April 2020. This includes over 3000 SARS-CoV-2 genomic sequences which were collected in this period (and subsequently deposited into GISAID), as well as arrival and departure information (provided by official statistics from each country), human mobility data collected from mobile phones (by Apple), and COVID-19 case data (released by the World Health Organisation). Even early on during the COVID-19 pandemic, the properties of SARS-CoV-2 - including the reproduction number and mutation rate - were well characterised, and a range of these estimates have been covered in our article. Our Bayesian phylodynamic models, including their prior distributions, are informed by all of the above sources of information. Finally, we have incorporated all of the available information on COVID-19 transmission clusters identified by the New Zealand Ministry of Health during this period. Added value of this studyWe quantified the decline in the reproduction number of SARS-CoV-2, following the decline in human mobility, in four "island" countries. We also demonstrated how importation events of SARS-CoV-2 into each considered country declined markedly following the reduction of international travel. Our results shed a different light on these patterns because of (i) our locations of choice - the four countries had success in dealing with the first pandemic wave, with their geographic isolation contributing to cleaner signals of human mobility, and (ii) our novel and empirically driven phylodynamic model, which we built from explicitly modelling mobile phone data in the four islands. Furthermore, by crossing epidemiological against ge3nomic data, our paper quantitatively assesses the ability of contact tracing, as implemented by the New Zealand Ministry of Health (NZMH), in identifying COVID-19 transmission clusters. We find evidence for a high efficacy of the specific measures taken - and when they were taken - by the NZMH in identifying transmission clusters, considered worldwide to have been successful in its response to the pandemic. Our analyses also illustrate the power of viral genomic data in assisting contact tracing. Implications of all the available evidenceThe conclusions drawn from this research inform effective policy for locations pursuing an elimination strategy. We confirm the accuracy of standard contact tracing methods at identifying clusters and show how these methods are improved using genomic data. We demonstrate how the overseas introduction rates and domestic transmission rates of an infectious viral agent can be surveilled using genomic data, and the important role each plays in overall transmission. Specifically, we have quantified these processes for four countries and have shown that they did decline significantly following declines in human travel and mobility. The phylodynamic methods used in this work is shown to be robust and applicable to a range of scenarios where appropriate subsampling is used.
