Independent report

Chapter 1: understanding the pathogen

Updated 10 January 2023

Introduction

Particularly in the early days of the pandemic, there was pressure to develop rapid evidence on SARS-CoV-2 and COVID-19. This was driven by important operational and policy questions at the outset of this public health emergency, such as:

• what were the sensible options for response, and were there public health interventions that could interrupt transmission?
• were there any therapeutics or a vaccine that could be deployed for this pathogen?
• what should the clinical response be – and what would this mean for health system response?
• how extensive did the response need to be – should measures target only cases, or all of society?
• how long would these measures be needed for?
• what strength of evidence would be needed for different responses?
• what could be communicated to the public – what was known about this pathogen and the disease it caused?

Policy decisions were for ministers to take and they involved multiple non-health as well as health-related trade-offs. However, there was a need for clinical and scientific advice on the evidence base about the pathogen and the disease it caused in order to support decision-makers. Of course, there were particular windows for policy decisions and the evidence base did not always give a definitive answer to support one option or another at the time a decision had to be taken. In such cases, there was a need to use basic epidemiological principles and be open and clear about what the evidence base did and did not say, and with what level of certainty any conclusions could be reached. The evidence base evolved throughout the course of the pandemic, and so it was important to keep an open mind and consider all feasible possibilities. It was also important to bring together a range of disciplines and types of evidence to get a fuller, more certain and more nuanced picture.

Some key scientific questions at the outset of this pandemic concerning the pathogen, the disease and its epidemiology are set out below.

The pathogen

1. What was this pathogen?

2. What information could be gathered about the pathogen that could help develop an initial diagnostic test?

3. What information about the pathogen and the disease could support targeting of appropriate repurposed and newly developed pharmaceutical interventions?

4. How could viral evolution be monitored?

The disease

5. How severe was this disease, and were there longer-term sequelae?

6. What was the duration of naturally acquired and vaccine acquired immunity, and the risk of reinfection over time?

Epidemiology

7. What were the case definitions?

8. What were the important routes of transmission?

9. What were the higher risk settings for transmission?

10. What was the proportion of asymptomatic infection and transmission, and could this maintain R over 1?

11. How long were people infectious?

In this chapter, we explore for each question how the evidence base was developed, highlighting important methods that may come into play in a future pandemic.

Our focus is on the UK’s experience. However, the science of COVID-19 is a global science and a good part of the evidence base comes from the excellent work of colleagues across the world.

Questions on the pathogen

1. What was this pathogen?

At the onset of the COVID-19 pandemic, when information on SARS-CoV-2 itself was limited, initial risk assessments and hypothesis generation for research drew upon what was already known about similar pathogens. Fortunately, identification and initial characterisation of the causative virus came swiftly. This early virological information fed into risk assessments about the nature of the virus and its risk to the population, when and whether it would be imported into the UK, as well as supporting the development of a diagnostic molecular test. It is likely that future pandemics and significant epidemics will see similarly rapid dissemination of initial information about the pathogen, particularly if they emerge and establish in countries with significant scientific capacity but, even given this, the speed of international information flow from the start of 2020 was impressive.

Early emergence and first sequences

Following the first official reports of pneumonia of unknown origin in Wuhan, China, at the end of December 2019, very early information about the pathogen came from China and other countries that experienced early imported cases. Within days, the causative pathogen was identified as a beta coronavirus, and was subsequently named as SARS-CoV-2. Chinese scientists rapidly performed laboratory-based characterisation (virus culture, electron microscopy) and sequencing (unbiased meta-genomic techniques) of the pathogen from clinical samples.^{[footnote 1]}, ^{[footnote 2]} The first genomic sequence was generated on 3 January 2020, and publicly released on 10 January 2020. Within weeks, the virus receptor was identified as ACE2, with TMPRSS2 also flagged as important for viral entry.

Early on, phylogenetic analysis of available genomes and epidemiological studies of early cases gave signals that the virus had recently emerged, and consideration was given to the possible origin.^{[footnote 3]}, ^{[footnote 4]}

Local expertise and access to high-end technology in China enabled rapid identification and characterisation of SARS-CoV-2. Nonetheless, detection of a newly emerged pathogen could take longer if, for example, presence of genomic material was short-lived or difficult to detect, the pathogen was difficult to culture in the laboratory, or if the outbreak had arisen in a region with more limited diagnostic capacity. In the earliest stages, knowledge and expert opinion was reliant on accessible international data. Channels to access this rapidly such as the Global Initiative on Sharing Avian Influenza Data (GISAID) were key.^{[footnote 5]}

Using existing knowledge from similar pathogens

Comparison of genome sequences with other known human pathogens demonstrated that SARS-CoV-1 was the closest related human pathogen, with around 80% genomic similarity to SARS-CoV-2. It was known that SARS-CoV-1 caused severe human infections and used the same ACE2 receptor. Other related human pathogens were also drawn upon for scientific insight, including:

• MERS-CoV, which showed around 50% genomic similarity but did not use ACE2
• NL63, an endemic coronavirus that used ACE2
• other endemic coronaviruses: OC43, 229E and HKU1
• influenza, as a pandemic respiratory virus

As data about SARS-CoV-2 accumulated with time, it became apparent that SARS-CoV-2 was different from SARS-CoV-1 in several aspects, such as in its pre-symptomatic infectiousness, levels of asymptomatic or subclinical infections, and routes of transmission.

In the early stages of the pandemic, before robust data on SARS-CoV-2 itself became available, prior experience and knowledge about these related pathogens guided early understanding and public health actions – for example:

• facilitating prioritisation of potential therapeutics that had already shown in vitro or clinical activity against human and zoonotic coronaviruses
• signalling the potential for reinfections due to prior observations of waning immunity to seasonal coronaviruses

Prior knowledge also fed into early estimates of the incubation period, which was known to be longer for coronaviruses than influenza. Reviewing existing data on the environmental persistence of coronaviruses informed early policy on decontamination.^{[footnote 6]}

In characterising the pathogen from early clinical material, relationships between public health agencies and laboratory networks were key in prioritising distribution of virus isolate (to those with established biocontainment facilities) and planning further investigations. Academic laboratories with technical expertise collaborated with those running approved biocontainment facilities in other organisations to set up and lead work on virus characterisation, such as sequencing, in vitro studies and animal models. This supported assay development and furthered our knowledge of the virus. Clinical studies, in particular use of established protocols via the UK’s International Severe Acute Respiratory Infection Consortium (ISARIC) Clinical Characterisation Protocol and, later, human challenge studies, also delivered important data about the virus and the disease it caused.^{[footnote 7]}, ^{[footnote 8]}

As the virus reached the UK, early recognition and detection of cases was important in supporting further research into SARS-CoV-2. After the first case was detected in the UK in late January 2020, the virus was cultured and sequenced within days and shared with academic partners, enabling early virological work and feeding into wider research to develop our understanding of the pathogen. This wider research, including into potential pharmaceutical interventions, the duration of protective immunity to this pathogen and likelihood of reinfection, and the nature of severe and long-term disease, is set out in the following sections.

2. What information could be gathered about the pathogen that could help develop an initial diagnostic test?

Testing to identify cases had multiple applications throughout this pandemic, supporting clinical management, infection prevention and control (especially in health and care settings), contact tracing, surveillance, and to understand transmission force, transmission routes and severe disease rates. Testing was especially important because the symptoms of COVID-19 were often non-specific, minimal or absent. It was therefore an early priority – in the UK and globally – to develop diagnostic tests for the SARS-CoV-2 virus. This is likely to be the case for future pandemics and major epidemics.

The early diagnostic test (as is the case for many viruses) was molecular (reverse transcription polymerase chain reaction, or RT-PCR), though development of serological assays was also a major strand from an early stage, and later commercially developed antigen tests were also deployed at scale (for further detail on test technologies see Chapter 6: testing). Had this been a virus whose genetic material (DNA or RNA) is only briefly detectable (such as dengue virus), serology may have played a greater role for diagnostic purposes. In this pandemic, and in contrast to, for example, HIV, serology was primarily used to monitor seroprevalence and support research (such as understanding rates of asymptomatic infection, the risks of reinfection and vaccine efficacy). Self-performed viral antigen-based tests were implemented for widespread community-based asymptomatic testing and, in the later stages of the pandemic, as a signal for infectiousness to guide isolation timelines.

There was a need for multiple modes and types of testing. The speed of initial development of several different test modalities in this pandemic was impressive, with scale-up being more rate limiting. Scale-up was also hampered by the lack of a significant diagnostics industry capability in the UK (again, this is covered in more detail in Chapter 6: testing).

Evidence informing molecular testing

Sequencing

Typically, with current methods, the development of specific molecular diagnostics for any new emerging viral pathogen requires knowledge of the virus genomic sequence. Once the target sequence is known, sensitivity and specificity of PCR-based or nucleic acid amplification test (NAAT)-based diagnostics is typically greater than 95% and 99%, respectively; this will likely change and improve by the time of the next pandemic. Very early in this pandemic Chinese scientists performed genomic sequencing of SARS-CoV-2 and shared the full sequence globally via a public database.^{[footnote 1]} It was important to have the entire viral sequence for SARS-CoV-2 because different regions of the viral genome could be used for different purposes for diagnostic detection. Within each virus family for RNA and DNA viruses, there tend to be regions of the viral genome which are highly conserved, usually containing family-specific sequences. Such regions of the viral genome have been used to develop family specific diagnostics – for example, pan-coronavirus, influenza A, or herpes virus diagnostics.

Whole genome sequencing also enabled identification of genetic similarity with other coronaviruses, particularly SARS-CoV-1, for which diagnostic expertise and clinical materials existed in several public health laboratories across the world, including the UK. This facilitated rapid development of a diagnostic assay through international collaboration between public health laboratories. SARS-CoV-1 clinical samples were used as control material during the early development of an RT-PCR assay.^{[footnote 9]}

Ongoing sequencing surveillance was important for testing, throughout the pandemic, to highlight mutations within primer sites that could affect test performance.^{[footnote 10]} Multi-target PCR assays helped to reduce this risk – as, for example, when S gene target dropout was observed with the Alpha and Omicron SARS-CoV-2 variants.^{[footnote 11]} Links with industry for rapid development, distribution and validation of laboratory standards to support the monitoring of test performance were essential.

Sampling

Serial clinical sampling from multiple anatomical sites (respiratory and non-respiratory samples) from the first 10 to 20 UK cases that were contained in high consequence infectious disease (HCID) units provided valuable early data on viral shedding.^{[footnote 12]} By mid February 2020 there was growing clarity on which sites the virus was shed from and when, based on sequential sampling studies from small cohorts and case reports.^{[footnote 13]}, ^{[footnote 14]}, ^{[footnote 15]} Clinical data and case series began to show that nose and throat swabs were reasonable samples for detection of the virus, and that although faecal shedding occurred there was limited evidence for viraemia. ^{[footnote 11]} It is worth noting that there were initial difficulties moving samples around due to their HCID classification, and this is relevant for future pandemics that will likely require rapid moving and investigation into such samples. HCID classification should not be extended beyond the period it is required.

Understanding the kinetics of viral infection in the upper respiratory tract during an acute infection helped to inform the interpretation of PCR test results – in other words, what positive, negative and ‘positive at the limit of detection’ PCR results imply in terms of infectiousness, at different stages of infection. It was noted that low-level PCR positivity can remain for some time after an acute infection, without infectiousness. Therefore, for example, a single low positive (high cycle threshold value) PCR test result could indicate early infection when the individual is about to become highly infectious, late infection with lower infectiousness, or inadequate sample quality – understanding this nuance was important in interpreting test results for infection control or public health actions.^{[footnote 16]}

As the pandemic progressed and testing was scaled up, the value of easy-to-perform sampling, particularly that which can be performed by the patient themselves at the point of care, became increasingly important. In this pandemic, saliva samples for PCR-based diagnosis, different upper respiratory tract swabbing locations (anterior nares versus nasopharyngeal sampling), oral fluid or dried blood spots versus venous blood sampling for serology were all explored. Longitudinal and cross-sectional sampling studies, collecting novel sample types alongside existing validated sample types, enabled validation of diagnostics.^{[footnote 17]}, ^{[footnote 18]}

With the ongoing evolution of SARS-CoV-2 and emergence of variants, it has been necessary to repeat and review virological sampling studies to monitor any impact on test performance as pathogen biology changes. We anticipate this will be needed in future epidemics and pandemics.

Virus culture

In general, virus culture work was constrained by requirement for Biosafety Level 3 containment facilities and technical expertise. Distribution of the first live virus isolates required appropriate safety licensing in place at receiving research laboratories, which is a potential rate-limiting step in the event of a pandemic. Virus isolation from clinical material taken from one of the first UK clinical cases had occurred by early February 2020 – this was needed to generate RT-PCR assay control material for diagnostic laboratories. Of note, had the UK not experienced a clinical case of COVID-19 for some time, this material would have needed to be sourced promptly from an international partner to prevent delays to diagnostic test development and rollout. The same was true for testing new variants where appropriate samples were not available in the UK to use for neutralisation studies. Throughout the pandemic, virus culture, performed ad hoc on clinical samples from cohort studies, provided valuable information about infectiousness timelines (see section 11: How long were people infectious?), which in turn aided interpretation of diagnostic tests results for infection control and public health purposes.^{[footnote 19]}

Evidence informing serological testing

Serological assay development, deployment and interpretation was supported by an understanding of when and which antibodies (IgA/IgM/IgG) develop after infection, to which pathogen antigen (such as SARS-CoV-2 spike, nuclear protein), at which anatomical sites, and for how long. Of course, all serological tests signalled some type of immune response to the virus – however, they had differential sensitivity and specificity depending on the assay and target. It was important to understand potential cross-reactivity with other coronaviruses as well as any differences in the magnitude of the serological response depending on the severity of illness (asymptomatic, mild, severe) or demographics (such as age).

Evaluating test performance requires access to well-characterised positive and negative serum samples. In this pandemic, paired serology was actively collected from persons with suspected COVID-19 in the first months of the pandemic who tested RT-PCR negative. This, in addition to existing banked serum and residual serum from NHS diagnostic laboratories, contributed vital assay control material.^{[footnote 20]} Longitudinal serological sampling studies such as SARS-CoV2 immunity and reinfection evaluation (SIREN) and Enhanced Seroprevalence for COVID-19 Antibodies (ESCAPE) also provided valuable clinical material to help validate assays in development. For example, the ESCAPE study collected oral fluid at the same time as serum to facilitate validation of this sample type. These studies also furthered our understanding of the kinetics of the immune response to infection (such as when people develop detectable antibodies) and of the duration of protective immunity (such as how long antibodies are able to protect us from a further infection). It was then possible, in close collaboration with academic partners, to develop and validate assays for detection of neutralising antibodies, particularly surrogate assays not requiring containment level 3, and to understand their correlation with commercially available serological tests such as enzyme-linked immunosorbent assay (ELISA) tests. Later in the pandemic, having internationally recognised serological standards enabled better comparison between vaccine clinical trials, and specific serological testing was used to differentiate natural from vaccine-derived immunity.

3. What information about the pathogen and the disease could support targeting of appropriate repurposed and newly developed pharmaceutical interventions?

Pharmaceutical interventions (PIs) were an early priority as a means to reduce morbidity and mortality – both directly due to COVID-19 disease and indirectly from healthcare disruption due to high numbers of severe cases.

This section sets out the information required on the pathogen and the host response to guide and support PI development in this pandemic; how this evidence was generated; advice based on this experience. This is set broadly into 2 sections covering the different types of interventions: vaccines and therapeutic agents, including disease modifying host directed therapeutics and antiviral therapeutics. The process of developing and deploying PIs is covered in more detail in Chapter 9.

Early research focused on viral pathophysiology, host susceptibilities and disease course in order to:

• identify targets for preventative, disease modifying and antiviral therapeutics
• shortlist repurposed pharmaceutical candidates
• focus research and development of novel options

Vaccines

The development of vaccines against SARS-CoV-2 was informed by the host immune response to the virus following natural infection and required information on the antigenic target of antibodies that neutralised virus entry into cells. Review of existing data on related human coronavirus structure and host cell binding and vaccine studies for SARS-CoV-1 and MERS-CoV identified the SARS-CoV-2 spike protein as a primary antigenic target for vaccine development and suggested the likely success of vaccines targeting this region of the virus in the first months of the pandemic.^{[footnote 21]}, ^{[footnote 22]}, ^{[footnote 23]}, ^{[footnote 24]}, ^{[footnote 25]}, ^{[footnote 26]} Prior knowledge of the mutation rates and duration of immune responses to highly related human coronaviruses also helped to predict the need for repeated vaccinations and regular adaptation of vaccine content.^{[footnote 27]}

By April 2020, spike glycoprotein sequence analysis and structural analysis using cryogenic electron microscopy had confirmed ACE2 as the human host cell receptor.^{[footnote 28]}, ^{[footnote 29]}, ^{[footnote 30]} Laboratory studies from early clinical samples enabled a better understanding of the viral lifecycle and identification of the interaction between the SARS-CoV-2 spike protein receptor binding domain (RBD) and ACE2.

The rapid development and validation of neutralisation assays provided methodology for assessing the development of antibodies that could neutralise viral entry into cells and were used to show that antibodies targeting the SARS-CoV-2 spike protein neutralise the virus. This corroborated the use of the spike protein as a target for vaccine development and identified anti-viral monoclonal antibodies with potential for therapeutic use.^{[footnote 31]}

Neutralisation assays were also used to monitor the immune response following natural infection to examine correlates of protection and duration of immunity, informing protocols for vaccine trials and the need for booster doses.^{[footnote 32]}, ^{[footnote 33]}, ^{[footnote 34]}

All of these processes required rapid access to viral specimens to analyse the genetic sequence of the virus and obtain live virus isolates, clinical characterisation of patients with different disease severity, and blood samples to assess antibodies from convalescent patients. This necessitated the early set-up of cohort studies from the outset of the pandemic, with sequential sampling from people across the spectrum of disease, a process that was undertaken first in China and then rapidly across the globe as the pandemic spread.^{[footnote 35]}

A variety of samples (serum, whole blood, peripheral blood mononuclear cell (PBMC), oral fluid) from affected individuals during acute and convalescent phases were obtained. The processing of these samples can be more involved in terms of time and materials than standard diagnostic samples, but these samples were key to understanding the nature and duration of pathogen-specific immune memory and for the identification of further vaccine targets.

Knowledge of the high mutation rate of other human coronaviruses highlighted the need for vigilant monitoring of the genetic evolution of the virus, which was facilitated through the set-up of the COVID-19 Genomics Consortium. This identified new viral variants and guided hypotheses regarding the likely generation of resistance to vaccines and antiviral agents, as well as likelihood of reinfection due to evasion of host immunity.^{[footnote 36]}

By early summer 2020, concurring with earlier studies from China, a cohort study using samples collected from the first infected people in the UK showed that the majority of individuals mounted a detectable antibody response, including neutralising antibodies, following laboratory confirmed infection. This suggested that individuals were likely to respond to vaccination with a protective immune response. The same study also found a higher neutralising antibody associated with more severe disease and highlighted the potential for convalescent plasma as a therapeutic intervention.^{[footnote 37]}

Animal models were an important route to testing hypotheses and delivered early signals on likely host responses to vaccination. In August 2020 non-human primate models indicated protection from re-infection following a primary infection with SARS-CoV-2 or passive immunisation with SARS-CoV-2 specific monoclonal antibodies, supporting the postulation of the likely success of future vaccination programmes in protective immunity, at least in the short term.^{[footnote 38]}, ^{[footnote 39]}, ^{[footnote 40]} Simultaneously, SARS-CoV-2 virus-specific B cells were found to be detectable by flow cytometry following mild and severe infection, and for several months following infection, irrespective of waning neutralising antibody titres.^{[footnote 41]}, ^{[footnote 42]} This demonstrated the presence of a pool of antigen specific immune memory cells primed to respond on re-exposure.

