Independent report

Chapter 6: testing

Updated 10 January 2023

Introduction

In pandemics and major epidemics, the development of a test or tests and their scaling up is often a rate-limiting step to:

• optimising clinical care
• deploying control measures
• developing a clear epidemiological picture
• assessing community-based countermeasures

This was true with HIV and Ebola virus, for example. Testing technology evolves, but the centrality of developing and scaling tests will remain. This is particularly true where an infection has non-specific symptoms, or can be asymptomatic, both of which were the case for SARS-CoV-2. In COVID-19, tests were developed rapidly, but the time taken to scale up limited the response in the early phases. COVID-19 was over time notable for widespread use of self-testing for an acute infection. As with all tests accurate, rapid reporting systems and clear use cases were as important as the test technology.

This chapter explores the test technologies needed, the testing strategies deployed and the scale and speed of the systems required to deliver those strategies in this pandemic. It sets out some of the most important innovations and approaches in testing from this pandemic, including mass population symptomatic and asymptomatic testing with the use of lateral flow devices (LFDs) which helped people manage risk in day-to-day activities, the widespread use of self-testing and the unprecedented scale of testing operations reaching across the UK.

It also explores the challenges in scaling up testing – a particular issue in the initial stages of the pandemic – evaluating and assessing new technologies and formulating appropriate testing strategies for different stages of the pandemic.

Broadly, testing types and the use cases they supported in this pandemic were as follows:

1. Diagnostic testing, including in the general population, helped guide clinical care and infection prevention and control in high-risk settings. It also enabled targeted isolation guidance and contact tracing to flag linked cases, supporting cluster and outbreak identification.

2. Asymptomatic testing enabled case-finding for outbreak management and routine detection of asymptomatic or pre-symptomatic cases for higher-risk settings, for those at higher risk of severe outcomes and for some critical infrastructure roles such as health and social care staff. It also supported risk management for a number of day-to-day activities across the population (such as for international travel or attending events), and data from asymptomatic testing supported surveillance – for example, at borders. Scientific consensus on the effectiveness of asymptomatic testing evolved in the early stages of the pandemic. Initially, it wasn’t clear if asymptomatic people had a lower viral load and how this might impact sensitivity and specificity – but this question was resolved scientifically quite early.

3. Large-scale diagnostic testing across the population enabled monitoring of prevalence and spread of infections at an unprecedented level. Surveying the virus genome from affected patients helped inform the clinical and public health measures needed to minimise severe disease.

4. Wastewater testing provided signals on the presence or absence of cases and variants in an area or setting. This was most useful when case, variant or testing rates were very low – for example, where a variant was newly detected in an area or to assess whether local community infections had ended. Results were also relatively fast, so offered a good chance to triangulate signals with other sources in advance of surveillance programmes such as the Office for National Statistics (ONS) Coronavirus (COVID-19) Infection Survey (CIS).

5. Testing for research enabled us to track outcomes within a range of studies – for example, on vaccine effectiveness in preventing severe disease or on rates of reinfection. Linked data from mass community testing was also key to research – for example, supporting calculation of the infection fatality rate – which in turn supported projections of potential case hospitalisation and fatality rates.

Of course, testing strategies had a number of broader aims beyond this, for example keeping sectors of the economy open or building trust with parents when schools reopened to enable them to manage potential risks. Political leaders had to consider these wider aims when taking decisions.

To support such aims, there was a need throughout the pandemic for:

• accurate and reliable tests (the technology) to determine both current and previous infection (though testing for current infection made up the bulk of demand) – there was a continual balance in selecting technologies between sensitivity and specificity, and turnaround times from sample to result
• a testing network to deliver those tests at scale and speed, analyse them, process and return results to individuals and/or professionals as needed
• scaled-up genomic sequencing to monitor new and emerging variants of concern (VOCs)
• summary data from testing to inform clinical and wider health system management, outbreak management and public health response at a local and national level, and policy decisions
• ongoing review of testing systems and strategies to continually adapt processes, strategies, communications and technologies. In practice, evaluations often focused on acceptability, uptake and outputs of testing (such as number of cases identified). It was difficult to formally assess the public health impact of testing in the rapidly changing context of a multi-wave pandemic, though there is little doubt that testing was an important part of pandemic response that provided surveillance, enabled interventions like isolation of cases, supported research and helped guide pandemic control strategies

Some of these elements were less developed at the outset of the pandemic – most notably the ability to rapidly scale systems to support widespread community testing. For COVID-19 the initial development and validation of molecular tests was very rapid by historical standards, but there were delays scaling up to mass community testing at the outset of the pandemic and this was of critical importance.