By September 2020, understanding of immune differences between those with mild and severe disease further expanded with T cell enzyme-linked immunosorbent spot (ELISPOT) on peripheral blood mononuclear cells using synthetic peptides of SARS-CoV-2, finding functional CD4+ and CD8+ memory T cell responses in COVID-19 survivors.^{[footnote 43]} The presence of responses to multiple viral epitopes, including those outside the key spike region of the virus, highlighted novel vaccine targets with the potential to be less susceptible to viral escape mutations within the spike region.

Early detection of neutralising antibodies from patients recovered from SARS-CoV-2 infection were important in the development of monoclonal antibodies blocking the interaction between SARS-CoV-2 and the host cell receptor.^{[footnote 31]} Along the same principles the potential utility of convalescent plasma therapy was considered in the early stages of the pandemic based on historical use in SARS-CoV-1, influenza and other respiratory viral infections.^{[footnote 44]}, ^{[footnote 45]}, ^{[footnote 46]} This required sampling from known positive cases and substantial operational input and coordination between public health and blood transfusion services to obtain donations for analysis and therapeutic use. Evaluation in multi-site platform trials subsequently demonstrated this not to have a survival benefit in hospitalised patients, the reasons for which remain unclear, but convalescent plasma may be a useful option to consider in the absence of other therapeutics early in the course of a newly discovered infectious agent.^{[footnote 47]}

In this pandemic, vaccine development was focused on the S protein which was the most obvious and most defined antigenic target. Targeting a wider selection of target proteins, such as the N protein, could potentially be helpful. These targets are less well defined but could be more conserved and offer more durable protection particularly given the possibility of vaccine-escaping new variants. The first targeted antigen for a pandemic organism may not ultimately be the best, so there may need to a broader scientific lens and incentives and support for industry to explore other protein targets. This is important to keep in mind whatever the pathogen.

Therapeutic agents

Therapeutic agents were required for different purposes in different scenarios: in intensive care (ICU) settings the primary aim was to reduce mortality; in hospitalised patients outside of ICU the goal was to reduce escalation to ICU or requirement for oxygen therapy; in the community the aim was preventing hospital admission by treating high risk individuals early or targeting prophylaxis at high-risk individuals who had been exposed. For post-exposure prophylaxis in the community (both vaccines and therapeutics), early studies on the secondary attack rate were helpful in clarifying the incubation period.

Potential therapeutic agents included those acting directly against the virus, immunomodulatory agents directed against the host immune response to infection, and therapeutics directed against other organ system effects of the infection. At the very outset of the pandemic, hypothesis generation and identification of candidate therapeutics for trials relied on existing knowledge of similar pathogens. Knowledge of other human coronaviruses, including SARS-CoV-1 and MERS-CoV, enabled a rapid assessment of potentially viable therapeutic agents, both direct acting antivirals and immunomodulatory agents. In vitro studies, animal models and human safety data were key in generating early candidates for clinical trials – though caution and expert input were essential when interpreting such evidence in light of SARS-CoV-2.

Virus-directed agents

Initial assessments suggested 2 antiviral candidates to begin trials:

• combination lopinavir/ritonavir, protease inhibitors with activity shown in limited experience with SARS-CoV-1 and in non-human primate models of MERS-CoV
• remdesivir, a nucleoside analogue with activity against MERS-CoV^{[footnote 48]}, ^{[footnote 49]}

This is covered in Chapter 9: pharmaceutical interventions.

Host-directed therapeutics

Initial selection of host-directed countermeasures for evaluation (such as immunomodulators or anti-thrombotics) depended upon careful clinical characterisation of mild, moderate and severe cases and the mechanisms of pathogenesis, as the efficacy (and safety) of host-directed therapies can depend on the stage and severity of disease. Large-scale cohort studies provided information on the contribution of immune-mediated disease to the pathogenesis of infection and rates of complications, such as thrombosis. They delivered results fast in this pandemic. By March 2020, multi-centre cohort trials with sequential sampling from individuals across the spectrum of disease severity measuring a range of markers highlighted the role of inflammation in pathogenesis of SARS-CoV-2 infection and identified interleukin 6 (IL-6), in particular, as a potential therapeutic target.^{[footnote 50]}, ^{[footnote 51]} These findings resulted in the inclusion of steroids, tocilizumab, a monoclonal antibody targeting the IL-6 molecule, and sarilumab, a monoclonal antibody inhibiting the IL-6 molecule receptor, in clinical trials in April 2020. This is covered in more detail in Chapter 9: pharmaceutical interventions.

4. How could viral evolution be monitored?

Although whole genome sequencing of viruses during epidemics (for example, during the Ebola outbreak of 2014 and the H1N1 influenza 2009 pandemic) has been employed over the past decade, the SARS-CoV-2 pandemic marked a turning point with many countries, particularly the UK, investing substantially in sequencing large numbers of genomes.^{[footnote 52]} This allowed for fine epidemiological tracking, to understand the introduction of virus and variants into the UK, and rapid detection of novel variants. However, it is important to note that large scale sequencing on its own was not sufficient to understand variant emergence, nor to make meaningful risk assessments to inform policy responses, until it was later coupled with phenotypic analyses including antigenic studies and epidemiologic analyses of clinical severity. It also required robust, large scale epidemiological sampling.

Wastewater sampling helped signal human circulation of SARS-CoV-2 variants of concern and supported tracking lineages of SARS-CoV-2. It could have a potential role in future pandemics, but in this pandemic in the UK there were a number of important caveats to its use, such as the potential to detect viral fragments from past, resolved infections. These are covered in more detail in Chapter 4: situational awareness, analysis and assessment.

Wild type

Large scale sequencing revealed that SARS-CoV-2 arrived in the UK by hundreds of separate introductions carried by travellers returning in large part from Europe after the half-term holidays in February.^{[footnote 53]} This first wave was largely clonal, with the single exception of the early emergence, and rapid worldwide dominance, of the B.1 lineage.^{[footnote 54]} B.1 contained 4 mutations including the D614G substitution in the spike gene. It was not clear until several months later, when detailed phenotypic analyses were performed, that this was something other than a founder effect – in other words, a predominance of a lineage without a clear fitness advantage, largely due to early import and stochastic growth. Subsequent phenotypic work showed the single mutation D614G worked by exposing the part of the spike protein that bound to the ACE2 receptor, and thus increased infectivity.^{[footnote 55]} This analysis was only possible due to a combination of the development of large scale sequencing, which was at the time rapidly scaled up, including by the nascent COVID-19 Genomics UK consortium (COG-UK), coupled with phenotypic characterisation by multidisciplinary collaborators.

Alpha

For several months after summer 2020 there was again relative stasis in SARS-CoV-2 evolution within the UK, with only a few minor and fairly inconsequential mutant lineages emerging. Again, many were carried to UK from mainland Europe by travellers.^{[footnote 56]} Towards late 2020, however, rising case rates in the south-east of the UK were investigated and found to correlate with a negative result for the S gene target, one of the commonly used probe sets for quantitative polymerase chain reaction (qPCR) tests. This variant was later labelled the ‘Alpha’ variant and was relatively easy and fast to track using S gene target failure in qPCR testing.^{[footnote 57]} This underscored the importance of using several different PCR targets in combination for large scale testing of an RNA virus; had this not been done, Alpha infections would have gone undetected until later in the wave. Alpha drove a large wave of cases in the winter of 2020 to 2021, and genome sequencing revealed a constellation of mutations throughout its genome.^{[footnote 11]} Alpha was revealed through later phenotypic testing to have increased transmissibility conferred by changes in receptor binding and also changes in innate immune control.^{[footnote 58]}, ^{[footnote 59]} With the emergence of Alpha (and, shortly after, Beta detected in Southern Africa), effort was expanded to sequence and rapidly identify and characterise any further variants arising.^{[footnote 60]}

Delta

By spring 2021 signals were seen in India of potential new variants, with a surge in cases reported. These variants were later classified as Delta and Kappa. In the UK, cases of Delta and Kappa were initially predominantly in those travelling from India (see Chapter 8: NPIs, for further epidemiological context on travel restrictions).^{[footnote 61]} Initially, Kappa was assessed to be the larger threat as imports into the UK consisted mostly of Kappa, which contained a mutation at spike position 484 (484Q) that was flagged as a likely antigenic escape mutant due to its similarity to E484K (found in Beta and Gamma). However, Delta began to exhibit a more rapid growth rate and went on to dominate globally in 2021. This was occurring at the same time as the UK was rapidly vaccinating its population and gradually lifting NPIs. Laboratory studies showed that Delta was intrinsically more transmissible than previous variants.^{[footnote 62]} It also showed some modest immune escape properties, potentially allowing it to break through immunity granted by vaccination or prior infection from wild type SARS-CoV-2 with greater efficiency than Alpha.^{[footnote 63]}

Omicron

By November 2021 many countries worldwide, including the UK, were reaching their highest rates of sequencing. Sequencing from Southern Africa and travel-related sequencing from Hong Kong allowed the rapid identification of a novel variant of concern, Omicron, as soon as the first 4 sequences had been uploaded by Southern African researchers to the online sequence database GISAID.^{[footnote 64]} Omicron was characterised by a very large number of mutations, including 35 across the spike gene, many at known antigenic epitopes. The large antigenic distance between Omicron and the wild type spike protein, combined with antibody waning, resulted in poor neutralisation of Omicron by sera from vaccines – and this necessitated rapid implementation of vaccine booster programmes to counter immunological waning associated with the establishment of this variant.^{[footnote 65]}

Discussion

The origin of variants remains an open question. However, immunocompromised hosts have been a hypothetical population for variant emergence prior to the pandemic, and a similar route was implicated in this pandemic by the fact that Alpha and Omicron were phylogenetically similar to much older sequences that circulated 6 to 18 months before their emergence.^{[footnote 66]}, ^{[footnote 67]}

Whole genome sequencing has been a huge boon to the UK in the pandemic and was probably world-leading in terms of genomic epidemiology, identification of novel variants and understanding the evolution of viruses in real time. This has been a mix of both population-wide surveillance, allowing for high quality epidemiological resolution of new variants, as well as surveillance targeted to hospital populations allowing rapid detection of imported variants or chronic infections in hospitalised patients. It was extremely fortuitous (and very unlikely to be repeated in a future pandemic) that one of the main qPCR toolsets bound to a region of the SARS-CoV-2 genome was not present for some variants, allowing for rapid detection of certain potential variants. Although the UK deployed qPCR-based targeted genotyping sparingly, this could be very important for future pandemics where a rapid detection method like S gene target failure is unlikely to occur. Furthermore, the UK (like many other countries) invested in associated phenotypic characterisation of variants, allowing rapid risk assessment of emerging variants to feed into public health policy. It has been important to bring together multidisciplinary groups of public health academics including epidemiologists, genomics scientists, bioinformaticians and virologists together to rapidly assess new variants.

Table 1: summary of key SARS-CoV-2 variants and their emergence, 2020 to 2021

Event	Timeline	Description
D614G becomes predominant	Spring 2020	Genomic signal, confirmatory studies
Alpha first found	November 2020	Epidemiological signal from Kent, genotyping signal, genomic signal
Delta takes over	April 2021	Genomics signal from UK and India, travel-related signals
Omicron first found	November 2021	Genomics signal from South Africa, rapid global response

Questions on the disease

5. How severe was this disease, and were there longer-term sequelae?

Gauging the potential impact of COVID-19, and the appropriate response to take, heavily relied on understanding both the severity of acute disease and its possible longer-term sequelae.^{[footnote 68]} The degree of severity and its underlying causes will be central to the management of any future pandemic or epidemic. This section sets out evidence evolved on mortality and morbidity, both acute and chronic, for COVID-19.

Mortality

Mortality rates were difficult to define in the initial stages of this pandemic, as was the case for H1N1 influenza and SARS-CoV-1 – but for slightly different reasons. For SARS-CoV-1 in 2003, initial case fatality rate (CFR) figures underestimated severity due to early estimates missing delayed deaths – though statistical methods were developed to provide a more robust estimate of severity in similar situations which were useful in this pandemic.^{[footnote 69]} For H1N1 influenza in 2009, initial CFR estimates were about 500 times higher than the later agreed infection fatality rate (IFR) of 0.001% to 0.002% due to initially measuring only symptomatic or confirmed cases and missing milder and asymptomatic ones.^{[footnote 70]}, ^{[footnote 71]} Later, more accurate estimations of the IFR for H1N1 influenza arose from studies on outbreaks, such as one in a school in New York which included milder cases – though with important caveats on the demographic representativeness of those within specific settings like schools.^{[footnote 72]}

For SARS-CoV-2, too, there were varying estimates of CFRs in the early stages. In the UK, before widespread surveillance was set up, initial estimates of the CFR came from dividing numbers of reported deaths by the estimated number of cases in Wuhan, China at a given time.^{[footnote 73]} These estimates were greatly improved by Chinese Centres for Disease Control (CCDC) data: in mid February 2020, for example, the CCDC weekly bulletin provided a CFR estimate of 2.3% from 72,314 cases identified using either PCR testing (63%) or clinical diagnosis (37%).^{[footnote 74]} Of this group 1.3% were thought asymptomatic. Of the PCR confirmed cases, 81% were classified as mild (which included non-pneumonia or mild pneumonia) and 19% were described as severe or worse (which was classified as dyspnoea, low oxygen saturations and/or greater than 50% lung infiltrates on imaging). The CFR for those with severe disease was high at 49% and increased substantially with age (though the age distribution of this cohort was relatively young compared to the UK, with 68.8% of patients under 60). Another early study incorporated a wider range of cases from PCR testing for international travellers arriving to China, alongside cases and deaths in Wuhan, and reported a CFR of 1.4% for symptomatic COVID-19 cases.^{[footnote 75]} It was initially difficult to interpret such studies for a UK context, in part because denominators and numerators varied and in part because their source populations differed from the UK in several important ways (such as age distribution).

Population-wide surveillance (positive tests, syndromic surveillance) linked to outcomes (hospitalisation, deaths) provided high quality data for the routine calculation of CFRs in particular by providing a robust denominator. In the UK this was initially done using serology, which was difficult to interpret due to waning antibody levels, and after late spring 2020 by large scale surveillance studies such as the Office for National Statistics (ONS) COVID-19 Infection Survey (CIS), Real-time Assessment of Community Transmission (REACT) and Early Assessment of Vaccine and anti-viral Effectiveness 2 (EAVE-2), and in cohorts such as SIREN (healthcare workers) and Vivaldi (care homes). The calculation of an accurate IFR required serological testing of a representative random sample of the population, and establishing a regular serological survey allowed us to estimate the severity of disease on a regular basis. However, this took time to set up and for results to indicate severity more clearly and CFR was available much more quickly. Early establishment of data storage and linkage systems was important for the timely calculation of these statistics. Securely sharing data with academic groups facilitated rapid analysis.

Investigations of large outbreaks of COVID-19, similar to previous experience with H1N1 influenza, also supported CFR and IFR estimates early on, as well as giving signals on the proportion of asymptomatic infections. An outbreak on the cruise ship Diamond Princess in February 2020 provided early data on outcomes for 3,711 passengers and crew, and gave a CFR of 2.6% and an IFR of 1.3%, likely due to testing across the ship picking up asymptomatic cases.^{[footnote 76]}, ^{[footnote 77]} Studies of Wuhan residents outlining the likely delay distribution between onset and death were critical in estimating both CFRs and, as testing and surveillance expanded, IFRs.^{[footnote 78]} Other opportunities for screening were passengers on flights from affected areas. However, these figures needed to be interpreted in context, and could not readily be applied to different population groups with different demographic characteristics.

It was not until late spring 2020, when many countries were experiencing high transmission and testing was being ramped up alongside surveillance studies, that a shift from CFR to IFR occurred and estimates converged towards an overall IFR of around 1%.

The presence of asymptomatic cases and asymptomatic transmission for COVID-19 was particularly problematic in early mortality rate estimates, and this had not been the case for the closely related SARS-CoV-1 (for which peak infectiousness matched peak clinical symptoms). Many early studies missed asymptomatic cases in the absence of widespread testing and community surveillance, and in the UK in February to April 2020 a number of cases due to COVID-19 occurred in the community without confirmatory testing. This was likely the reason behind higher early CFR estimates: collated data in England from 31 January to 22 April 2020, for example, recorded 99,137 cases with 16,271 deaths, a crude mortality ratio of 16.4%.^{[footnote 79]} Around the same time, adjusting for age and using serological data alongside case data gave an IFR of 1.6% for the UK.^{[footnote 80]}

As noted above, global comparisons proved difficult as hospitalisation criteria, testing availability and case definitions varied over time and across different health jurisdictions. Mortality itself also varied significantly from country to country, likely due to different age structures of populations as well as differences in a range of other risk factors such as obesity, levels of social deprivation and important comorbidities (see Chapter 2: disparities). A study in Italy, where 37.6% of cases were aged 70 years or older, gave an estimated CFR of 7.3% up to 15 March 2020, compared to a much lower CFR in a Chinese study where just 11.9% of cases were over 70.^{[footnote 81]} Understanding of how these complex and interacting demographic factors influenced severe disease evolved throughout the pandemic and underscored the importance of continual evaluation of variation in severity. Heterogeneity of infection risk and disease severity is covered in more detail in Chapter 2: disparities.

Obesity was also an important driver of mortality rates. A large study of over 13,000 hospital admissions in England found a J-shaped relationship between BMI and death from COVID-19, with a nadir at 23 kg/m², and a linear rise with BMI values higher than this.^{[footnote 82]} A BMI of 40 was associated with about a 2-fold increased risk of death. Geography, level of social deprivation and the presence of co-morbidities, often linked to ethnicity, played an important part in understanding rates of severe COVID-19 and disease outcomes overall.^{[footnote 83]}, ^{[footnote 84]} Gender, too, has been flagged as a risk factor for mortality: in the working-age population, COVID-19 death rates were consistently and markedly higher for men than women throughout the pandemic.^{[footnote 85]} Early reports during the pandemic were often not able to link and adjust for all relevant variables. This is covered in more detail in Chapter 2: disparities.

In light of these differences, changes in all-cause mortality across different countries was a helpful indicator as it was not sensitive to differences in diagnostic or testing data and encompassed both direct and indirect mortality impacts from the pandemic.^{[footnote 86]} Nevertheless, geographical comparisons even with all-cause mortality needed to be handled very carefully. Future developments in infectious disease modelling may allow more precise determinations of severity earlier in a pandemic.

Morbidity

Mortality was not the only measure of severity; admissions to hospital and ICU with COVID-19 were also important metrics in this pandemic – particularly to help plan healthcare delivery. Understanding delays between infection and severe disease was also crucial in estimating the correct denominator and likely rates of severe disease at any given point. For COVID-19, the mean delay from infection to death was around 4 weeks but with wide variation.

Initial clinician impressions from the first cases can give early signals but can be misleading. Many of the early patients seen in the UK with COVID-19 were returning travellers from Europe, the majority of whom were young and fit patients with greater rates of mild disease than the wider population. Within about 2 weeks the disease had spread more widely in the population and hospitals were faced with large numbers of older patients with severe disease and high mortality.

As case rates rose, determining wider population levels of morbidity was complex. Although routine statistics on hospitalisations within the UK were available from early on, a need to prioritise tests during times of limited testing capacity meant that it was difficult to estimate the proportion of cases likely to require hospital admission or ICU care. Early, large-scale testing within the population is of course the best way to gauge severity more accurately, but this is not always feasible, especially when tests need to be developed, or are limited in supply and need to be prioritised to high-risk settings.