Timeline of testing

Throughout the pandemic, the capacity and effectiveness of laboratory processing, delivery and distribution routes and global demand and supply of materials continually changed. Testing strategies were continually adapted in response, and as the epidemiology changed and wider pandemic strategies also adjusted (for example, where routine testing enabled strategies supporting the labour market).^{[footnote 1]} Testing strategies also evolved as new technologies became available and as evidence emerged on the potential needs, use cases and population responses to different testing options – such as self-testing, as opposed to that undertaken by a health professional or in clinical settings only, or accessibility of public testing centres. Testing evaluation initiatives were important throughout in understanding this and helped shape government policy.

Early 2020: targeted testing

As the first few hundred cases reached the UK, testing of symptomatic patients was central to refining the clinical case definition, confirming clinical diagnoses, and conducting epidemiological studies to understand the speed and extent of the transmission to inform public health control measures. Diagnostic polymerase chain reaction (PCR) tests, processed by existing lab infrastructure, were primarily used in hospitals for case finding and early outbreak management, to prevent incursions of SARS-CoV-2 into healthcare settings, and for infection prevention and control in clinical settings. Some genomic sequencing was available from an early stage and, following the research funding provided to the COVID-19 Genomics UK consortium (COG-UK), enabled retrospective analysis of incursions into the UK.^{[footnote 2]}

In these early weeks, when we had a relatively small number of imported cases to test, existing capacity using the rapidly developed reverse transcription PCR (RT-PCR) test in existing lab arrangements was sufficient for diagnostics for those meeting early clinical and epidemiological case definition in the community. At this early stage, rapid expansion of local testing in clinical settings would have been helpful in identifying whether any further high-risk patients had COVID-19 as well as giving early signals on the possible proportion of asymptomatic infection and transmission, though this relied on local testing capacity and available resources.

The early days of the pandemic brought much academic, practitioner and media attention about the reliability and performance characteristics of new diagnostics. It was essential to have robust validation and verification data to support the mass deployment of new tests. This was both to ensure clinical and public confidence in the use and reliability of any test but also to minimise the necessity for further development work, lest the test prove to be insufficiently robust when field tested.

Spring 2020: widespread community transmission and testing scale-up

Once community transmission rose steeply in the UK in early spring 2020, community cases soon outstripped the supply of tests and existing systems were not capable of the rapid scale needed to meet demand. We anticipate this may well be a repeated problem in future pandemics and epidemics.

There was a need for rapid scaling up of capacity and wider infrastructure to enable high throughput of (particularly diagnostic) tests. Eventually this was achieved but there was a period in which testing supply did not meet demand and this was a rate-limiting factor for a number of interventions.^{[footnote 3]} This difficulty scaling existing systems was for several reasons, including:

• the limited size of the pre-existing diagnostic industry (which was not the case in all comparable countries, some of which were able to scale more quickly)
• the fact that pre-existing testing systems used multiple small labs with multiple platforms and space constraints

Although they had an expert workforce, many smaller labs also faced difficulties rapidly expanding the workforce.

At the same time, global testing supplies (particularly swabs and reagents) were significantly impacted by both increasing demand and reduced production in spring 2020 as the epidemic spread more widely, including to regions producing test materials. This was exacerbated by the fact that testing platforms were previously only validated for certain swab types, so it was difficult to flex to alternative supply routes when existing supplies were disrupted. Rising demand also put extreme pressure on existing testing systems in many countries, particularly where these were not set up in a resilient way enabling rapid scale-up. This should be anticipated in any future pandemic.

These pressures meant that early in the first wave, testing capacity was limited, and there was a need to prioritise testing. In the UK this prioritisation focused on:

• clinical care
• key workers
• vulnerable settings such as hospitals
• outbreaks in care homes, prisons and immigration and detention centres
• selected key studies to inform policy or clinical practice (such as the ‘Easter 6’ study on care homes in early April 2020)

Testing strategy was to support infection prevention and control in settings with vulnerable groups and to ensure essential services kept running by reducing ingress of COVID-19 while allowing those with COVID-19 symptoms to confirm if they needed to self-isolate. There were of course many further uses that were not implemented at this stage of limited testing capacity.