Comparisons using other nations’ case hospitalisation rates (CHRs), as noted above for CFRs and IFRs, was complicated by differing age structures and hospitalisation criteria and access. It was particularly challenging as some countries hospitalised all cases as an isolation method, while others hospitalised only those with clinical need for hospital care. An early report from Hubei province, China, found that 80% of identified cases were mild (no pneumonia or mild pneumonia) indicating that hospitalisation was unlikely to be required for the majority of cases – though its estimation of cases requiring hospitalisation was undoubtedly too high, most likely because it was restricted to symptomatic patients. Later, widespread testing enabled more accurate estimates which gave significantly lower percentages: a study in Indiana, USA, in early 2020 found an infection hospitalisation rate (IHR) of 2.3%, while a similar analysis in the UK at the end of 2020 (for the wild type strain) gave 3.5%.^{[footnote 87]}, ^{[footnote 88]}

Estimates of the demand for hospital and ICU beds were challenging. Levels of known risk factors for severe disease, such as population age profiles, were helpful in signalling potential levels of demand. Large scale surveillance, such as via ISARIC, has been important in giving early signals on risk factors.^{[footnote 89]} ICU admission criteria, and indeed the definition of ICU, varied between countries, again making international comparisons complex. There was significant variation in the number of critical care beds in different countries, and population characteristics (such as age) influenced likely need for ICU among COVID-19 cases.^{[footnote 90]} However, criteria for admission and quality of care in ICU were likely similar across comparable health systems, suggested by international comparisons of ICU mortality in early 2020 which showed broadly similar mortality rates of 35% to 40%.^{[footnote 91]} There were, of course, substantial changes in hospital fatality rates (HFRs) over the course of the pandemic: rates in the UK during the first wave had almost halved by summer 2020, but rose again during autumn 2020 and into the 2020 to 2021 Alpha wave as the new variant drove rapidly increasing case rates and hospitals came under significant pressures.

Longer-term consequences of COVID-19

By the summer of 2021, it was becoming apparent that many patients had ongoing symptoms after recovery which persisted for longer than 3 months. One prospective study of 431 individuals testing positive for COVID-19 in Switzerland, published in July 2021, found that 6 to 8 months after infection 55% of the cohort reported ongoing fatigue, 25% had some degree of breathlessness, and 26% fulfilled criteria for depression.^{[footnote 92]} Since that time, the range of chronic symptoms recorded for cases of COVID-19 has expanded greatly.^{[footnote 93]} A diagnostic definition of the condition has been made as post-COVID-19 syndrome by the National Institute for Clinical Excellence (NICE), more commonly referred to as ‘long COVID’ by sufferers and clinicians, although in reality it is likely to represent several overlapping syndromes.^{[footnote 94]}

The exact number who have experienced longer-term symptoms after COVID-19 is likely substantial but remains unclear, as does the aetiology of the syndrome, including whether it was one or (perhaps more likely) a number of different overlapping syndromes. In July 2022 the ONS CIS estimated that 1.4 million people in the UK were experiencing long COVID symptoms that adversely affected their day-to-day activities in the 4 weeks ending 4 June 2022.^{[footnote 95]}

Most children had very minimal medium and long-term health impacts from COVID-19, but rarely some children developed a multisystem inflammatory condition termed paediatric inflammatory multisystem syndrome (PIMS-TS) temporally associated with SARS-CoV-2, or multisystem inflammatory syndrome (in children) (MIS-C).^{[footnote 96]} The true incidence of PIMS-TS was unclear, as many childhood COVID-19 infections went undiagnosed. One study from the US estimated 316 cases per 10⁶ COVID-19 infections in persons under 21 years old.^{[footnote 97]} The relationship between the syndrome and COVID-19 infection was shown by about two-thirds of presentations being associated with seroconversion to SARS-CoV-2, and about one-third actually testing positive for SARS-CoV-2 on admission. In some cases, the association was suspected because of close contacts with a confirmed case but without seroconversion or positive viral PCR. Most cases presented between 2 to 4 weeks after COVID-19 infection was documented. About 70% of cases required ICU admission, though mortality was relatively low at 1.1%.^{[footnote 98]} Some children also experienced long COVID but at a much lower rate than adults.

It is important to note for future pandemic preparedness that there may be longer-term consequences of an infection affecting a large percentage of the population, and that adequate surveillance mechanisms should be in place to capture the epidemiology of the condition accurately to allow adequate planning of healthcare resources in the longer term.

Variants

Over time, new variants arose that led to different clinical outcomes. Detecting these differences was challenging, as it required linking large scale genomic data with hospitalisation and mortality rates. Greater severity was seen with one of the first variants (Alpha), although a subsequent group of variants (Omicron) was found to have had reduced hospitalisations and deaths per case, though due to higher transmissibility and therefore high case rates still resulted in large numbers of hospitalisations. ^{[footnote 99]}, ^{[footnote 100]} Changes in pathogenicity were difficult to measure and it was not possible to assume a shift towards less severe outcomes as the virus evolved. Levels of immunity (both natural and vaccine-derived) were an important confounding factor in determining the intrinsic severity of new variants, as were changing demographic factors (such as the age group predominantly infected) across different waves.

6. What was the duration of naturally acquired and vaccine acquired immunity, and the risk of reinfection over time?

Duration of immunity (natural or vaccine-derived) and risk of reinfection has varied widely in epidemic-potential infections, ranging from lifelong infections such as HIV, infections where a single infection generally confers lifelong protection such as measles, and infections where prior infection provides partial, temporary, or minimal protection from subsequent infection such as influenza and malaria. Cross-protection between different variants of a disease is also highly variable.

As a novel infection, understanding the duration of immunity and risk of reinfection over time for COVID-19 was important to enable individuals, scientists, and policymakers to determine who was protected against infection and for how long, to predict the likely duration of impact of any vaccines, and to inform epidemic modelling. Knowledge of the duration of passive immunity from antibodies was also important for understanding the potential role of antibody drugs.

This information is likely to be important in any new pandemic or major epidemic.

Throughout, there was a need to differentiate between sterilising immunity, which provides protection against both illness and infection, and non-sterilising immunity which provides some, or complete, protection against serious illness but not infection.^{[footnote 101]} Estimating protection against infection required routine systematic testing to detect infections in the presence or absence of symptoms, while symptom-based testing and data on hospitalisations or deaths supported understanding of protection against illness. There was an initial assumption, which had to be tested, that waning of immunity from severe disease would be significantly slower than waning of immunity from infection.

Initial hypotheses

Extrapolation from biologically similar or evolutionarily related pathogens provided the earliest clues to whether reinfection was likely, and after what interval.^{[footnote 102]} Immunity to SARS-CoV-1 and MERS-CoV was thought to wane over time, and there was evidence of confirmed reinfections with seasonal human coronaviruses.^{[footnote 103]}, ^{[footnote 104]}, ^{[footnote 105]}, ^{[footnote 106]}, ^{[footnote 107]}, ^{[footnote 108]} This meant that from an early stage there was an assumption that reinfections with SARS-CoV-2 were possible and it was possible to explore the impact of reinfection through mathematical models, monitor early case reports for evidence of proven reinfection and design studies to investigate reinfection rates.^{[footnote 109]} There was also a reasonable assumption that the virus would mutate over time which in turn could impact reinfection risk.

Characterisation of the immune response to infection with SARS-CoV-2 required exploration of both antibody and cell-mediated effects. However, the presence or absence of an antibody or T-cell response was insufficient to confirm protection against infection with SARS-CoV-2.^{[footnote 110]} Measurement of the duration of immunity therefore required establishment of correlates of protection which indicated the presence of an effective immune response.^{[footnote 102]}, ^{[footnote 111]}

Early data

By early 2020, data emerged indicating that the majority of individuals infected with SARS-CoV-2 displayed an antibody response between 10 to 14 days after symptom onset. ^{[footnote 102]} Data showed that in mild cases, antibodies took longer to appear or were low or undetectable during the timescale of completed studies. ^{[footnote 102]},^{[footnote 112]}, ^{[footnote 113]}, ^{[footnote 114]}, ^{[footnote 115]}, ^{[footnote 116]} Much data was gathered through observational studies with serial sampling on small numbers of participants – however, a lack of available validated assays to measure antibody or cell-mediated immunity in early 2020 hampered early attempts to characterise the immune response soon after the emergence of the pathogen. Around this time, data from animal models also signalled that the presence of antibody protected against reinfection when challenged with SARS-CoV-2.^{[footnote 117]}, ^{[footnote 118]}

Antibodies did not, however, inevitably mean protection from infection (nor did lack of antibodies preclude it due to other immunological mechanisms such as T-cell mediated immunity), so there was a need for further longitudinal studies to examine reinfection risk. The Vivaldi (care homes) and SIREN (healthcare workers) cohort studies were key to developing understanding of infection, transmission and immunity.^{[footnote 119]}, ^{[footnote 120]}, ^{[footnote 121]} These studies were initiated in the first half of 2020 and adapted to provide up-to-date information on issues as they emerged, through adjustment of protocols to include questions on vaccine effectiveness and variant characteristics.^{[footnote 122]} SIREN, for example, recruited its first participant in June 2020, investigated its first reinfection in September 2020, produced an initial reinfection analysis in December 2020, and published its first vaccine effectiveness analysis in January 2021.^{[footnote 123]}

Emerging evidence from the first wave

From early to mid 2020, evidence arose that there was variation in the antibody response produced by different individuals after infection.^{[footnote 102]}, ^{[footnote 114]}, ^{[footnote 124]} In May 2020, literature reports emerged of individuals testing positive for SARS-CoV-2 on PCR for 6 to 8 weeks, complicating the differentiation of new infections from ongoing detection.^{[footnote 55]} At this stage, the time to seroconversion and antibody dynamics over the first 3 months following infection were well-characterised for both total antibody and antibody classes.^{[footnote 125]} Mid 2020 also saw the emergence of early observational studies describing the T-cell response to SARS-CoV-2 infection, though there was little data on the T-cell response after the acute phase of infection. Robust evidence characterising the T-cell response to SARS-CoV-2 infection emerged later in the year.^{[footnote 43]}

The first published case reports of SARS-CoV-2 reinfection confirmed by whole genome sequencing also emerged in mid 2020.^{[footnote 126]} Several other reports of reinfection emerged at this time, though many did not have sufficient data to distinguish between persistent primary infection and reinfection.^{[footnote 127]}, ^{[footnote 128]}, ^{[footnote 129]}, ^{[footnote 130]} The corroboration of early reports of reinfections with SARS-CoV-2 was complicated due to restricted access to testing during the time period of primary infections. During the ‘first wave’, the great majority of infected persons did not have access to PCR testing, and viral isolates were not regularly obtained for sequencing.^{[footnote 131]} At this point, reliable information on the proportion of people likely to experience reinfection, the timeline of reinfection, and the characteristics that make reinfection more or less likely was still missing.

Accumulating evidence as time from infection increases

As time since the first infections with SARS-CoV-2 elapsed, the length of time over which the immune response was characterised increased. By the end of 2020, antibodies, in particular neutralising antibodies, were shown to be a useful correlate of protection against SARS-CoV-2, through a combination of animal studies, outbreak studies and cohort studies.^{[footnote 120]}, ^{[footnote 132]}, ^{[footnote 133]}, ^{[footnote 134]}, ^{[footnote 135]}, ^{[footnote 136]} Nevertheless, the concentration of antibody that correlated with protection was not yet established. The antibody response following natural infection was shown to persist for at least 3 to 6 months, and the cellular immune response for over 5 months, though seroprevalence studies in the UK showed a decline in the presence of antibody positivity and confirmed reports of reinfection began to emerge, suggesting a waning in protection over time.^{[footnote 41]}, ^{[footnote 128]}, ^{[footnote 129]}, ^{[footnote 137]}, ^{[footnote 138]} Evidence from longitudinal observational and cohort studies emerged to suggest that people who had experienced asymptomatic or mild SARS-CoV-2 infection could experience waning immunity over 3 to 5 months.^{[footnote 33]},^{[footnote 139]},^{[footnote 140]}

Data collection in longitudinal cohort studies included the demographic characteristics of participants, routine samples (systematic testing for the identification of the pathogen and its antibodies, with genetic sequencing of the pathogen where applicable), and routine collection of information on symptoms and exposures. Once established, these longitudinal cohort studies were cross-purpose sources of information, providing insight not only into reinfection risk, but also the duration of the protective effect of vaccination following rollout, and the prevalence and incidence of infections in defined populations. Healthcare workers were a useful target population as they were essential for the functioning of the health system, could provide insight into the effectiveness of personal protective equipment and assist in the understanding of nosocomial transmission, and facilitated the establishment of cohort studies at pace.^{[footnote 121]},^{[footnote 132]}

At this time, numerical estimates of the protective effect of baseline antibodies to SARS-CoV-2 against symptomatic reinfection, asymptomatic reinfection, or all infections combined over a period of 3 to 5 months, also became available. ^{[footnote 101]}, ^{[footnote 132]}, ^{[footnote 133]}, ^{[footnote 139]} The end of 2020 also brought the first clinical trial data demonstrating that SARS-CoV-2 vaccines could provide a high level of protection against disease – however, the duration of immunity provided remained unknown.

By mid 2021, descriptions of viral loads (as measured by cycle threshold (Ct) values) in reinfected individuals were available.^{[footnote 120]} Cultivable virus had also been isolated from reinfected individuals, demonstrating that reinfections presented a risk of onward transmission.^{[footnote 141]}, ^{[footnote 142]} Throughout the first half of 2021, understanding of the duration of the immune response to SARS-CoV-2 improved. Antibody was found to be detectable in saliva for at least 8 months following infection, and in blood for at least 9 months. The presence of antibody was shown to be associated with a protective effect against infection over at least 7 to 10 months, with a lower effect in those aged over 65.^{[footnote 110]} The cell-mediated immune response to SARS-CoV-2 was shown to be detectable up to 8 months after infection.^{[footnote 41]}, ^{[footnote 110]}, ^{[footnote 137]} Characterisation of neutralising antibody titres over time since either infection or vaccination or both (through longitudinal serological sampling) continued throughout 2022.^{[footnote 143]}, ^{[footnote 144]}

Variants

The duration of protection against infection and illness with SARS-CoV-2 was driven both by the immune response to either infection or vaccination or both, and the antigenic distance between circulating viruses.^{[footnote 145]} It was recognised that protection would not endure if the variant causing the primary infection (or against which the vaccine is directed) was replaced by a new variant that was antigenically distant from the first.^{[footnote 146]} In late 2020 and early 2021, the emergence of new SARS-CoV-2 variants which were significantly different to the Wuhan original necessitated exploration of the protection induced by natural infection and vaccines against variants that were antigenically different to the primary infection.^{[footnote 147]}, ^{[footnote 148]}

In March 2021, early evidence showed that the risk of reinfection with the Alpha variant was comparable to the risk of reinfection with the wild type, though these findings were confounded by the shorter time from primary infection in the case of the alpha variant.^{[footnote 149]}, ^{[footnote 150]} National surveillance data was used to monitor reinfections, including with newly emerging variants, and showed evidence of increased reinfections at the emergence of the delta and omicron variants.^{[footnote 151]},^{[footnote 152]},^{[footnote 153]}

Epidemiological questions

7. What were the case definitions?

Establishing case definitions is an essential step in any pandemic or major epidemic. As a new disease, the case definitions for COVID-19 evolved over time. During the SARS-CoV-2 pandemic, as with most common infectious diseases, case definitions were used for 3 differentiated but overlapping purposes:

• public health: contact tracing, outbreak investigations, and communication to the public – for example, on when to isolate
• epidemiological: surveillance
• clinical: provision of healthcare

Optimising case definitions to cover different use cases often required trade-offs, especially between sensitivity and specificity. Case definitions used epidemiological, clinical and testing criteria, but the balance of these changed over the course of the pandemic as knowledge of SARS-CoV-2 accumulated and as testing resources expanded to meet demand.

Epidemiological criteria

Initially, UK case definitions placed more emphasis on person and place (such as people who travelled from Wuhan, China) than on testing criteria – which would likely also occur in the initial stages of most future pandemics and major epidemics for which testing is limited.^{[footnote 154]}, ^{[footnote 155]} Symptoms were included but it was helpful to also include epidemiological information (such as where a person had recently been) due to non-specific symptom profiles for COVID-19 in early 2020.^{[footnote 154]}, ^{[footnote 155]}

The geographical scope of definitions widened as cases appeared in other countries until such time as it was no longer meaningful and most transmission was domestic.

Clinical criteria

The clinical criteria included in the case definition changed over time as data accumulated. For example, in spring 2020, loss of taste or smell were included in the COVID-19 case definition.^{[footnote 156]}

Robust estimates of the sensitivity and specificity of specific symptoms were not available until later in the pandemic, as much of the early evidence generated was affected by the following limitations:

1. Many studies reported only the frequency of symptoms in persons infected with SARS-CoV-2 and no comparative data on symptomatic people testing negative. This allows assessment of sensitivity but not specificity. Research should include non-infected comparator groups.^{[footnote 157]}
2. Many early symptom reports focused on people who were hospitalised, leaving it unclear whether symptoms would be similar in mild community cases.
3. Data from national testing programmes may be biased as these programmes often specify the symptoms for which they want people to test. This leads to an overestimation of the sensitivity of the symptoms described in the testing criteria.

Throughout the pandemic, there were frequent calls to include a wider range of symptoms in case definitions but there was an ongoing need to balance the need for sensitivity (increased by a broader list of symptoms) with specificity (increased by a narrower list of symptoms).^{[footnote 158]}, ^{[footnote 159]} Early in the pandemic when the infection was emerging, and the critical objective was to find as high a proportion of all cases as possible and reduce transmission through high impact public health contact tracing, the strategic aim of the case definition was high sensitivity.^{[footnote 160]}

Regular reviews of the sensitivity and specificity of specific symptoms and symptom complexes were undertaken to ensure that a reasonable balance was struck between the ability to correctly identify cases, and the ability to exclude non-cases, in a pragmatic and clinically useful way.^{[footnote 157]},^{[footnote 161]} Algorithmic approaches to case definitions, incorporating both symptoms and epidemiological data, could theoretically have been used to optimise the balance between sensitivity and specificity, but may have been challenging to implement and communicate.

When deliberating the balance between sensitivity and specificity, it was also necessary to consider the impact of changing case definitions. For example, using a highly sensitive case definition would have had a big impact on testing resources, and would have also increased the numbers of individuals who needed to self-isolate, potentially unnecessarily.

Testing criteria

Rapid diagnostic development meant that tests were available early in the pandemic, and testing criteria were included in some early case definitions. However, as the first wave rose in the UK, demand for testing rapidly outstripped capacity, and existing supply had to be prioritised for hospital settings. This impacted the ability to confirm cases in the community so other forms of case definition, such as symptomatic, were prioritised. (See Chapter 6: testing.) Test demand outstripping supply is likely to be the case in a future pandemic; it will be essential to ensure that diagnostic testing is scaled quickly and capacity is created for widespread community testing as early as possible. Understanding of the frequency of certain symptoms over the year (such as influenza-like illnesses in winter) can support preparations for this.^{[footnote 162]}

As testing capacity increased in spring 2020 and became more widely available in the community, testing criteria played a greater role in case definitions. Identifying cases using contact criteria, meanwhile, required effective contact tracing systems, which were under significant pressure during the first wave when community transmission rose rapidly. (See Chapter 6: testing and Chapter 7: contact tracing.) It also required a good understanding of what type of contact constituted a risk of infection, which took time to accumulate.

Evolution of COVID-19 case definitions

The earliest sources of information for the establishment of case definitions were case reports, case series and information shared by national health agencies in East Asia and the WHO.^{[footnote 15]}, ^{[footnote 163]}, ^{[footnote 164]} In December 2019, the Wuhan Municipal Health Commission reported a cluster of pneumonia cases in Wuhan, Hubei Province, China.^{[footnote 165]}, ^{[footnote 166]} By mid January 2020, the WHO had issued a report describing the clinical symptoms and signs associated with the pneumonia cluster.^{[footnote 165]} The first surveillance case definition for human infection with novel coronavirus followed soon afterwards.^{[footnote 167]}

Throughout January, reports describing the clinical signs and symptoms associated with SARS-CoV-2 infection continued to emerge, including the first published case reports and case series. ^{[footnote 15]},^{[footnote 163]}, ^{[footnote 164]}, ^{[footnote 168]}, ^{[footnote 169]} By the end of January, the New and Emerging Respiratory Virus Threats Advisory Group (NERVTAG), the Scientific Advisory Group for Emergencies (SAGE), Public Health England (PHE) (later the UK Health Security Agency or UKHSA), and the Department of Health and Social Care (DHSC) had agreed the first epidemiological case definition in the UK, the geographical element of which expanded over the following weeks.^{[footnote 154]}, ^{[footnote 155]}, ^{[footnote 160]}, ^{[footnote 170]}, ^{[footnote 171]}

In the UK, the First Few Hundred Cases Study (FF100) provided early insight into the symptom profiles of local cases, but these were generally younger and healthier cases.^{[footnote 172]} Existing surveillance studies (such as flu watch) provided useful negative controls against which to compare the symptom profile of positive cases.^{[footnote 157]}

With the passage of time, more sources of data were established. National surveillance data, with symptom surveys linked to test results, provided useful insight into symptom frequency in cases throughout. By mid to late 2020, systematic reviews and meta-analyses with large sample sizes had produced detailed summaries of symptom profiles in different age groups. Non-traditional academic sources, such as healthcare worker symptom reporting, symptom-tracker apps (such as the ZOE app) and social media, also provided information on symptom frequency, though many of these sources were not sampled in a randomised way and were therefore not representative of the population as a whole.

As new variants emerged later in the pandemic, ecological studies were used to compare symptom profiles over time.^{[footnote 150]} Observational studies with large sample sizes also allowed the accumulation of data on symptom profiles.

Population-wide or nationally representative case-control studies and longitudinal studies, and later systematic reviews and meta-analyses, ultimately provided the best insight into symptom profiles and case definitions, though they took time to establish. Studies that tested people regardless of symptoms (such as REACT and those coordinated by ONS) and compared symptom profiles in symptomatic test negative and symptomatic test positive people provided robust estimates of the sensitivity and specificity of specific symptoms, while avoiding the biases often present in national testing data.

Challenges and complexities

Throughout the pandemic, the public nature of case definitions for COVID-19 to direct people to take actions such as self-isolation added complexity. Case definitions for public use (as opposed to use by clinicians) had to be sufficiently simple to be remembered by the general public so that they could take appropriate public health actions, while correctly identifying cases sufficiently frequently for public health action. Evidence suggested that very sensitive case definitions, including many symptoms, could lead to reduced compliance with public health actions (such as testing or self-isolation) especially if they were triggered too frequently.^{[footnote 173]}

There were also important nuances to how symptoms were communicated. For example, many people did not have access to thermometers to measure fever, and so language such as feeling hot or feverish was helpful in addition to a technical definition of fever. It was also important to consider how symptoms were interpreted when transmitted into different languages.

Towards the end of 2020, co-circulation of other influenza-like illnesses threatened to impact the specificity of SARS-CoV-2 public case definitions.^{[footnote 174]} In the event, there was relatively limited co-circulation of SARS-CoV-2 and other influenza-like illnesses in the UK in winter 2020 to 2021 due to widespread implementation of NPIs, though co-circulation has since occurred.

As knowledge of symptom profiles and diagnostic testing capacity accumulated, the strategic objectives of each case definition had to be borne in mind (for example, correctly identifying as many infections as possible) and balanced with a requirement for consistency and public understanding.^{[footnote 145]}

Clinical case definitions were more widely defined throughout the pandemic than the public ones, recognising the wide range of rarer symptoms that people with COVID-19 could present with.

8. What were the important routes of transmission?

Evidence on routes of transmission was important for guiding the pandemic response, especially in the early stages where NPIs were the only interventions that were available.^{[footnote 175]}, ^{[footnote 176]}, ^{[footnote 177]}

Evidence of this kind has been important in previous pandemics and recent epidemics, such as HIV (sexual and intravenous), Ebola virus (touch) or Zika virus (vector), and it will be for any future pandemic or major epidemic.

It was established early that the likely principal route of transmission for COVID-19 was respiratory, although secondary routes including faeco-oral were not excluded. From early in the pandemic, 3 components have been considered potentially important for COVID-19: fomite, droplet and aerosol spread. However, global scientific consensus on the relative importance of these different transmission routes, and the potential role of other routes, shifted as new evidence emerged, and evidence has been continually reviewed as new variants of SARS-CoV-2 have become established.^{[footnote 178]}

There were important complexities in understanding transmission routes. First, transmission depends on multiple factors including:

• pathogen dynamics, such as viral load
• environmental factors, such as temperature and ventilation
• host-related factors, such as behavioural adaptation, immunity and contact patterns
• wider contextual factors, such as prevalence of the disease^{[footnote 175]}, ^{[footnote 176]}

Second, some routes of transmission were easier to measure than others. It was relatively rapidly identified that close contacts were at elevated risk and from that it was inferred that close range droplet transmission was likely to be important. It was less easy to identify the most likely pathway in those with more distant exposure – where respiratory particles will have been diluted by distance – as a contact event was often harder to identify.

Third, there was a need to balance the level of infection risk from a given transmission route with the frequency and likelihood of exposure to this route in day-to-day activities. Aerosol transmission across a room, for example, may present a low risk from any single exposure, but the ability for one infectious person to expose multiple people at the same time means it could present a higher population level risk in some settings than for close direct contact with an infectious person.

Finally, given the challenges inherent in attempting to determine the relative impacts of different routes of transmission, it was important to retain an open mind as understanding evolved over the course of the pandemic. It was also important to ensure that absence of evidence was not interpreted as evidence of absence, and that important transmission routes to which there were potential countermeasures were not ignored.

Expertise in public health, clinical medicine, microbiology, physics, behavioural science, built environment and data science was helpful to interpret a range of evidence on routes of transmission.

Outset: using existing knowledge

Initially, inference was drawn from studies of transmission routes for other respiratory viruses. Phylogenetic studies helped identify similarities to known viruses within the same family, in particular SARS-CoV-1.^{[footnote 178]} In retrospect, this provided mixed early indications – on the one hand, the airborne transmission capabilities of SARS-CoV-2 are similar to SARS-CoV-1; on the other, there are a number of important differences such as in timelines of transmission and the much greater role of asymptomatic transmission seen with SARS-CoV-2 (see section 10).^{[footnote 177]}

As a respiratory virus SARS-CoV-2 carried the potential for transmission via droplets and aerosols, direct physical contact, and indirect (fomite based) physical contact. Existing evidence suggested that close contact with a person with acute respiratory infection carried more risk than a more physically distant contact, implying the importance of close-range droplet and, as now understood, short-range aerosol transmission. Pre-pandemic research into other acute respiratory infections also showed the importance for transmission of exposure in public spaces including public transport, shops, restaurants, parties, theatres and places of worship, suggesting an additional potential role for more distant, primarily aerosol based, transmission.^{[footnote 179]} Existing systematic reviews showed that regular handwashing can reduce incidence of respiratory infections, implying a possible role for direct contact and/or fomite based transmission.^{[footnote 180]} This helped guide early control strategies, but the relative importance of these transmission routes for SARS-CoV-2 was initially unclear and required further investigation.

Early investigations

Early retrospective cohort studies were helpful in generating hypotheses about modes of transmission. In January 2020, for example, a retrospective cohort study of 41 patients in Wuhan, China, provided initial evidence of human transmission. The authors of the study suggested further investigation to exclude major alternate routes of transmission such as faeco-oral and recommended the use of precautions against airborne transmission.^{[footnote 178]}

Outbreaks – especially super-spreading events – also provided valuable opportunities to understand transmission dynamics at the outset of the pandemic, particularly when background prevalence was low. Well-designed outbreak investigations conducted during times of low prevalence could identify transmission from a single index case and describe the risk of infection according to proximity of contact. For example, early outbreaks in restaurants in China showed the highest risk of infection was for those with closest proximity to the index case. They also showed infections among people at distant tables, implying that some aerosol transmission had occurred – video evidence later discounted the role of fomite transmission.^{[footnote 181]}, ^{[footnote 182]} Similar findings were seen for outbreaks on coaches and trains.^{[footnote 183]}, ^{[footnote 184]} An early outbreak investigation in Germany in March 2020, combined with similar studies from China, also suggested the importance of pre-symptomatic transmission as some of those infected had only been exposed to the index case prior to that person becoming symptomatic.^{[footnote 185]}, ^{[footnote 186]}, ^{[footnote 187]} Gaining access to outbreak sites to gather samples, however, proved challenging, and at the outset of the pandemic protocols on containment levels hampered efforts to rapidly move samples. Having pre-approved emergency protocols for access and sample transportation, as well as adequate resources to investigate and take samples from outbreaks, will be important in a future pandemic. Adequate resource to undertake reviews of outbreaks occurring internationally is also important.

Systematic studies of contacts of known cases, such as the First Few Hundred approach, provided valuable evidence in the early stages of the pandemic.^{[footnote 188]} In order to describe secondary attack rates according to the nature and setting of exposure, these studies needed carefully to define the nature of the contact in terms of proximity, type of contact, duration and setting, to follow up both close and distant contacts, and to undertake regular testing of contacts regardless of symptoms.

Environmental studies were also important. One environmental study with air and surface sampling, conducted over a period of 2 weeks in a Singaporean hospital with COVID-19 patients, found environmental contamination suggestive of droplet spread, and possible faecal shedding.^{[footnote 189]} However, sampling live virus is difficult and it remained unclear whether shedding in this study indicated transmission risk.

Alongside the above relatively rapid investigations in the early months of the pandemic, there was a need to establish surveillance programmes across multiple settings to provide real-time information and therefore early warning signals on transmission by different routes in household, community, health and social care settings. However, this relied on large scale availability of testing, which was limited in early spring 2020 in the UK as testing capacity struggled to meet rapidly rising demand (for more on this process, see Chapter 6: testing).

The WHO-China Joint Mission analysis in early 2020 triangulated findings from phylogenetic and laboratory studies of COVID-19, outbreak analyses, in-depth analysis of disease progression, and published literature to outline what was known and not known with respect to COVID-19 in order to make recommendations for both China and the international community. This suggested that SARS-CoV-2 was likely to be primarily transmitted through respiratory droplets during close unprotected contact, and also by fomites, an assessment that did not change in their follow-up briefing in March 2020.^{[footnote 190]}, ^{[footnote 191]}

In recognition of the need to maintain an up-to-date overview of emerging evidence the SAGE Environment and Modelling group (EMG) was established in April 2020 to bring together a range of scientific experts to explore these issues in depth. The group continuously monitored best available evidence on transmission routes, in particular the growing evidence for the significant role of aerosol transmission.^{[footnote 192]}, ^{[footnote 193]}, ^{[footnote 194]}, ^{[footnote 195]}

Throughout the pandemic

Based on a further review of the existing evidence in July 2020, the WHO continued to recommend that direct or close contact with infected people via droplet remained the most likely principal route of transmission, and uncertainty remained about the fomite route. Multiple environmental sampling studies demonstrated presence of viable SARS-CoV-2 virus and/or RNA on surfaces for hours to days – however, there was an absence of case reports or outbreaks robustly demonstrating fomite transmission (most people who came into contact with infectious surfaces had also had close contact with an infectious person).^{[footnote 196]}, ^{[footnote 197]}

Quantitative microbial risk assessment methods, estimating viral exposure via hand–face touches based on measured environmental contamination, steadily added to the evidence base that fomite transmission risks were low, with one study concluding that each contact with a contaminated surface had less than a 1 in 10,000 chance of causing an infection.^{[footnote 198]} Epidemiological evidence for fomite transmission and the impact of interventions such as surface cleaning and hand hygiene was and remains very limited. There was a notable difference between calls for evidence of the importance of airborne transmission that were not replicated for fomite transmission, which was assumed despite little evidence to support it.

Though SARS-CoV-2 RNA had been detected in some samples of urine and faeces, there remained no published reports by summer 2020 that were able to link transmission to these routes.^{[footnote 189]} Bloodborne transmission was considered low risk due to low viral titres in blood, and there was still no evidence of intrauterine transmission.^{[footnote 199]}

As the evidence base grew, synthesis of evidence from completed studies on viral load across the respiratory tract, fluid dynamic studies examining dispersion of virus from household appliances, environmental air sampling outbreak reports, and studies in animal models all helped enhance understanding of short and long-range airborne transmission risks and the importance of ventilation.^{[footnote 196]}, ^{[footnote 200]}, ^{[footnote 201]}, ^{[footnote 202]}, ^{[footnote 203]} Despite accumulating evidence, reaching a position of confidence on the full range of transmission routes and their relative importance took longer than expected. A year into the pandemic, the WHO noted that high-quality research was still required to understand routes of transmission, infectious dose and settings in which transmission might be amplified.

As the pandemic progressed the importance of airborne transmission was increasingly recognised.^{[footnote 204]} It was established early on that transmission was far more likely indoors than outdoors, suggesting a role for the environment, and particularly dilution by air (but also the effects of sunlight), in influencing transmission. The evidence encompassed theory, observation and experiment, and included:^{[footnote 182]}, ^{[footnote 205]}, ^{[footnote 206]}, ^{[footnote 207]}, ^{[footnote 208]}, ^{[footnote 209]}, ^{[footnote 210]},^{[footnote 211]}, ^{[footnote 212]}, ^{[footnote 213]}

• outbreak reports relating to choir groups, restaurants and fitness classes
• long-range transmission in quarantine hotels between people who had had no contact with one another
• nosocomial transmission in settings where droplet-based precautions but not aerosol based ones were taken
• animal studies in caged animals which became infected despite only sharing air ducts
• air sampling studies showing infectivity of air for up to 3 hours in rooms occupied by patients with COVID-19
• experimental studies mimicking aerosol dispersion
• a substantial volume of cases arising from pre symptomatic transmission which was most likely to have occurred by the aerosol route

Some transmission events were reported to occur after an infected person had left a setting, indicating likely airborne transmission of the virus.^{[footnote 205]},^{[footnote 206]},^{[footnote 207]}

Although the fact that the respiratory route was dominant was established very early, teasing out the relative contributions of close range and longer distance airborne spread, and of fomites, presented significant challenges. Super-spreading events and rapid epidemiological studies made an important contribution to understanding transmission routes – however, relying solely on these at times led to misleading conclusions about transmission, especially because aerosol and fomite transmission were and remain harder to measure robustly than close range transmission.^{[footnote 214]} Even transmission at close range was subject to prior assumptions, with the belief that the risk was posed by large droplets rather than more concentrated small aerosols, resulting in reduced focus on masks for protection against inhalation for people at close proximity.

This pandemic highlighted the role of controlled laboratory settings in providing evidence on routes of transmission, as well as the importance of rapid investigations into survival of viable virus across different environments (using, for example, quantitative microbial risk assessment). ^{[footnote 198]},^{[footnote 215]} Different laboratory detection and sampling methods had differing abilities to detect differences between viable and non-viable virus. It is important to note that the level of viral RNA measured in an environment is not necessarily reflective of its infectivity. As an example, sampling of environments where people have influenza or Monkeypox show far more viral RNA than for SARS-CoV-2, yet the outbreak data indicate that both are much less transmissible. This suggests that a lower viral dose is needed to initiate a SARS-CoV-2 infection than for these other diseases.

There was a need to consider local circumstances when assessing the evidence. For example, early data from China suggested a limited role for healthcare settings in driving transmission, but this was in the context of important differences between these settings in China and the UK, including the imposition of different mitigation measures against aerosol transmission. ^{[footnote 190]}

9. What were the higher risk settings for transmission?

In this pandemic it has been important to understand higher risk settings for transmission in order to target mitigation measures at those locations where they would have the greatest impact.

Outset: using existing knowledge

At the outset, in the absence of specific evidence on mechanisms of transmission of SARS-CoV-2, the use of fundamental transmission principles alongside pre-existing research on respiratory-transmitted pathogens helped identify potential high-risk settings for transmission. Fundamental principles suggested that the highest risk of transmission would be in places where people from multiple households could meet, such as hospitality settings, especially if they were physically close and indoors. There were ongoing questions regarding mass events, particularly where these took place predominantly outdoors. Chapter 8 on NPIs covers this in more detail, outlining how greater understanding on this issue was reached, and outlining key epidemiological principles when considering transmission linked to mass events. Pre-existing research on respiratory pathogens supported this approach, with high transmission risks likely in settings including households, schools, hospitals, homeless hostels, prisons and nursing homes.^{[footnote 216]}, ^{[footnote 217]}, ^{[footnote 218]}, ^{[footnote 219]}, ^{[footnote 220]} There were, however, important caveats to using such evidence. The level of transmission risk within different settings can vary according to the characteristics of different infectious diseases, such as who uses such settings, who is vulnerable to severe disease, and how this might affect their behaviour. There was therefore a need to generate evidence on high-risk settings both in terms of transmission of SARS-CoV-2 and the consequences for those affected, rather than relying on existing evidence alone. It was also important to review findings as new variants became established, vaccines were rolled out, and both guidance and public behaviour changed.

Early investigations

In the first few months of the pandemic, early outbreaks gave an indication of potential high risk contexts including health and care settings, long-term living facilities particularly for older people, prisons and cruise ships.^{[footnote 190]}, ^{[footnote 191]}, ^{[footnote 221]} Later in spring 2020, evidence from early outbreaks in choir groups, restaurants and fitness classes was reported.^{[footnote 182]}, ^{[footnote 205]}, ^{[footnote 206]}, ^{[footnote 207]} Formal and informal information channels played a part in reporting possible outbreaks at speed; many apparent outbreaks were reported in the media or on social media long before they were formally described in preprints or journal articles. However, in addition to uncertainties about the reliability of such reports there was an additional important caveat to this early evidence: the majority of transmission did not take place within recognised large outbreaks, which are more likely to be identified in relatively closed settings than in more open venues such as shops or public transport where tracing of contacts is more difficult and the extent of contact often less clear. In addition, outbreak studies highlighting risks in particular settings had to be balanced with the overall epidemiological importance of that setting in a given population. For example, while shopping may not be inherently high risk, the fact that the majority of people need to shop for essential items means that it makes an important contribution to transmission.^{[footnote 222]} It should also be noted that in the early days testing was very limited, so outbreaks where multiple people were symptomatic or died would have been more likely to be reported.

Early mortality data, alongside outbreak studies, indicated that enclosed settings which housed vulnerable individuals (such as migrants, homeless people and prisoners), and health and care settings (hospitals, care homes, care settings for those with learning disabilities, domiciliary care, long stay mental health institutions) were of particular importance for both mitigation efforts and for research.^{[footnote 223]}, ^{[footnote 224]}

Differences in mortality by occupation also gave indications of potential higher risk contexts. Data from May 2020 showed that mortality was elevated in occupations with high levels of close contact with others (including health and care contact), and in those with low pay.^{[footnote 224]} Later analyses controlling for key comorbidities with COVID-19 showed that high levels of comorbidities in some occupational groups contributed to these variations, but setting and type of work remained an important factor.^{[footnote 225]} It is also important to note that industrial sectors concentrated in areas with high levels of community prevalence might have given a misleading impression that the type of business posed an elevated risk when this may in fact have primarily been a function of local prevalence or workers living close to one another or sharing social facilities.