There were simultaneously major efforts to expand systems and infrastructure to provide community testing at scale in spring 2020. Many offers were made by individual sequencing facilities or staff in laboratories. There was considerable expertise, existing workforce and technology across multiple smaller labs in universities, research institutes and the NHS, and many of these labs came forward to offer help. However, without a full and integrated system of testing and reporting and quality control mechanisms, using many such smaller facilities did not easily provide a solution to delivering rapidly scaled and integrated mass testing. It was also important to protect the resources and workforce in NHS and Public Health England (PHE) labs so that testing in clinical settings and other necessary ongoing testing was not impacted by expansion of community testing. Therefore a cross-UK, centrally funded testing network was established for mass community testing alongside a shared testing operations function.

By late April 2020 the COVID-19 Testing Delivery Programme had been stood up to provide mass community testing across the UK with a single shared IT system and end-to-end processes to manage the delivery and processing of tests and results at an unprecedented scale. Large ‘lighthouse’ laboratories were set up with the support of private sector and academic partners in the diagnostics industry and existing laboratory staff and experts to provide high throughput test processing at speed. Regional and mobile testing sites were set up for community testing and a digital infrastructure was created to track and locate tests and communicate results (which were linked to existing NHS records). It was important – and will likely remain so in the future – to link national infrastructure back to local teams. For example, regional and mobile testing sites needed to be set up in a way that gave access to testing for all communities and did not exacerbate inequalities in public health and healthcare use.

On 23 March 2020, testing in the community and study-based testing (often referred to as pillar 2, under the COVID-19 Testing Delivery Programme) tested 23 samples per day. By late April, testing capacity exceeded 100,000 tests a day and continued to expand throughout the pandemic. In the month of December 2021 alone the UK laboratory network processed over 13 million samples.^{[footnote 4]}

This was a major undertaking, particularly bearing in mind the processes, networks and skills to set this up were not already in place and so needed to be brought together at pace. The COVID-19 Testing Delivery Programme operated in a unique position across the UK as a service provider, service commissioner, product procurer and, while not the actual manufacturer, the legal device manufacturer for some products.

The following core capabilities were needed to deliver effective testing at scale across the UK:

• product development: progressing concepts from idea generation through to market entry at pace
• product management: ability to manage and launch products following an iterative approach with defined release dates, and ability to add and remove products from a system seamlessly
• technical validation and evaluation of testing technologies
• high throughput lab capacity to deliver low-cost testing, which end point PCR helped achieve – it is helpful to have this lab capacity sufficiently flexible to change application to new pathogens or use new assays and equipment
• a digital platform for ordering and reporting tests – this needed to be rapidly designed and coordinated with digital partners, ensuring platforms remain up to date for latest policy requirements
• access to a national distribution network to move sensitive tests or samples around the country (including from homes and hospitals) and across the UK labs network
• supply chain and logistics expertise about both international and national inbound and outbound supplies, including supply chain planning, business operations and manufacturing – industry leaders brought valuable expertise here
• ability to operationalise policy at pace, leveraging public or private partnerships where beneficial

As a result of this scaling up in testing, from May 2020 onwards symptomatic diagnostic testing was widened from clinical and keyworker testing to the general public, though there were points at which peaks in demand and operational issues resulted in delays processing tests. As PCR capacity expanded, testing strategies were adapted – for example:

• widening criteria for diagnostic testing to anyone in the community with key symptoms
• regular asymptomatic testing in high-risk settings, such as for staff in care homes and healthcare settings

Although efforts to scale up testing were unprecedented and constituted a major achievement, there were critical months in which testing capacity did not meet demand and in which testing capacity limited options for a number of strategies and interventions – and this bears consideration for future preparedness in being able to rapidly scale up testing systems if needed and appropriate.