Throughout the pandemic

From the early pandemic onwards a number of different scientific approaches were needed to understand high transmission risk settings. In the early stages, outbreak investigations, contact tracing, surveillance studies, environmental sampling, modelling studies and behavioural analysis were the approaches most likely to be able to collect data rapidly. As the pandemic progressed, longer-term methodologies such as case control studies, repeated cross-sectional studies, cohort studies, sequencing and phylogenetic studies, intervention studies and meta-analyses became possible and assumed greater importance.^{[footnote 176]} Implementing such studies required deployment of a variety of robust surveillance programmes and research to gather real-time information on cases in household, community, health and social care settings as well as rapid outbreak analysis.

In prioritising the focus of these studies, it was crucial to understand transmission dynamics and populations at risk from the pathogen as quickly as possible through live surveillance. With H1N1 influenza in 2009, young adults were most at risk, while other infectious disease such as measles generally affect children most.^{[footnote 226]}, ^{[footnote 227]} With COVID-19, demographics of those at risk became clear through outbreak and mortality patterns analysed prospectively and retrospectively in cohort studies, with the aid of electronic healthcare data.^{[footnote 223]}, ^{[footnote 224]}, ^{[footnote 228]}, ^{[footnote 229]}

Well-designed epidemiological studies took considerable time to generate statistically robust data. This required the development of reliable methods for testing and sequencing, and the rollout of these at scale. The speed at which this can happen is likely to depend on how similar the pathogen is to existing microorganisms and whether surveillance and sampling approaches for other infections can be easily adapted. It was important to have funding mechanisms, data governance and data sharing agreements in place, and to plan and initiate them as rapidly as possible. They also relied heavily on availability of testing and contact tracing, both of which were running at very limited capacity in the early part of the UK’s first wave, and on community surveillance such as the ONS CIS, which went live in April 2020, the same month the UK’s first wave peaked nationally.

It was more difficult to generate new evidence on potential high transmission risk settings such as hospitality, some workplaces, or schools during periods of intense restrictions as many such settings were highly restricted or closed down, and thus unable to contribute to generating evidence. Analyses of the Virus Watch cohort submitted to SAGE in December 2021 showed that during restrictions in winter 2020 to 2021 leaving home for work, using public transport, and shopping were all important risk factors for transmission. Following lifting of restrictions all of these activities remained relevant, but other activities which had previously been restricted (such as visiting pubs and restaurants) increased their relative importance as risk factors.^{[footnote 222]}

Analyses that brought together multiple study types were helpful in highlighting consistent signals from particular settings. For example, an analysis of COVID-19 outbreaks in hospitality, retail and leisure facilities in the UK and worldwide, presented to SAGE in January 2021, used multiple analytical approaches to examine transmission risks in these settings including:

• social contacts over time
• case-control studies
• secondary attack rates
• cluster concordance ^{[footnote 230]}

It reinforced the initial fundamental principles outlined above that transmission risks were highest in settings that were poorly ventilated and crowded, where mixing was for extended periods of time, and where population turnover was high. Analysis of cases by occupation and sector also highlighted that risk is not necessarily the same across a particular sector or indeed within a setting, with a range of socio-economic factors influencing risks.^{[footnote 231]} For example, food processing is a sector that has been associated with multiple large outbreaks, with analysis suggesting that the likelihood of transmission depends not only on the characteristics of the settings (such as ventilation, social distancing), but also the socio-economic characteristics of the workforce, including shared housing, lack of sick pay (creating pressure to continue working even if unwell), and use of shared transport.^{[footnote 232]} It was difficult to differentiate beyond fundamental principles to attribute causation to particular properties of these specific settings which increased or reduced risk.^{[footnote 230]}

Understanding transmission risks in different settings was a complex process for a number of reasons, some of which have been outlined above. First, transmission risk in settings was linked to factors that changed throughout the pandemic, across different settings and communities, in response to changing guidance, behaviours and mitigating measures: contact patterns (the type, frequency, proximity and duration of contacts and networks of contacts), levels of immunity, and environmental factors such as ventilation or occupant density.^{[footnote 176]},^{[footnote 233]}

Second, transmission risk may vary depending on factors particular to specific settings (rather than setting types or sectors) such as ventilation or proximity to others in a building. Society-wide guidance for different setting types needed to be accompanied by risk assessments tailored to particular locations, and adaptations that considered the range of activities as well as the environment. Third, background community prevalence and the changing epidemiology of the pandemic needed to be considered. For example, a retrospective study examining outbreaks recorded in educational settings between June and July 2020 when community prevalence was relatively low noted that outbreaks were uncommon; transmission in educational settings was higher later in the pandemic as new variants became established and prevalence rose again.^{[footnote 234]}

Transmission risk in settings was a dynamic factor throughout the pandemic, and this ongoing risk of time-varying and contextual confounding meant that although some settings were indicated as potentially higher risk through epidemiological studies, the level of that risk was complex to assess. Cross-disciplinary expertise across epidemiology, health, microbiology and understanding of specifics behaviours and environments supported interpretation of potential risks.

10. What was the proportion of asymptomatic infection and transmission, and could this maintain R over 1?

Overview

From the outset, asymptomatic infection and transmission were considered possible, but the extent of each was not understood. Existing knowledge of other related human coronaviruses suggested that asymptomatic infection and transmission were possible, but it was difficult to extrapolate directly, and work was needed to clarify:

• the proportion of infections that were asymptomatic
• the role of asymptomatic transmission

These parameters are complex and quantitative, and their estimation required the continual balancing of multiple types of emerging evidence. Continual reassessment of this evidence was also required, as the immunity profile of the population changed due to infection-induced and vaccine-derived immunity, and as new variants emerged. There was conflation of asymptomatic infection and asymptomatic transmission in some public reporting, and it was necessary to highlight that asymptomatic infection does not necessarily lead to asymptomatic transmission (though it was a prerequisite).

Knowing the proportion of infections that were asymptomatic was important for case detection strategies and determining the infection fatality rate. Understanding the role of asymptomatic transmission was important for identifying which public health measures would likely bring R below 1. Transmission of infection from asymptomatic cases can be difficult to control, and the infectious timeline is difficult to establish in the absence of symptoms as a marker of infection or infectiousness, adding complexity to disease control.^{[footnote 235]}, ^{[footnote 236]}

Asymptomatic cases cannot be detected in the absence of testing, and in the early pandemic the global and UK constraints on test availability significantly slowed the estimation of asymptomatic cases.

Proportion of infections that were asymptomatic

The proportion of SARS-CoV-2 infections that were asymptomatic was defined using 2 different numerators:

• PCR positivity
• antibody positivity

PCR positivity was technically easier to assess but had a shorter duration, which may have resulted in undercounting of infections in some studies. Serology was more labour intensive to collect and analyse, but has a longer duration, providing a more accurate estimate of infection proportions.

There was difficulty in identifying asymptomatic cases as the majority of testing took place in those who were symptomatic, particularly in the early stages of the pandemic when limited tests had to be prioritised.

Simpler study designs (such as cross-sectional studies) were unable to differentiate between asymptomatic and pre-symptomatic infections.^{[footnote 237]} Although these produced estimates of the proportion of asymptomatic infections at pace, they were likely inflated by the inclusion of some pre and post-symptomatic individuals.^{[footnote 238]}

Role of asymptomatic transmission

It was likewise challenging to distinguish between asymptomatic, pauci-symptomatic and pre-symptomatic transmission.^{[footnote 239]} Where studies had designs which did not enable the differentiation of pre and asymptomatic transmission, there was a tendency to over-report cases resulting from asymptomatic transmission.^{[footnote 236]}

Transmission from one person to another depends on a number of factors including shedding of viable virus and behaviours and contact patterns, noting that asymptomatic people may be more likely to be unaware of infection than symptomatic people.

Methods to understand the proportion and relative infectiousness of asymptomatic infections

1. Case series and cluster investigations provided early signals that asymptomatic infection and transmission were possible while more robust data was being collected.^{[footnote 240]}, ^{[footnote 241]}
2. Longitudinal designs which collected information on symptoms over time (and thus were able to differentiate between asymptomatic and pre-symptomatic infections) were needed to calculate reliable estimates of the asymptomatic proportion.^{[footnote 242]}
3. Longitudinal studies were also required to understand the potential for transmission from asymptomatic cases. These studies addressed secondary attack rates in households with asymptomatic infections and/or included serial viral culture to indicate the presence of live, infectious virus.^{[footnote 243]}
4. Studies in institutional settings (nursing homes, army barracks) were among the earliest established, and enabled the estimation of asymptomatic proportions and relative infectiousness more quickly.^{[footnote 244]}, ^{[footnote 245]}, ^{[footnote 246]} However, their applicability to the general population was potentially limited.^{[footnote 236]}
5. Viral culture was the optimal tool for assessing infectiousness in both symptomatic and asymptomatic cases but was not widely available.

Summary of the types of evidence available, and broad timelines

For SARS-Cov-2, the asymptomatic proportion and the relative infectiousness of asymptomatic individuals varied substantially depending on the setting and characteristics of the individuals involved. In addition, they changed over time as the population gained protection from prior infection or vaccination and viral variants with different biological properties emerged.^{[footnote 238]}

Early case and cluster reports raised the possibility of asymptomatic infection and transmission but often with poor differentiation between asymptomatic and pre-symptomatic transmission.^{[footnote 240]}, ^{[footnote 247]}, ^{[footnote 248]} At this stage, robust data on asymptomatic infections and whether they may be infectious to others was lacking, and estimates of the asymptomatic proportion varied widely. ^{[footnote 249]}

After a few months, outbreak studies in closed or institutional environments provided early estimates of the asymptomatic proportion of PCR-confirmed cases, but may have included pre-symptomatic cases. Descriptive reports of transmission chains and clusters described apparently asymptomatic transmission.^{[footnote 241]}, ^{[footnote 250]}

Over time, evidence of positive tests in asymptomatic individuals mounted, and more robust data on asymptomatic transmission emerged. Estimates of the asymptomatic proportion were high. Cross-sectional studies were conducted which were unable to differentiate between pre and asymptomatic transmission.

By mid 2020, further estimates of the asymptomatic proportion in closed and/or institutional settings had been published, and the first evidence that infectious virus could be recovered from asymptomatic individuals emerged.^{[footnote 244]}, ^{[footnote 245]}, ^{[footnote 246]}, ^{[footnote 251]}, ^{[footnote 252]}, ^{[footnote 253]} Early systematic reviews and meta analyses of asymptomatic proportions followed, with wide variation in the estimates of the asymptomatic proportion, and lower estimates from studies that were better able to differentiate between pre and asymptomatic cases.^{[footnote 238]}, ^{[footnote 242]} Around this time, early data comparing cycle threshold (Ct) values between asymptomatic and symptomatic individuals became available, though the link between Ct values and infectiousness was not firmly established.^{[footnote 245]}, ^{[footnote 254]}, ^{[footnote 255]}, ^{[footnote 256]}

Eventually, large random-sample swabbing studies, such as REACT and those led by the ONS, were established and provided robust estimates of the asymptomatic proportion on a regular basis. By mid to late 2020, studies of household transmission had been established that were able to robustly identify asymptomatic infections and transmission, and the viral load dynamics in asymptomatic individuals had been characterised.^{[footnote 243]}, ^{[footnote 254]}, ^{[footnote 257]}

Establishing that asymptomatic transmission occurred was well in advance of establishing what proportion of transmission was from asymptomatic people, and whether, if all symptomatic transmission ceased (for example, due to case isolation) asymptomatic transmission alone was capable of sustaining the reproduction number (R) above 1.

11. How long were people infectious?

Understanding duration of infectiousness is central to infection prevention and control and will be for any future pandemic or epidemic. Infections vary widely in their duration of infectiousness from a few days to lifelong (the last major new pandemic, HIV, was lifelong when untreated). It was important to understand the duration of the infectious period of SARS-CoV-2 in order to make informed decisions on the duration of isolation and contact tracing windows, to optimise prevention of transmission in health and care settings, and to be able to understand and model the dynamics of the pandemic.

For SARS-CoV-2, epidemiological and virological methods were primarily used to develop this understanding.

Detecting the presence of SARS-CoV-2 virus

The presence of SARS-CoV-2 virus was an essential piece of information for determining timelines of infectiousness. Presence of SARS-CoV-2 virus can be detected in several ways:

• RT-PCR testing (see Chapter 6: testing): detects the presence of virus genetic material but does not reliably indicate viable infectious viral particles. It provides Ct values, which allow estimation of the amount of virus present in a sample. Ct values correlate with, but are not a predictor for, infectiousness^{[footnote 258]}
• virus culture: detects the presence of live infectious virus, thus can be used as a proxy for infectiousness^{[footnote 19]}
• rapid antigen: detects the presence of viral antigen in a clinical sample

RT-PCR testing was used to detect infection, and measurement of Ct values on RT-PCR allowed quantification of the amount of virus present in a sample. Serial Ct values, obtained using the same type of assay, were used to show the variation in viral load in an individual over time.^{[footnote 259]}, ^{[footnote 260]} Ct values were also used as a proxy for infectiousness. A reasonably firm correlation between cycle threshold values and the presence of live, infectious virus was established approximately 6 months into the COVID-19 pandemic through studies with serial sampling, RT-PCR testing and viral culture.^{[footnote 261]}, ^{[footnote 262]}

Early in the SARS-CoV-2 pandemic, clinical sampling was of variable quality and there was wide variation in diagnostic targets and sensitivity.^{[footnote 263]} Clinical samples were obtained on relatively small numbers of individuals, often after symptom onset and without systematic follow up. Estimates of trends in viral load throughout the entire course of illness, as measured by RT-PCR, were available but low certainty until 6 to 8 months into the pandemic.^{[footnote 261]}, ^{[footnote 262]}, ^{[footnote 264]}

Viral culture was used to infer infectiousness. Results were not available in the UK until 3 to 4 months into the pandemic, and at this time, studies assessing the presence of infectious virus through viral culture were few and based on small numbers of persons and datapoints.^{[footnote 265]} The most timely datasets came from both international sources and PHE’s laboratories, which shared results with expert groups at 6 months.^{[footnote 261]}

Ultimately, longitudinal studies with serial sampling of cases, quantitative RT-PCR and viral culture allowed the most direct measures of the kinetics of infectiousness.^{[footnote 19]},^{[footnote 259]} As would be expected, SARS-CoV-2 viral load dynamics and kinetics of infectiousness were found to vary between individuals depending on symptom severity, immune response, prior infection and vaccination status.^{[footnote 259]}, ^{[footnote 260]}, ^{[footnote 266]}

Timeline of discovery

Initially, knowledge of other coronaviruses (SARS-CoV-1 and MERS-CoV) was used to develop broad estimates of the expected kinetics of viral shedding of SARS-Cov-2, but this needed to be supplemented with pathogen-specific evidence.^{[footnote 267]}, ^{[footnote 268]}, ^{[footnote 269]}, ^{[footnote 270]}

Epidemiological studies of transmission chains provided the earliest estimates of infectious periods. Studies of clusters and chains of transmission, and early models of transmission dynamics, were used to infer the infectious period.

After 3 to 4 months, initial estimates of the infectious period, informed by longitudinal data on viral shedding, were available.^{[footnote 14]},^{[footnote 50]},^{[footnote 271]},^{[footnote 272]} The first viral culture results from the UK became available in April 2020.^{[footnote 265]}, ^{[footnote 273]} At this time, absolute numbers of data points and persons investigated remained small.

By mid 2020, accumulating data on viral dynamics (as measured by RT-PCR) had demonstrated a peak in viral load at the onset of symptoms, followed by a gradual decline in viral load.^{[footnote 263]} Viral culture data suggested that cultivable virus levels were correlated with PCR values and time after symptom onset, and that viable virus could be isolated from pre-symptomatic cases, providing support for infectiousness of pre-symptomatic cases.^{[footnote 261]}, ^{[footnote 262]}, ^{[footnote 274]} Longitudinal or cross-sectional sampling and culture showed that beyond 14 days the majority of infected people shed virus at amounts lower than could be cultured, suggesting they were no longer infectious.^{[footnote 261]}, ^{[footnote 275]}, ^{[footnote 276]}

By the end of 2020 there was a robust understanding of viral dynamics over time. Further data emerged to suggest a strong relationship between Ct values and ability to recover viable virus.^{[footnote 19]}, ^{[footnote 264]}, ^{[footnote 277]} Throughout 2021, comparisons of viral kinetics across people infected with different variants were undertaken, as well as across vaccinated and unvaccinated individuals.^{[footnote 259]}, ^{[footnote 260]}, ^{[footnote 278]}, ^{[footnote 279]}

Later in the pandemic, human challenge studies in controlled environments with systematic daily sampling allowed complete characterisation of the viral dynamics of infection, though these were often limited to young, healthy volunteers.^{[footnote 8]}

Reflections and advice for a future CMO or GCSA

Most of the reflections are in the body of the text above, but in addition we would highlight the following.

Point 1

Scientific and medical advice will often need to be formulated on the basis of limited data.

This was the case for SARS-CoV-2 in early 2020 with respect to several areas, including, for example, asymptomatic transmission or spread via aerosols.

This cannot be avoided but it is critical therefore to explain in the advice the strength of the evidence and the degree of uncertainty about the conclusions, and to prepare the ground for the advice to change as evidence accumulates.

Point 2

Understanding the pathogen and the disease was a global effort, particularly at the outset, and sharing data and expertise from the beginning was key.

Reports from China and Italy were critical in this respect. Personal and professional networks of CMOs, the GCSA, public health leaders and SAGE participants were invaluable. In some cases, rapid identification of counterparts in other countries was difficult and establishing clear points of contact in preparation for future emergencies would be helpful.

Point 3

Gaining a clear understanding of the pathogen and the disease required an array of cross-disciplinary studies to be initiated quickly.

Many study types and disciplines were needed but some study designs set up early in the pandemic delivered useful evidence across multiple areas. These included:

• longitudinal cohort studies with relevant baseline measures and systematic symptom review
• linked or shared surveillance data with demographic details
• clinical studies of patients with severe disease

Point 4

Building on and adapting existing research systems and networks was usually much faster than setting up new systems, but strong leadership, direction and coordination are required.

‘Peacetime’ processes were adapted, bringing together funders, researchers, CMOs, the GCSA and PHE (later UKHSA) to mobilise sufficient resources and stand up research rapidly.

Point 5

Viral variants, population behaviours and population immunity changed significantly over time requiring continuation of studies.

In contrast to some infectious agents, pathogenesis and disease characteristics of COVID-19 continually changed over the first 2 years. This needed continual review and re-validation of tools, for example:

• revalidating assays for testing
• revalidating vaccine efficacy
• adapting models

References

1. Zhou P, Yang X-L, Wang X-G, Hu B, Zhang L, Zhang W, et al. A pneumonia outbreak associated with a new coronavirus of probable bat origin. Nature. 2020 Mar 12;579(7798):270–3 ↩ ↩²

2. Lu R, Zhao X, Li J, Niu P, Yang B, Wu H, et al. Genomic characterisation and epidemiology of 2019 novel coronavirus: implications for virus origins and receptor binding. Lancet. 2020 Feb;395(10224):565–74 ↩

3. Andersen KG, Rambaut A, Lipkin WI, Holmes EC, Garry RF. The proximal origin of SARS-CoV-2. Nat Med. 2020 Apr 17;26(4):450–2 ↩

4. The Novel Coronavirus Pneumonia Emergency Response Epidemiology Team. The Epidemiological Characteristics of an Outbreak of 2019 Novel Coronavirus Diseases (COVID-19) in China. China CDC Weekly, 2020, 2(8): 113-122. ↩

5. GISAID ↩

6. SARS-CoV-2 (COVID-19) environmental persistence and potential infection risk: review of data, 14 February 2020 ↩

7. ISARIC, COVID-19 research and resources ↩

8. Killingley B, Mann AJ, Kalinova M, Boyers A, Goonawardane N, Zhou J, et al. Safety, tolerability and viral kinetics during SARS-CoV-2 human challenge in young adults. Nat Med. 2022 Mar 31 ↩ ↩²

9. Corman VM, Landt O, Kaiser M, Molenkamp R, Meijer A, Chu DK, et al. Detection of 2019 novel coronavirus (2019-nCoV) by real-time RT-PCR. Eurosurveillance. 2020 Jan 23;25(3). ↩

10. COVID-19 – The role of viral genomics in shaping the response to the pandemic ↩

11. Walker AS, Vihta K-D, Gethings O, Pritchard E, Jones J, House T, et al. Tracking the Emergence of SARS-CoV-2 Alpha Variant in the United Kingdom. N Engl J Med. 2021 Dec 30;385(27):2582–5. ↩ ↩² ↩³

12. Singanayagam A and Zambon M, PHE Virology Cell. Clinical virology of SARS-CoV-2, 17 February 2020 ↩

13. Pan Y, Zhang D, Yang P, Poon LLM, Wang Q. Viral load of SARS-CoV-2 in clinical samples. Lancet Infect Dis. 2020 Feb;0(0). ↩

14. Zou L, Ruan F, Huang M, Liang L, Huang H, Hong Z, et al. SARS-CoV-2 Viral Load in Upper Respiratory Specimens of Infected Patients. N Engl J Med. 2020 Mar 19;382(12):1177–9. ↩ ↩²

15. Holshue ML, DeBolt C, Lindquist S, Lofy KH, Wiesman J, Bruce H, et al. First Case of 2019 Novel Coronavirus in the United States. N Engl J Med. 2020 Jan 31;NEJMoa2001191. ↩ ↩² ↩³

16. Understanding cycle threshold in SARS-CoV-2 RT-PCR. A guide for health protection teams ↩

17. To KK-W, Tsang OT-Y, Yip CC-Y, Chan K-H, Wu T-C, Chan JM-C, et al. Consistent Detection of 2019 Novel Coronavirus in Saliva. Clin Infect Dis. 2020 Jul 28;71(15):841–3. ↩

18. Tu Y-P, Jennings R, Hart B, Cangelosi GA, Wood RC, Wehber K, et al. Swabs Collected by Patients or Health Care Workers for SARS-CoV-2 Testing. N Engl J Med. 2020 Jul 30;383(5):494–6. ↩

19. Singanayagam A, Patel M, Charlett A, Lopez Bernal J, Saliba V, Ellis J. Duration of infectiousness and correlation with RT-PCR cycle threshold values in cases of COVID-19. England Eurosurveillance. 2020;25:32. ↩ ↩² ↩³ ↩⁴

20. Serological surveillance of COVID-19 in England: Sera Collection Protocol ↩

21. Tortorici, M. A. & Veesler, D. Structural insights into coronavirus entry. Adv. Virus Res. 105, 93–116 (2019). ↩

22. Saif LJ. Animal coronavirus vaccines: lessons for SARS. Dev Biol (Basel). 2004;119:129-40. ↩

23. Wang Y, Tai W, Yang J, Zhao G, Sun S, Tseng CK, et al. Receptor-binding domain of MERS-CoV with optimal immunogen dosage and immunization interval protects human transgenic mice from MERS-CoV infection. Hum Vaccin Immunother. 2017;13(7):1615-24. ↩

24. Tai W, Zhao G, Sun S, Guo Y, Wang Y, Tao X, et al. A recombinant receptor-binding domain of MERS-CoV in trimeric form protects human dipeptidyl peptidase 4 (hDPP4) transgenic mice from MERS-CoV infection. Virology. 2016;499:375-82. ↩

25. Qiu M, Shi Y, Guo Z, Chen Z, He R, Chen R, et al. Antibody responses to individual proteins of SARS coronavirus and their neutralization activities. Microbes Infect. 2005;7(5-6):882-9. ↩

26. Li J, Ulitzky L, Silberstein E, Taylor DR, Viscidi R. Immunogenicity and protection efficacy of monomeric and trimeric recombinant SARS coronavirus spike protein subunit vaccine candidates. Viral Immunol. 2013;26(2):126-32. ↩

27. Sariol A, Perlman S. Lessons for COVID-19 Immunity from Other Coronavirus Infections. Immunity. 2020;53(2):248-263 ↩

28. Wan Y, Shang J, Graham R, Baric RS, Li F. Receptor Recognition by the Novel Coronavirus from Wuhan: an Analysis Based on Decade-Long Structural Studies of SARS Coronavirus. J Virol. 2020;94(7) ↩

29. Walls AC, Park Y-J, Tortorici MA, Wall A, McGuire AT, Veesler D. Structure, Function, and Antigenicity of the SARS-CoV-2 Spike Glycoprotein. Cell. 2020;181(2):281-92.e6 ↩

30. Wrapp D, Wang N, Corbett KS, Goldsmith JA, Hsieh CL, Abiona O, et al. Cryo-EM structure of the 2019-nCoV spike in the prefusion conformation. Science. 2020;367(6483):1260-3 ↩

31. Liu L, Wang P, Nair MS, Yu J, Rapp M, Wang Q, Luo Y, Chan JF, Sahi V, Figueroa A, Guo XV, Cerutti G, Bimela J, Gorman J, Zhou T, Chen Z, Yuen KY, Kwong PD, Sodroski JG, Yin MT, Sheng Z, Huang Y, Shapiro L, Ho DD. Potent neutralizing antibodies against multiple epitopes on SARS-CoV-2 spike. Nature. 2020 Aug;584(7821):450-456. doi: 10.1038/s41586-020-2571-7. Epub 2020 Jul 22. PMID: 32698192 ↩ ↩²

32. Wajnberg A, Amanat F, Firpo A, Altman DR, Bailey MJ, Mansour M, McMahon M, Meade P, Mendu DR, Muellers K, Stadlbauer D, Stone K, Strohmeier S, Simon V, Aberg J, Reich DL, Krammer F, Cordon-Cardo C. Robust neutralizing antibodies to SARS-CoV-2 infection persist for months. Science. 2020 Dec 4;370(6521):1227-1230. Epub 2020 Oct 28. PMID: 33115920; PMCID: PMC7810037 ↩

33. Seow J, Graham C, Merrick B, Acors S, Pickering S, Steel KJA et al. Longitudinal observation and decline of neutralizing antibody responses in the three months following SARS-CoV-2 infection in humans. Nature Microbiology. 2020 Dec 1;5(12):1598-1607 ↩ ↩²

34. Ju B, Zhang Q, Ge J, Wang R, Sun J, Ge X, Yu J, Shan S, Zhou B, Song S, Tang X, Yu J, Lan J, Yuan J, Wang H, Zhao J, Zhang S, Wang Y, Shi X, Liu L, Zhao J, Wang X, Zhang Z, Zhang L. Human neutralizing antibodies elicited by SARS-CoV-2 infection. Nature. 2020 Aug;584(7819):115-119. Epub 2020 May 26. PMID: 32454513 ↩

35. Boddington N, et al. COVID-19 in Great Britain: epidemiological and clinical characteristics of the first few hundred (FF100) cases: a descriptive case series and case control analysis ↩

36. COG-UK: A UK-wide collaborative network for SARS-CoV-2 genomics, research and training ↩

37. Harvala H, Mehew J, Robb ML, Ijaz S, Dicks S, Patel M, et al. Convalescent plasma treatment for SARS-CoV-2 infection: analysis of the first 436 donors in England, 22 April to 12 May 2020. Euro Surveill. 2020;25(28) ↩

38. Chandrashekar A, Liu J, Martinot AJ, McMahan K, Mercado NB, Peter L, et al. SARS-CoV-2 infection protects against rechallenge in rhesus macaques. Science. 2020;369(6505):812-7. ↩

39. Deng W, Bao L, Liu J, Xiao C, Liu J, Xue J, et al. Primary exposure to SARS-CoV-2 protects against reinfection in rhesus macaques. Science. 2020;369(6505):818-23. ↩

40. Shi R, Shan C, Duan X, Chen Z, Liu P, Song J, et al. A human neutralizing antibody targets the receptor-binding site of SARS-CoV-2. Nature. 2020;584(7819):120-4. ↩

41. Dan JM, Mateus J, Kato Y, Hastie KM, Yu ED, Faliti CE, et al. Immunological memory to SARS-CoV-2 assessed for up to 8 months after infection. Science. 2021;371(6529):eabf4063. ↩ ↩² ↩³

42. Rodda LB, Netland J, Shehata L, Pruner KB, Morawski PA, Thouvenel CD, et al. Functional SARS-CoV-2-Specific Immune Memory Persists after Mild COVID-19. Cell. 2021;184(1):169-83.e17. ↩

43. Peng Y, Mentzer AJ, Liu G, Yao X, Yin Z, Dong D, et al. Broad and strong memory CD4(+) and CD8(+) T cells induced by SARS-CoV-2 in UK convalescent individuals following COVID-19. Nat Immunol. 2020;21(11):1336-45. ↩ ↩²

44. Luke TC, Kilbane EM, Jackson JL, et al. Meta-analysis: convalescent blood products for Spanish influenza pneumonia: a future H5N1 treatment? Ann Intern Med. 2006;145:599–609. ↩

45. Hung IF, To KK, Lee CK, et al. Convalescent plasma treatment reduced mortality in patients with severe pandemic influenza A (H1N1) 2009 virus infection. Clin Infect Dis. 2011;52:447–456. ↩

46. Mair-Jenkins J, Saavedra-Campos M, Baillie JK, et al. The effectiveness of convalescent plasma and hyperimmune immunoglobulin for the treatment of severe acute respiratory infections of viral etiology: a systematic review and exploratory meta-analysis. J Infect Dis. 2015;211:80–90 ↩

47. RECOVERY Collaborative Group. Convalescent plasma in patients admitted to hospital with COVID-19 (RECOVERY): a randomised controlled, open-label, platform trial. Lancet. 2021 May 29;397(10289):2049-2059. Epub 2021 May 14. PMID: 34000257; PMCID: PMC8121538 ↩

48. Sheahan TP, Sims AC, Graham RL, Menachery VD, Gralinski LE, Case JB, et al. Broad-spectrum antiviral GS-5734 inhibits both epidemic and zoonotic coronaviruses. Sci Transl Med. 2017;9(396). ↩

49. Chu CM, Cheng VC, Hung IF, Wong MM, Chan KH, Chan KS, et al. Role of lopinavir/ritonavir in the treatment of SARS: initial virological and clinical findings. Thorax. 2004;59(3):252-6. ↩

50. Zhou F, Yu T, Du R, Fan G, Liu Y, Liu Z, et al. Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study. Lancet. 2020;395(10229):1054-62. ↩ ↩²

51. Choy, E.H., De Benedetti, F., Takeuchi, T. et al. Translating IL-6 biology into effective treatments. Nat Rev Rheumatol 16, 335–345 (2020) ↩

52. An integrated national scale SARS-CoV-2 genomic surveillance network. The Lancet Microbe. 2020;1(3):e99-e100. ↩

53. Plessis Ld, McCrone JT, Zarebski AE, Hill V, Ruis C, Gutierrez B, et al. Establishment and lineage dynamics of the SARS-CoV-2 epidemic in the UK. Science. 2021;371(6530):708-12. ↩

54. Volz E, Hill V, McCrone JT, Price A, Jorgensen D, O’Toole Á, et al. Evaluating the Effects of SARS-CoV-2 Spike Mutation D614G on Transmissibility and Pathogenicity. Cell. 2021;184(1):64-75.e11. ↩

55. Korber B, Fischer WM, Gnanakaran S, Yoon H, Theiler J, Abfalterer W, et al. Tracking Changes in SARS-CoV-2 Spike: Evidence that D614G Increases Infectivity of the COVID-19 Virus. Cell. 2020;182(4):812-27.e19. ↩ ↩²

56. Aggarwal D, Page AJ, Schaefer U, Savva GM, Myers R, Volz E, et al. Genomic assessment of quarantine measures to prevent SARS-CoV-2 importation and transmission. Nature Communications. 2022;13(1):1012. ↩

57. Rambaut A, Loman N, Pybus O, Barclay W, Barrett J, Carabelli A, et al. Preliminary genomic characterisation of an emergent SARS-CoV-2 lineage in the UK defined by a novel set of spike mutations. Virological; 2020. ↩

58. Thorne LG, Bouhaddou M, Reuschl A-K, Zuliani-Alvarez L, Polacco B, Pelin A, et al. Evolution of enhanced innate immune evasion by SARS-CoV-2. Nature. 2022;602(7897):487-95. ↩

59. Liu Y, Liu J, Plante KS, Plante JA, Xie X, Zhang X, et al. The N501Y spike substitution enhances SARS-CoV-2 infection and transmission. Nature. 2022;602(7896):294-9. ↩

60. Tegally H, Wilkinson E, Giovanetti M, Iranzadeh A, Fonseca V, Giandhari J, et al. Detection of a SARS-CoV-2 variant of concern in South Africa. Nature. 2021;592(7854):438-43. ↩

61. McCrone JT, Hill V, Bajaj S, Pena RE, Lambert BC, Inward R, et al. Context-specific emergence and growth of the SARS-CoV-2 Delta variant. medRxiv. 2021:2021.12.14.21267606 ↩

62. Campbell F, Archer B, Laurenson-Schafer H, Jinnai Y, Konings F, Batra N, et al. Increased transmissibility and global spread of SARS-CoV-2 variants of concern as at June 2021. Eurosurveillance. 2021;26(24):2100509. ↩

63. Mlcochova P, Kemp SA, Dhar MS, Papa G, Meng B, Ferreira IATM, et al. SARS-CoV-2 B.1.617.2 Delta variant replication and immune evasion. Nature. 2021;599(7883):114-9 ↩

64. Viana R, Moyo S, Amoako DG, Tegally H, Scheepers C, Althaus CL, et al. Rapid epidemic expansion of the SARS-CoV-2 Omicron variant in southern Africa. Nature. 2022. ↩

65. Cao Y, Wang J, Jian F, Xiao T, Song W, Yisimayi A, et al. Omicron escapes the majority of existing SARS-CoV-2 neutralizing antibodies. Nature. 2022;602(7898):657-63 ↩

66. Harari S, Tahor M, Rutsinsky N, Meijer S, Miller D, Henig O, et al. Drivers of adaptive evolution during chronic SARS-CoV-2 infections. Nature Medicine. 2022. ↩

67. Hill V, Du Plessis L, Peacock TP, Aggarwal D, Colquhoun R, Carabelli AM, et al. The origins and molecular evolution of SARS-CoV-2 lineage B.1.1.7 in the UK. bioRxiv. 2022:2022.03.08.481609 ↩

68. Fineberg HV. Report of the Review Committee on the Functioning of the International Health Regulations (2005) in relation to Pandemic (H1N1) 2009 Geneva 2011. ↩

69. Ghani AC, Donnelly CA, Cox DR, Griffin JT, Fraser C et al. Methods for Estimating the Case Fatality Ratio for a Novel, Emerging Infectious Disease. American Journal of Epidemiology 2005;162(5):479-486 ↩

70. Donaldson LJ, Rutter PD, Ellis BM, Greaves FEC, Mytton OT et al. Mortality from pandemic A/H1N1 2009 influenza in England: public health surveillance study. BMJ (Clinical research ed 2009;339:b5213 ↩

71. Wong JY, Kelly H, Ip DKM, Wu JT, Leung GM et al. Case fatality risk of influenza A (H1N1pdm09): a systematic review. Epidemiology (Cambridge, Mass) 2013;24(6):830-841 ↩

72. Lessler J, Reich NG, Cummings DA, New York City Department of H, Mental Hygiene Swine Influenza Investigation T et al. Outbreak of 2009 pandemic influenza A (H1N1) at a New York City school. N Engl J Med 2009;361(27):2628-2636. ↩

73. Kucharski et al. Lancet ID 2020. Early dynamics of transmission and control of COVID-19: a mathematical modelling study ↩

74. The Novel Coronavirus Pneumonia Emergency Response Epidemiology T. The Epidemiological Characteristics of an Outbreak of 2019 Novel Coronavirus Diseases (COVID-19) — China, 2020. China CDC Weekly 2020;2(8):113-122 ↩

75. Wu JT, Leung K, Bushman M, Kishore N, Niehus R et al. Estimating clinical severity of COVID-19 from the transmission dynamics in Wuhan, China. Nature Medicine 2020;26(4):506-510 ↩

76. Russell TW, Hellewell J, Jarvis CI, van Zandvoort K, Abbott S et al. Estimating the infection and case fatality ratio for coronavirus disease (COVID-19) using age-adjusted data from the outbreak on the Diamond Princess cruise ship, February 2020. Eurosurveillance 2020 ↩

77. Russell et al. Estimating the infection and case fatality ratio for coronavirus disease (COVID-19) using age-adjusted data from the outbreak on the Diamond Princess cruise ship, February 2020. Eurosurveillance 2020 ↩

78. Linton et al. Journal of Clinical Medicine. 2020;9(2):538 ↩

79. UK Health Security Agency blog: Coronavirus (COVID-19): Using data to track the virus. 23 April 2020. ↩

80. Verity R, Okell LC, Dorigatti I, Winskill P, Whittaker C et al. Estimates of the severity of coronavirus disease 2019: a model-based analysis. The Lancet infectious diseases 2020;20(6):669-677 ↩

81. Onder G, Rezza G, Brusaferro S. Case-Fatality Rate and Characteristics of Patients Dying in Relation to COVID-19 in Italy. JAMA 2020;323(18):1775-1776 ↩

82. Gao M, Piernas C, Astbury NM, Hippisley-Cox J, O’Rahilly S et al. Associations between body-mass index and COVID-19 severity in 6 to 9 million people in England: a prospective, community-based, cohort study. The Lancet Diabetes & Endocrinology 2021;9(6):350-359 ↩

83. PHE. Beyond the data: Understanding the impact of COVID-19 on BAME groups. June 2020. ↩

84. PHE. Disparities in the risk and outcomes of COVID-19. 2020 ↩

85. COVID-19 Health Inequalities Monitoring for England (CHIME) tool ↩

86. Wang, Haidong et al. Estimating excess mortality due to the COVID-19 pandemic: a systematic analysis of COVID-19-related mortality, 2020–21 The Lancet, Volume 399, Issue 10334, 1513 - 1536 ↩

87. Menachemi N, Dixon BE, Wools-Kaloustian KK, Yiannoutsos CT, Halverson PK. How Many SARS-CoV-2-Infected People Require Hospitalization? Using Random Sample Testing to Better Inform Preparedness Efforts. J Public Health Manag Pract 2021;27(3):246-250 ↩

88. Nyberg T, Twohig KA, Harris RJ, Seaman SR, Flannagan J et al. Risk of hospital admission for patients with SARS-CoV-2 variant B.1.1.7: cohort analysis. BMJ Clinical research ed 2021;373:n1412 ↩

89. ISARIC4C (Comprehensive Clinical Characterisation Collaboration) home page ↩

90. OECD. [Beyond Containment: Health systems responses to COVID-19 in the OECD](https://read.oecd-ilibrary.org/view/?ref=119_119689-ud5comtf84&title=Beyond_Containment:Health_systems_responses_to_COVID-19_in_the_OECDI]. April 2020 ↩

91. Armstrong RA, Kane AD, Cook TM. Outcomes from intensive care in patients with COVID-19: a systematic review and meta-analysis of observational studies. Anaesthesia 2020;75(10):1340-1349 ↩

92. Menges D, Ballouz T, Anagnostopoulos A, Aschmann HE, Domenghino A et al. Burden of post-COVID-19 syndrome and implications for healthcare service planning: A population-based cohort study. PloS one 2021;16(7):e0254523 ↩

93. NIHR. Living with COVID-19. 2020. ↩

94. NICE. COVID-19 rapid guideline: managing the long-term effects of COVID-19. 2020 ↩

95. ONS COVID-19 Infection Survey summary. Prevalence of ongoing symptoms following coronavirus (COVID-19) infection in the UK. Weekly summaries. ↩