Alongside cross-UK community testing at scale in the cross-UK laboratory network, each nation also maintained its own testing capacity, which was predominantly in NHS labs but flexibly used for both clinical and non-clinical settings as needed. In Scotland, for example, the 3 NHS Scotland regional hub laboratories were established to provide resilience during rapid spikes in demand. There was also some variation in testing deployment and the detail and timing of testing policies – for example, workplace testing continued for longer in Wales, Scotland and Northern Ireland than in England.

Other testing methods, including reverse-transcription loop-mediated isothermal amplification (RT LAMP) and other testing functions, notably antibody testing, were also explored. Technologies (and why some were used over others) are explored in more detail below.

Later 2020 to 2022: expansion of asymptomatic testing

Although PCR had been used for asymptomatic testing during 2020, turnaround times, the need for laboratory processing, relatively high costs and high resource needs to conduct PCR testing at routine mass population scale, and the fact that PCR testing was sensitive to viral fragments long after infection had resolved, all meant that it was not realistic for routine, mass asymptomatic and pre-symptomatic testing. The arrival of LFDs to the market in later 2020 and 2021, which produced results for self-testing within 15 minutes on average, enabled increasingly widespread self-testing across the population during this period, and testing strategies adapted accordingly.

Initially, there were few LFD products available on the market, and the quality of initial products was very variable with some showing low sensitivity and/or specificity. Manufacturer claims about the performance of individual tests were often not matched when tests were analysed against criteria for quality and reliability. A national scheme was set up to address this and evaluate test quality and reliability. This process is set out in more detail below under ‘Quality evaluation, improvement and validation of testing technologies’.

Of course, LFDs generally have lower sensitivity and specificity than PCR (see ‘Technologies’ below), but once LFDs of sufficient quality became available at scale they were sufficiently reliable and accurate to enable mass routine asymptomatic testing. This supported individuals to assess their likelihood of infectiousness on a day-to-day basis, and was important for key settings such as schools, hospitals and care homes. The asymptomatic testing programmes operated in 4 main testing groups:

• group 1: repeated testing to detect positive cases among asymptomatic individuals (and remove them from circulation) – for example, testing regimes for staff working in high-risk settings such as the NHS, social care, homeless shelters and prisons
• group 2: testing prior to an activity to reduce risk (this may be one or more tests)
• group 3: asymptomatic testing where there was a signal of a potential outbreak (or where there had been an outbreak) to control infections, or where there was perceived to be a higher risk
• group 4: daily testing of contacts to identify positive cases early

As evidence emerged during 2021 on the potential use of LFDs at a population level to highlight potential infectiousness and as a rapid diagnostic tool, further use cases evolved:

• to guide antiviral prescription for those eligible
• for those isolating and contacts to assess their infectiousness (and exit isolation where tests on 2 subsequent days from day 5 onwards were negative)

In 2021 and 2022, more transmissible variants established and many population-wide non-pharmaceutical interventions (NPIs) were eased as both natural and vaccine-derived immunity rose and weakened the link between infections and severe disease. This led to much higher case rates – and testing demand – than that seen in 2020. As a result, LFDs became increasingly central in testing strategies as a way to rapidly test millions of (symptomatic and asymptomatic) people on a weekly basis without needing to further expand laboratory capacity. In under 5 months, the number of LFD tests reported using the existing UK National Testing Programme digital infrastructure had risen from 73 in the week commencing 22 October 2020 to more than 7.6 million in the week commencing 11 March 2021.^{[footnote 5]}

Throughout this period, groups, nations and regions in the UK innovated ways in which testing could be used, with initiatives such as the Events Research Programme which examined the risk of COVID-19 transmission from attendance at events and interventions to reduce that risk, and the commissioning of school and general population trials of daily testing for contacts versus self-isolation.^{[footnote 6]}, ^{[footnote 7]}, ^{[footnote 8]}

The city-wide Liverpool voluntary COVID-19 rapid antigen testing pilot provided community open-access LFD testing for those with or without symptoms to understand the possible role of mass LFD testing in various pandemic control strategies.^{[footnote 9]} More than half the population took up asymptomatic testing, and the evaluation found that LFDs identified most COVID-19 cases with high viral load and that the pilot led to an estimated 21% reduction in cases during its first 6 weeks. It also found that test uptake was lower and infection rates were higher in deprived areas, in areas with fewer digital resources or lower digital literacy, and among non-white ethnic groups. Fear of income loss from self-isolation was a key barrier to testing. These were important findings informing not only the possible role of LFDs in pandemic control strategies but also considerations in their deployment. LFD testing differed slightly across the UK nations, predominantly in its deployment timing, but overall strategic aims were similar.