96. Hoste L, Van Paemel R, Haerynck F. Multisystem inflammatory syndrome in children related to COVID-19: a systematic review. Eur J Pediatr. 2021 Jul;180(7):2019-2034 ↩

97. Payne AB, Gilani Z, Godfred-Cato S, et al. Incidence of Multisystem Inflammatory Syndrome in Children Among US Persons Infected With SARS-CoV-2. JAMA Netw Open. 2021;4(6):e2116420 ↩

98. Flood J, Shingleton J, Bennett E, et al. Paediatric multisystem inflammatory syndrome temporally associated with SARS-CoV-2 (PIMS-TS): Prospective, national surveillance, United Kingdom and Ireland, 2020. The Lancet Regional Health Europe. March 2021 ↩

99. Davies, N.G., Jarvis, C.I., CMMID COVID-19 Working Group. et al. Increased mortality in community-tested cases of SARS-CoV-2 lineage B.1.1.7. Nature 593, 270–274 (2021) ↩

100. Nyberg T.,et al, Comparative analysis of the risks of hospitalisation and death associated with SARS-CoV-2 omicron (B.1.1.529) and delta (B.1.617.2) variants in England: a cohort study, The Lancet,399,(2022)1303-1312 ↩

101. Horby P, Hayward A, Barclay W, Openshaw P, Edmunds J, Ferguson N, et al. NERVTAG meeting paper - Immunity Certification. 2020 Nov p. 16 ↩ ↩²

102. Kellam P, Barclay W 2020. The dynamics of humoral immune responses following SARS-CoV-2 infection and the potential for reinfection. J Gen Virol. 101(8):791–7. ↩ ↩² ↩³ ↩⁴ ↩⁵

103. Choe PG, Perera RAPM, Park WB, Song KH, Bang JH, Kim ES, et al. MERS-CoV Antibody Responses 1 Year after Symptom Onset, South Korea, 2015. Emerg Infect Dis. 2017 Jul;23(7):1079–84. ↩

104. Cao WC, Liu W, Zhang PH, Zhang F, Richardus JH. Disappearance of Antibodies to SARS-Associated Coronavirus after Recovery. N Engl J Med. 2007 Sep 13;357(11):1162–3. ↩

105. Kiyuka PK, Agoti CN, Munywoki PK, Njeru R, Bett A, Otieno JR, et al. Human Coronavirus NL63 Molecular Epidemiology and Evolutionary Patterns in Rural Coastal Kenya. J Infect Dis. 2018 May 5;217(11):1728–39. ↩

106. Callow KA, Parry HF, Sergeant M, Tyrrell DA. The time course of the immune response to experimental coronavirus infection of man. Epidemiol Infect. 1990 Oct;105(2):435–46. ↩

107. Alshukairi AN, Khalid I, Ahmed WA, Dada AM, Bayumi DT, Malic LS, et al. Antibody Response and Disease Severity in Healthcare Worker MERS Survivors. Emerg Infect Dis. 2016 Jun;22(6):1113–5. ↩

108. NERVTAG. Immunity to SARS-CoV-2 and the concept of an Immunity Certifcate (viewed on 30 March 2022) ↩

109. NERVTAG. Minutes of the NERVTAG COVID-19 Ninth Meeting: 13 March 2020[viewed on 28 March 2022] ↩

110. Openshaw P, Huntley C, Horby P, Barclay W, Siggins MK, Thwaites RS, et al. NERVTAG: Update note on immunity to SARS-CoV-2 after natural infection 2021 May, p. 14 ↩ ↩² ↩³

111. Plotkin SA, Plotkin SA. Correlates of Vaccine-Induced Immunity. Clin Infect Dis. 2008 Aug 1;47(3):401–9. ↩

112. To KKW, Tsang OTY, Leung WS, Tam AR, Wu TC, Lung DC, et al. Temporal profiles of viral load in posterior oropharyngeal saliva samples and serum antibody responses during infection by SARS-CoV-2: an observational cohort study. Lancet Infect Dis. 2020 May 1;20(5):565–74. ↩

113. Okba NMA, Müller MA, Li W, Wang C, GeurtsvanKessel CH, Corman VM, et al. SARS-CoV-2 specific antibody responses in COVID-19 patients. medRxiv; 2020 [viewed on 28 March 2022] ↩

114. Zhao J, Yuan Q, Wang H, Liu W, Liao X, Su Y, et al. Antibody Responses to SARS-CoV-2 in Patients With Novel Coronavirus Disease 2019. Clin Infect Dis. 2020 Nov 19;71(16):2027–34. ↩ ↩²

115. Lou B, Li TD, Zheng SF, Su YY, Li ZY, Liu W, et al. Serology characteristics of SARS-CoV-2 infection after exposure and post-symptom onset. Eur Respir J. 2020 Aug 1 [viewed on 28 March 2022] ↩

116. Thevarajan I, Nguyen THO, Koutsakos M, Druce J, Caly L, van de Sandt CE, et al. Breadth of concomitant immune responses prior to patient recovery: a case report of non-severe COVID-19. Nat Med. 2020 Apr;26(4):453–5. ↩

117. Bao L, Deng W, Gao H, Xiao C, Liu J, Xue J, et al. Lack of Reinfection in Rhesus Macaques Infected with SARS-CoV-2. bioRxiv; 2020 [viewed on 28 March 2022] ↩

118. Munster VJ, Feldmann F, Williamson BN, van Doremalen N, Pérez-Pérez L, Schulz J, et al. Respiratory disease in rhesus macaques inoculated with SARS-CoV-2. Nature. 2020 Sep;585(7824):268–72. ↩

119. Shrotri M, Krutikov M, Nacer-Laidi H, Azmi B, Palmer T, Giddings R, et al. Duration of vaccine effectiveness against SARS-CoV2 infection, hospitalisation, and death in residents and staff of Long-Term Care Facilities (VIVALDI): a prospective cohort study, England, Dec 2020-Dec 2021. Infectious Diseases (except HIV/AIDS); 2022 Mar [viewed on 17 June 2022] ↩

120. Krutikov M, Palmer T, Tut G, Fuller C, Shrotri M, Williams H, et al. Incidence of SARS-CoV-2 infection according to baseline antibody status in staff and residents of 100 long-term care facilities (VIVALDI): a prospective cohort study. Lancet Healthy Longev. 2021 Jun;2(6):e362–70. ↩ ↩² ↩³

121. Hall VJ, Foulkes S, Charlett A, Atti A, Monk EJM, Simmons R, et al. SARS-CoV-2 infection rates of antibody-positive compared with antibody-negative health-care workers in England: a large, multicentre, prospective cohort study (SIREN). Lancet Lond Engl. 2021;397(10283):1459–69. ↩ ↩²

122. Krutikov M, Palmer T, Donaldson A, Lorencatto F, Forbes G, Copas AJ, et al. Study Protocol: Understanding SARS-Cov-2 infection, immunity and its duration in care home residents and staff in England (VIVALDI). Wellcome Open Research; 2021 [viewed on 17 June 2022] ↩

123. Hall V. Personal communication. 2022. ↩

124. Long QX, Liu BZ, Deng HJ, Wu GC, Deng K, Chen YK, et al. Antibody responses to SARS-CoV-2 in patients with COVID-19. Nat Med. 2020 Jun;26(6):845–8. ↩

125. Post N, Eddy D, Huntley C, Schalkwyk MCI van, Shrotri M, Leeman D, et al. Antibody response to SARS-CoV-2 infection in humans: A systematic review. PLOS ONE. 2020 Dec 31;15(12):e0244126. ↩

126. To KKW, Hung IFN, Ip JD, Chu AWH, Chan WM, Tam AR, et al. Coronavirus Disease 2019 (COVID-19) Re-infection by a Phylogenetically Distinct Severe Acute Respiratory Syndrome Coronavirus 2 Strain Confirmed by Whole Genome Sequencing. Clin Infect Dis. 2021 Nov 1;73(9):e2946–51. ↩

127. Selhorst P, Van Ierssel S, Michiels J, Mariën J, Bartholomeeusen K, Dirinck E, et al. Symptomatic SARS-CoV-2 re-infection of a health care worker in a Belgian nosocomial outbreak despite primary neutralizing antibody response. Infectious Diseases (except HIV/AIDS); 2020 Nov [viewed on 28 March 2022] ↩

128. Tillett RL, Sevinsky JR, Hartley PD, Kerwin H, Crawford N, Gorzalski A, et al. Genomic evidence for reinfection with SARS-CoV-2: a case study. Lancet Infect Dis. 2021 Jan 1;21(1):52–8. ↩ ↩²

129. Bongiovanni M. COVID-19 reinfection in a healthcare worker. J Med Virol. 2021;93(7):4058–9. ↩ ↩²

130. Larson D, Brodniak SL, Voegtly LJ, Cer RZ, Glang LA, Malagon FJ, et al. A Case of Early Reinfection With Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2). Clin Infect Dis. 2021 Nov 1;73(9):e2827–8. ↩

131. Boyton RJ, Altmann DM. Risk of SARS-CoV-2 reinfection after natural infection. The Lancet. 2021 Mar 27;397(10280):1161–3. ↩

132. Lumley SF, O’Donnell D, Stoesser NE, Matthews PC, Howarth A, Hatch SB, et al. Antibody Status and Incidence of SARS-CoV-2 Infection in Health Care Workers. N Engl J Med. 2021 Feb 11;384(6):533–40. ↩ ↩² ↩³

133. Houlihan CF, Vora N, Byrne T, Lewer D, Kelly G, Heaney J, et al. Pandemic peak SARS-CoV-2 infection and seroconversion rates in London frontline health-care workers. The Lancet. 2020 Jul 25;396(10246):e6–7. ↩ ↩²

134. Addetia A, Crawford KHD, Dingens A, Zhu H, Roychoudhury P, Huang ML, et al. Neutralizing Antibodies Correlate with Protection from SARS-CoV-2 in Humans during a Fishery Vessel Outbreak with a High Attack Rate. J Clin Microbiol. 58(11):e02107-20. ↩

135. Imai M, Iwatsuki-Horimoto K, Hatta M, Loeber S, Halfmann PJ, Nakajima N, et al. Syrian hamsters as a small animal model for SARS-CoV-2 infection and countermeasure development. Proc Natl Acad Sci. 2020 Jul 14;117(28):16587–95. ↩

136. Jeffery-Smith A, Iyanger N, Williams SV, Chow JY, Aiano F, Hoschler K, et al. Antibodies to SARS-CoV-2 protect against re-infection during outbreaks in care homes, September and October 2020. Eurosurveillance. 2021 Feb 4;26(5):2100092. ↩

137. Zuo J, Dowell A, Pearce H, Verma K, Long HM, Begum J, et al. Robust SARS-CoV-2-specific T-cell immunity is maintained at 6 months following primary infection. bioRxiv; 2020 [viewed on 30 March 2022] ↩ ↩²

138. Ward H, Cooke G, Atchison C, Whitaker M, Elliott J, Moshe M, et al. Declining prevalence of antibody positivity to SARS-CoV-2: a community study of 365,000 adults. medRxiv; 2020 [viewed on 27 March 2022] ↩

139. Hall V, Foulkes S, Charlett A, Atti A, Monk EJM, Simmons R, et al. Do antibody positive healthcare workers have lower SARS-CoV-2 infection rates than antibody negative healthcare workers? Large multi-centre prospective cohort study (the SIREN study), England: June to November 2020. medRxiv; 2021 [viewed on 28 March 2022] ↩ ↩²

140. Gudbjartsson DF, Norddahl GL, Melsted P, Gunnarsdottir K, Holm H, Eythorsson E, et al. Humoral Immune Response to SARS-CoV-2 in Iceland. N Engl J Med. 2020 Oct 29;383(18):1724–34. ↩

141. NERVTAG. Minutes of the NERVTAG COVID-19 Forty-fifthmeeting: 05 February 2021. [viewed on 28 March 2022] ↩

142. Jeffery-Smith A, Rowland TAJ, Patel M, Whitaker H, Iyanger N, Williams SV, et al. Reinfection with new variants of SARS-CoV-2 after natural infection: a prospective observational cohort in 13 care homes in England. Lancet Healthy Longev. 2021 Dec;2(12):e811–9. ↩

143. Evans JP, Zeng C, Carlin C, Lozanski G, Saif LJ, Oltz EM, et al. Neutralizing antibody responses elicited by SARS-CoV-2 mRNA vaccination wane over time and are boosted by breakthrough infection. Sci Transl Med. 2022 Mar 23;14(637):eabn8057. ↩

144. Muecksch F, Wang Z, Cho A, Gaebler C, Tanfous TB, DaSilva J, et al. Increased Potency and Breadth of SARS-CoV-2 Neutralizing Antibodies After a Third mRNA Vaccine Dose. bioRxiv. 2022 Feb 15;2022.02.14.480394. ↩

145. Hiscox J, Barclay W, Evans C. NERVTAG Paper: SARS-CoV-2 variants. 2020 May [viewed on 28 March 2022] ↩ ↩²

146. Weisblum Y, Schmidt F, Zhang F, DaSilva J, Poston D, Lorenzi JC, et al. Escape from neutralizing antibodies by SARS-CoV-2 spike protein variants. Marsh M, van der Meer JW, Montefiore D, editors. eLife. 2020 Oct 28;9:e61312. ↩

147. Horby P, Bell I, Breuer J, Cevik M, Challen R, Davies N, et al. Update note on B.1.1.7 severity. 2021 Feb, p. 14 ↩

148. Public Health England. SARS-CoV-2 variants of concern and variants under investigation in England. Technical Brifeing 14. 2021 June p.66 ↩

149. Public Health England. Investigation of novel SARS-CoV-2 variant. Variant of Concern 202012/01. Technical briefing 2. 2020 Dec [viewed on 28 March 2022] ↩

150. Graham MS, Sudre CH, May A, Antonelli M, Murray B, Varsavsky T, et al. Changes in symptomatology, re-infection and transmissibility associated with SARS-CoV-2 variant B.1.1.7: an ecological study. medRxiv; 2021 [viewed on 28 March 2022]. p. 2021.01.28.21250680 ↩ ↩²

151. Information on COVID-19 reinfection surveillance in England. GOV.UK. [viewed on 28 March 2022] ↩

152. Public Health England. SARS-CoV-2 variants of concern and variants under investigation in England. Technical briefing 19. 2021 Jul [viewed on 28 March 2022] ↩

153. UKHSA. SARS-CoV-2 variants of concern and variants under investigation in England. Technical briefing 33. 2021 Dec [viewed on 28 March 2022] p. 42 ↩

154. NERVTAG. Minutes of the NERVTAG Wuhan Novel Coronavirus Second Meeting: 21 January 2020 2020 Jan [viewed on 29 March 2022] ↩ ↩² ↩³

155. Precautionary SAGE 1 minutes: Coronavirus (COVID-19) response, 22 January 2020[viewed on 29 March 2022] ↩ ↩² ↩³

156. Statement from the UK Chief Medical Officers on an update to coronavirus symptoms: 18 May 2020[viewed on 25 March 2022] ↩

157. Hayward A. NERVTAG Paper: Case definitions for contact tracing, 7 May 2020[viewed on 25 March 2022] ↩ ↩² ↩³

158. Sohal. Change covid case definition. BMJ. 2020 Dec 21 [viewed on 11 March 2022] ↩

159. Crozier A, Dunning J, Rajan S, Semple MG, Buchan IE. Could expanding the covid-19 case definition improve the UK’s pandemic response? BMJ. 2021 Jul 1;374:n1625. ↩

160. NERVTAG. Minutes of the NERVTAG Wuhan Novel Coronavirus Sixth Meeting: 07 February 2020 (viewed on 29 March 2022) ↩ ↩²

161. NERVTAG. NERVTAG: Community case definitions for Covid-19, September 2020(viewed on 29 March 2022) ↩

162. Eyre MT, Burns R, Kirkby V, Smith C, Denaxas S, Nguyen V, et al. Impact of baseline cases of cough and fever on UK COVID-19 diagnostic testing rates: estimates from the Bug Watch community cohort study. medRxiv; 2020. p. 2020.09.03.20187377 (viewed on 6 April 2022) ↩

163. Zhu N, Zhang D, Wang W, Li X, Yang B, Song J, et al. A Novel Coronavirus from Patients with Pneumonia in China, 2019. N Engl J Med. 2020 Feb 20;382(8):727–33. ↩ ↩²

164. Tang JW, Tambyah PA, Hui DSC. Emergence of a novel coronavirus causing respiratory illness from Wuhan, China. J Infect. 2020 Mar 1;80(3):350–71. ↩ ↩²

165. World Health Organization. Pneumonia of unknown cause – China (viewed on 29 March) ↩ ↩²

166. World Health Organization. Listings of WHO’s response to COVID-19 (viewed on 29 March) ↩

167. World Health Organization. Surveillance case definitions for human infection with novel coronavirus (nCoV): interim guidance, 11 January 2020 (viewed on 29 March) ↩

168. World Health Organization. Statement on the meeting of the International Health Regulations (2005) Emergency Committee regarding the outbreak of novel coronavirus 2019 (n-CoV) on 23 January 2020 (viewed on 1 April 2022) ↩

169. World Health Organization. Statement on the second meeting of the International Health Regulations (2005) Emergency Committee regarding the outbreak of novel coronavirus (2019-nCoV) (viewed on 1 April 2022) ↩

170. SAGE 5 minutes: Coronavirus (COVID-19) response, 6 February 2020 (viewed on 29 March 2022) ↩

171. NERVTAG. Minutes of the NERVTAG Wuhan Novel Coronavirus Fourth Meeting: 30 January 2020 (viewed on 29 March 2022) ↩

172. World Health Organization. The first few X cases and contacts (‎FFX)‎ investigation protocol for coronavirus disease 2019 (‎COVID-19)‎, version 2.2 (viewed on 25 March 2022) ↩

173. SPI-B. Symptom-based contact tracing is likely to reduce adherence to advice to quarantine in comparison to test-based approaches: 29 April 2020 (viewed on 29 March 2022) ↩

174. European Centre for Disease Prevention and Control. COVID-19 surveillance guidance. Transition form COVID-19 emergency surveillance to routine surveillance of respiratory pathogens. 2021 Oct;13. ↩

175. Rutter, H., et al. Visualising SARS-CoV-2 transmission routes and mitigations. Bmj, 2021. 375: p. e065312. ↩ ↩²

176. PHE Transmission Group Factors contributing to risk of SARS-CoV2 transmission associated with various settings - November presented at SAGE 70 (viewed on 8 March 2022) ↩ ↩² ↩³ ↩⁴

177. Morawska, L. and D.K. Milton. It Is Time to Address Airborne Transmission of Coronavirus Disease 2019 (COVID-19). Clin Infect Dis, 2020. 71(9): p. 2311-2313 ↩ ↩²

178. Huang, C., et al. Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China. Lancet, 2020. 395(10223): p. 497-506. ↩ ↩² ↩³

179. Hayward, A., et al. Public activities preceding the onset of acute respiratory infection syndromes in adults in England - implications for the use of social distancing to control pandemic respiratory infections. [version 1; peer review: 2 approved]. 2020. 5(54). ↩

180. Warren-Gash, C., E. Fragaszy, and A.C. Hayward. Hand hygiene to reduce community transmission of influenza and acute respiratory tract infection: a systematic review. Influenza and other respiratory viruses, 2013. 7(5): p. 738-749 ↩

181. Li, Y., et al. Probable airborne transmission of SARS-CoV-2 in a poorly ventilated restaurant. Building and environment, 2021. 196: p. 107788-107788. ↩

182. Lu, J., et al. COVID-19 Outbreak Associated with Air Conditioning in Restaurant, Guangzhou, China, 2020. Emerg Infect Dis, 2020. 26(7): p. 1628-1631. ↩ ↩² ↩³

183. Zheng, R., et al. Spatial transmission of COVID-19 via public and private transportation in China. Travel Med Infect Dis, 2020. 34: p. 101626. ↩

184. Zhao, S., et al. The association between domestic train transportation and novel coronavirus (2019-nCoV) outbreak in China from 2019 to 2020: A data-driven correlational report. Travel Med Infect Dis, 2020. 33: p. 101568. ↩

185. Rothe, C., et al. Transmission of 2019-nCoV Infection from an Asymptomatic Contact in Germany. N Engl J Med, 2020. 382(10): p. 970-971. ↩

186. Chan, J.F., et al. A familial cluster of pneumonia associated with the 2019 novel coronavirus indicating person-to-person transmission: a study of a family cluster. Lancet, 2020. 395(10223): p. 514-523. ↩

187. Li, Q., et al. Early Transmission Dynamics in Wuhan, China, of Novel Coronavirus-Infected Pneumonia. N Engl J Med, 2020. 382(13): p. 1199-1207 ↩

188. Public Health England “The First Few Hundred (FF100)” Enhanced Case and Contact Protocol v12, 2016 (viewed on 13 April 2022) ↩

189. Ong, S.W.X., et al. Air, Surface Environmental, and Personal Protective Equipment Contamination by Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) From a Symptomatic Patient. JAMA, 2020. 323(16): p. 1610-1612. ↩ ↩²

190. World Health Organization. Report of the WHO-China Joint Mission on Coronavirus Disease 2019 (COVID-19) 16-24 February 2020 (viewed on 8 March 2022) ↩ ↩² ↩³

191. World Health Organization. Modes of Transmission of virus causing COVID-19: implications for IPC precaution recommendations (viewed on 8 March 2022) ↩ ↩²

192. NERVTAG and Environmental Modelling Group. Evidence of environmental dispersion of COVID-19 for different mechanisms, 14 April 2020. ↩

193. Environmental Modelling Group of SAGE. Environmental influence on transmission - 12th May. 2020. ↩

194. Environmental Modelling Group of SAGE. Transmission of SARS-CoV-2 and Mitigating Measures - 4 June 2020 (viewed on 8 March 2022) ↩

195. Review of two metre social distancing guidance: Summary of review findings. 2020 (viewed on 8 March 2022) ↩

196. World Health Organization. Transmission of SARS-CoV-2: implications for infection prevention precautions: Scientific Brief - July 2020 (viewed on 8 March 2022) ↩ ↩²

197. Pastorino, B., et al. Prolonged Infectivity of SARS-CoV-2 in Fomites. Emerg Infect Dis, 2020. 26(9): p. 2256-7. ↩

198. Wilson, A.M., et al. Modeling COVID-19 infection risks for a single hand-to-fomite scenario and potential risk reductions offered by surface disinfection. American journal of infection control, 2021. 49(6): p. 846-848. ↩ ↩²

199. Chang, L., et al. Severe Acute Respiratory Syndrome Coronavirus 2 RNA Detected in Blood Donations. Emerg Infect Dis, 2020. 26(7): p. 1631-1633. ↩

200. Zou, L., et al. SARS-CoV-2 Viral Load in Upper Respiratory Specimens of Infected Patients. 2020. 382(12): p. 1177-1179 ↩

201. Li, Y.Y., J.X. Wang, and X. Chen. Can a toilet promote virus transmission? From a fluid dynamics perspective. Phys Fluids (1994), 2020. 32(6): p. 065107. ↩

202. Kim, Y.-I., et al. Infection and Rapid Transmission of SARS-CoV-2 in Ferrets. Cell Host & Microbe, 2020. 27(5): p. 704-709.e2. ↩

203. Tang, S., et al. Aerosol transmission of SARS-CoV-2? Evidence, prevention and control. Environment international, 2020. 144: p. 106039-106039. ↩

204. Greenhalgh, T., et al. Ten scientific reasons in support of airborne transmission of SARS-CoV-2. Lancet (London, England), 2021. 397(10285): p. 1603-1605. ↩

205. Hamner L, et al. High SARS-CoV-2 Attack Rate Following Exposure at a Choir Practice — Skagit County, Washington, March 2020. MMWR Morb Mortal Wkly Rep. 2020;69:606-1. 2020 ↩ ↩² ↩³

206. Jang, S., S.H. Han, and J.-Y. Rhee. Cluster of Coronavirus Disease Associated with Fitness Dance Classes, South Korea. Emerging infectious diseases, 2020. 26(8): p. 1917-1920. ↩ ↩² ↩³

207. Groves, L.M., et al. Community Transmission of SARS-CoV-2 at Three Fitness Facilities - Hawaii, June-July 2020. MMWR Morb Mortal Wkly Rep, 2021. 70(9): p. 316-320. ↩ ↩² ↩³

208. Eichler, N., et al. Transmission of Severe Acute Respiratory Syndrome Coronavirus 2 during Border Quarantine and Air Travel, New Zealand (Aotearoa). Emerg Infect Dis, 2021. 27(5): p. 1274-1278. ↩

209. Klompas, M., et al. A SARS-CoV-2 Cluster in an Acute Care Hospital. Ann Intern Med, 2021. 174(6): p. 794-802. ↩

210. Kutter, J.S., et al. SARS-CoV and SARS-CoV-2 are transmitted through the air between ferrets over more than one meter distance. Nat Commun, 2021. 12(1): p. 1653. ↩

211. van Doremalen, N., et al. Aerosol and Surface Stability of SARS-CoV-2 as Compared with SARS-CoV-1. 2020. 382(16): p. 1564-1567. ↩

212. Somsen, G.A., et al. Small droplet aerosols in poorly ventilated spaces and SARS-CoV-2 transmission. Lancet Respir Med, 2020. 8(7): p. 658-659. ↩

213. Chen, W., et al. Short-range airborne route dominates exposure of respiratory infection during close contact. Building and Environment, 2020. 176: p. 106859. ↩

214. Heneghan, C., et al. SARS-CoV-2 and the role of airborne transmission: a systematic review [version 2; peer review: 1 approved with reservations, 2 not approved]. 2021. 10(232). ↩

215. Freeman, A.L., et al. Expert elicitation on the relative importance of possible SARS-CoV-2 transmission routes and the effectiveness of mitigations. BMJ Open, 2021. 11(12): p. e050869. ↩

216. O’Connell, J.J. Nontuberculous respiratory infections among the homeless. Semin Respir Infect, 1991. 6(4): p. 247-53. ↩

217. Lambert, L.A., et al. Tuberculosis in Jails and Prisons: United States, 2002-2013. American journal of public health, 2016. 106(12): p. 2231-2237. ↩

218. Tsang, T.K., et al. Household Transmission of Influenza Virus. Trends Microbiol, 2016. 24(2): p. 123-133. ↩

219. Jackson, C., et al. School closures and influenza: systematic review of epidemiological studies. 2013. 3(2): p. e002149. ↩

220. Vanhems, P., T. Bénet, and E. Munier-Marion. Nosocomial influenza: encouraging insights and future challenges. Curr Opin Infect Dis, 2016. 29(4): p. 366-72. ↩

221. Rocklöv, J., H. Sjödin, and A. Wilder-Smith. COVID-19 outbreak on the Diamond Princess cruise ship: estimating the epidemic potential and effectiveness of public health countermeasures. J Travel Med, 2020. 27(3). ↩

222. Hayward, A. and S. Hoskins Relative importance of different non household activities for COVID-19 transmission during period of intense restrictions compared to period of no restrictions (viewed on 11 April 2022) ↩ ↩²

223. Aldridge, R.W., et al. Household overcrowding and risk of SARS-CoV-2: analysis of the Virus Watch prospective community cohort study in England and Wales. 2021: p. 2021.05.10.21256912. ↩ ↩²

224. Hayward, A. Impact of occupational exposure to disease, proximity to others during work and income on mortality from COVID-19. 2020 (viewed on 11 April 2022) ↩ ↩² ↩³

225. Nafilyan, V., et al. Occupation and COVID-19 mortality in England: a national linked data study of 14.3 million adults. 2021: p. 2021.05.12.21257123. ↩

226. Neumann, G., T. Noda, and Y. Kawaoka. Emergence and pandemic potential of swine-origin H1N1 influenza virus. Nature, 2009. 459(7249): p. 931-939. ↩

227. Hayman, D.T.S., et al. Global importation and population risk factors for measles in New Zealand: a case study for highly immunized populations. Epidemiol Infect, 2017. 145(9): p. 1875-1885. ↩

228. Williamson, E.J., et al. Risks of covid-19 hospital admission and death for people with learning disability: population based cohort study using the OpenSAFELY platform. 2021. 374: p. n1592. ↩

229. Williamson, E.J., et al. Factors associated with COVID-19-related death using OpenSAFELY. Nature, 2020. 584(7821): p. 430-436. ↩

230. Nervtag and Environmental Modelling Group. Insights on transmission of COVID-19 with a focus on the hospitality, retail and leisure sector. 2021 (viewed on 10 March 2022) ↩ ↩²

231. Environmental Modelling Group of SAGE COVID-19 Risk by Occupation and Workplace, 11 February 2021 (viewed on 7 October 2022) ↩

232. Hosseini, P., et al. Transmission and Control of SARS-CoV-2 in the Food Production Sector: A Rapid Narrative Review of the Literature. 2022. 19(19): p. 12104. ↩

233. Rutter, H., et al. Visualising SARS-CoV-2 transmission routes and mitigations. Bmj, 2021. 375: p. e065312. ↩

234. Ismail, S.A., et al. SARS-CoV-2 infection and transmission in educational settings: a prospective, cross-sectional analysis of infection clusters and outbreaks in England. Lancet Infect Dis, 2021. 21(3): p. 344-353. ↩

235. C. Fraser, S. Riley, R. M. Anderson, and N. M. Ferguson. Factors that make an infectious disease outbreak controllable ↩

236. A. L. Rasmussen and S. V. Popescu. SARS-CoV-2 transmission without symptoms ↩ ↩² ↩³

237. E. A. Meyerowitz, A. Richterman, I. I. Bogoch, N. Low, and M. Cevik, ‘Towards an accurate and systematic characterisation of persistently asymptomatic infection with SARS-CoV-2’, Lancet Infect. Dis., vol. 21, no. 6, pp. e163–e169, Jun. 2021, doi: 10.1016/S1473-3099(20)30837-9. ↩

238. A. Hayward and P. Horby. NERVTAG Paper: Asymptomatic SARS-CoV-2 Infection, May 2020 (viewed on 29 March 2022) ↩ ↩² ↩³

239. W. Gao, J. Lv, Y. Pang, and L.-M. Li. Role of asymptomatic and pre-symptomatic infections in covid-19 pandemic BMJ, vol. 375, p. n2342, Dec. 2021 ↩

240. C. Rothe et al. Transmission of 2019-nCoV Infection from an Asymptomatic Contact in Germany N. Engl. J. Med., vol. 382, no. 10, pp. 970–971, Mar. 2020 ↩ ↩²

241. Y. Bai et al. Presumed Asymptomatic Carrier Transmission of COVID-19 JAMA, vol. 323, no. 14, pp. 1406–1407, Apr. 2020 ↩ ↩²

242. S. Beale, A. Hayward, L. Shallcross, R. W. Aldridge, and E. Fragaszy. A rapid review and meta-analysis of the asymptomatic proportion of PCR-confirmed SARS-CoV-2 infections in community settings Wellcome Open Research, Nov. 05, 2020 ↩ ↩²

243. Public Health England PHE enhanced surveillance of household contacts: interim analysis’, July 2020 (viewed on 29 March 2022) ↩ ↩²

244. M. M. Arons et al. Presymptomatic SARS-CoV-2 Infections and Transmission in a Skilled Nursing Facility N. Engl. J. Med., vol. 382, no. 22, pp. 2081–2090, May 2020 ↩ ↩²

245. H. Taylor et al. Cross sectional investigation of a COVID-19 outbreak at a London Army barracks: Neutralising antibodies and virus isolation Lancet Reg. Health – Eur., vol. 2, Mar. 2021 ↩ ↩² ↩³

246. A. C. Roxby. Detection of SARS-CoV-2 Among Residents and Staff Members of an Independent and Assisted Living Community for Older Adults — Seattle, Washington, 2020 MMWR Morb. Mortal. Wkly. Rep., vol. 69, 2020 ↩ ↩²

247. J. F.-W. Chan et al. A familial cluster of pneumonia associated with the 2019 novel coronavirus indicating person-to-person transmission: a study of a family cluster The Lancet, vol. 395, no. 10223, pp. 514–523, Feb. 2020 ↩

248. L. Luo et al. Modes of contact and risk of transmission in COVID-19 among close contacts medRxiv, p. 2020.03.24.20042606, Mar. 26, 2020 ↩

249. PHE Virology Cell Are asymptomatic people with 2019nCoV infectious? ↩

250. K. Mizumoto, K. Kagaya, A. Zarebski, and G. Chowell. Estimating the asymptomatic proportion of coronavirus disease 2019 (COVID-19) cases on board the Diamond Princess cruise ship, Yokohama, Japan, 2020 Eurosurveillance, vol. 25, no. 10, p. 2000180, Mar. 2020 ↩

251. S. Y. Park et al.Coronavirus Disease Outbreak in Call Center, South Korea - Volume 26, Number 8, August 2020 - Emerging Infectious Diseases journal - CDC ↩

252. K. Danis et al. Cluster of Coronavirus Disease 2019 (COVID-19) in the French Alps, February 2020 Clin. Infect. Dis., vol. 71, no. 15, pp. 825–832, Jul. 2020 ↩

253. T. P. Baggett, H. Keyes, N. Sporn, and J. M. Gaeta. Prevalence of SARS-CoV-2 Infection in Residents of a Large Homeless Shelter in Boston JAMA, vol. 323, no. 21, pp. 2191–2192, Jun. 2020 ↩

254. S. N. Ladhani et al. Investigation of SARS-CoV-2 outbreaks in six care homes in London, April 2020 EClinicalMedicine, vol. 26, p. 100533, September 2020 ↩ ↩²

255. R. Wölfel et al. Virological assessment of hospitalized patients with COVID-2019 Nature, vol. 581, no. 7809, Art. no. 7809, May 2020 ↩

256. NERVTAG. Minutes of the NERVTAG Fifteenth Meeting: 24 April 2020 (viewed on 29 March 2022) ↩

257. M. Cevik, M. Tate, O. Lloyd, A. E. Maraolo, J. Schafers, and A. Ho. SARS-CoV-2, SARS-CoV, and MERS-CoV viral load dynamics, duration of viral shedding, and infectiousness: a systematic review and meta-analysis Lancet Microbe, vol. 2, no. 1, pp. e13–e22, Jan. 2021 ↩

258. Public Health England. Understanding cycle threshold (Ct) in SARS-CoV-2 RT-PCR:12. ↩

259. Ke R, Martinez PP, Smith RL, Gibson LL, Mirza A, Conte M, et al. Daily sampling of early SARS-CoV-2 infection reveals substantial heterogeneity in infectiousness medRxiv; 2021p. 2021.07.12.21260208 (viewed on 31 March 2022) ↩ ↩² ↩³ ↩⁴

260. Singanayagam A, Hakki S, Dunning J, Madon KJ, Crone MA, Koycheva A, et al. Community transmission and viral load kinetics of the SARS-CoV-2 delta (B.1.617.2) variant in vaccinated and unvaccinated individuals in the UK: a prospective, longitudinal, cohort study. Lancet Infect Dis. 2022 Feb 1;22(2):183–95. ↩ ↩² ↩³

261. Public Health England. Supplementary Data June 2020: Analysis of virus isolation data (viewed on 31 March 2022) ↩ ↩² ↩³ ↩⁴ ↩⁵

262. van Kampen JJA, van de Vijver DAMC, Fraaij PLA, Haagmans BL, Lamers MM, Okba N, et al. Duration and key determinants of infectious virus shedding in hospitalized patients with coronavirus disease-2019 (COVID-19). Nat Commun. 2021 Jan 11;12(1):267. ↩ ↩² ↩³

263. Evans C, Barclay W, Zambon M, Horby P, Hiscox J. NERVTAG: : Dynamics of infectiousness and antibody responses. 2020 Jun (viewed on 31 March 2022) ↩ ↩²

264. Cevik M, Tate M, Lloyd O, Maraolo AE, Schafers J, Ho A. SARS-CoV-2, SARS-CoV, and MERS-CoV viral load dynamics, duration of viral shedding, and infectiousness: a systematic review and meta-analysis. Lancet Microbe. 2021 Jan;2(1):e13–22. ↩ ↩²

265. NERVTAG. Minutes of the NERVTAG Twelfth Meeting: 03 April 2020 (viewed on 29 March 2022) ↩ ↩²

266. Lyngse FP, Mølbak K, Franck KT, Nielsen C, Skov RL, Voldstedlund M, et al. Association between SARS-CoV-2 Transmissibility, Viral Load, and Age in Households medRxiv; 2021 p. 2021.02.28.21252608 [viewed on 27 March 2022] ↩

267. Peiris J, Chu C, Cheng V, Chan K, Hung I, Poon L, et al. Clinical progression and viral load in a community outbreak of coronavirus-associated SARS pneumonia: a prospective study. The Lancet. 2003 May 24;361(9371):1767–72. ↩

268. Bermingham A, Heinen P, Iturriza-Gómara M, Gray J, Appleton H, Zambon MC. Laboratory diagnosis of SARS. Philos Trans R Soc Lond B Biol Sci. 2004 Jul 29;359(1447):1083–9. ↩

269. Min CK, Cheon S, Ha NY, Sohn KM, Kim Y, Aigerim A, et al. Comparative and kinetic analysis of viral shedding and immunological responses in MERS patients representing a broad spectrum of disease severity. Sci Rep. 2016 May 5;6(1):25359. ↩

270. Oh M don, Park WB, Choe PG, Choi SJ, Kim JI, Chae J, et al. Viral Load Kinetics of MERS Coronavirus Infection. N Engl J Med. 2016 Sep 29;375(13):1303–5. ↩

271. Young BE, Ong SWX, Kalimuddin S, Low JG, Tan SY, Loh J, et al. Epidemiologic Features and Clinical Course of Patients Infected With SARS-CoV-2 in Singapore. JAMA. 2020 Apr 21;323(15):1488–94. ↩

272. Lan L, Xu D, Ye G, Xia C, Wang S, Li Y, et al. Positive RT-PCR Test Results in Patients Recovered From COVID-19. JAMA. 2020 Apr 21;323(15):1502–3. ↩

273. NERVTAG. NERVTAG Paper: Duration of infectiousness following symptom onset in COVID. 2020 Apr (viewed on 31 March 2022) ↩

274. Arons MM, Hatfield KM, Reddy SC, Kimball A, James A, Jacobs JR, et al. Presymptomatic SARS-CoV-2 Infections and Transmission in a Skilled Nursing Facility. N Engl J Med. 2020 May 28;382(22):2081–90. ↩

275. Wölfel R, Corman VM, Guggemos W, Seilmaier M, Zange S, Müller MA, et al. Virological assessment of hospitalized patients with COVID-2019. Nature. 2020 May;581(7809):465–9. ↩

276. Kujawski SA, Wong KK, Collins JP, Epstein L, Killerby ME, Midgley CM, et al. Clinical and virologic characteristics of the first 12 patients with coronavirus disease 2019 (COVID-19) in the United States. Nat Med. 2020 Jun;26(6):861–8. ↩

277. Cevik M, Kuppalli K, Kindrachuk J, Peiris M. Virology, transmission, and pathogenesis of SARS-CoV-2. BMJ. 2020 Oct 23;371:m3862. ↩

278. Public Health England. SARS-CoV-2 variants of concern and variants under investigation in England Technical briefing 17. 2021 Jun p. 69 ↩

279. Public Health England. SARS-CoV-2 variants of concern and variants under investigation in England. Technical briefing 20. 2021 Aug p. 44 ↩