Throughout the pandemic

Testing supported research and clinical trials throughout the pandemic, but no systematic process existed to link people engaging in testing to trials or research studies. Approaches by individual trials to different parts of the testing infrastructure enabled some level of linking, for example with testing supporting trials for the AstraZeneca vaccine, followed by Valneva phase 1 and phase 2, and Novavax phase 3. Testing was also made available to key studies such as SIREN and also the ZOE app study. Systems were put in place to enable this, such as text alerts referring eligible people to National Institute for Health and Care Research (NIHR) trials or processes to enable access to testing infrastructure for research. However, in the future an established process to proactively identify eligible people for trials in advance of a pandemic and through the routine testing infrastructure would be helpful.

Throughout the pandemic, sequencing enabled baseline surveillance for emerging variants and changes in existing variants. This was used in conjunction with more timely case data to assess the potential impacts of new variants and adapt strategies accordingly. There was also a continuous need to evaluate the effectiveness of assays for new variants. Viral neutralising assay studies informed, for example, our understanding of potential vaccine escape. This process is outlined below under ‘Quality evaluation, improvement and validation of testing technologies’.

There were some differences across the UK in variant testing – for example, ‘surge’ testing following identification of variant cases in the first half of 2021 was predominantly conducted in England. At the time of writing there is not yet conclusive evidence on the effectiveness of surge testing to slow or stop the spread of variants, though it was also implemented to gather data supporting early assessment of the characteristics and dynamics of a given variant. This highlights the need for ongoing evaluations during pandemic response to be embedded in all areas.

Technologies

To expand capacity, reduce risk of supply failure and to service anticipated use cases, a wide selection of diagnostic technologies were supported into development and evaluation in this pandemic. At the inception of testing, a number of technologies were explored as it was unclear how effective, scalable or reliable each was. Broadly, there were 3 methods:

• molecular, to detect viral ribonucleic acid (RNA)
• antigen, to detect viral proteins
• serology, to detect host antibodies

What was available and what tests were used for changed over the course of the pandemic, and we anticipate this will be true also of future pandemics and epidemics. It was important to establish the sensitivity and specificity under specific conditions (for example, depending on viral load) and at which points after infection testing was most effective. The training and equipment needed was also important, as was acceptability of different testing methods and sample sites. It was important to engage operational and industry experts early on to understand, alongside the technological capabilities of a given test, what would be feasible in practical terms for its widespread use.

Alternatives to improve accessibility were continually reviewed, such as saliva sampling. However:

• LFDs with saliva sampling methods did not pass UK Health Security Agency (UKHSA) performance tests until relatively late on in the pandemic
• PCR with saliva sampling would have required changes to lab and logistics infrastructure (or parallel lab and logistics infrastructure) which would be costly, while the benefits in increased uptake were not compelling in evaluations
• LAMP testing using saliva samples did exist, but it was challenging to deploy LAMP testing at scale

Nasal and throat swabs were predominantly used, later switching to nasal only. We anticipate testing technologies will continue to evolve, and the next pandemic may well have technologies not currently available, but the broad principles of lab-based or point-of-use testing for acute infection, and serology or similar for prior infection, will remain.

Table 1: advantages and disadvantages of different methods

Method type	Advantages	Disadvantages
Molecular (PCR)	- High clinical and analytical sensitivity and specificity - Samples can easily be moved off-site and sent to labs far away -Can use self-collected samples - Possible to conduct at scale - Can indicate some variants in advance of genomic testing	Turnaround time longer than LFD (this varied depending on a number of factors – see main text) - Cannot easily distinguish between whole virus and viral fragments, so can continue to show positive after active infection has subsided
Molecular (RT-LAMP)	- Rapid results (less than 20 minutes) - Performed well in pre-infectious and infectious phase - Comparable performance to RT-quantitative PCR (qRT-PCR)	- RT-LAMP machines usually need to be on-site with regular quality checks - Staff training requirements to use
Antigen (LFD)	- Rapid results (10 to 30 minutes) - Results at point of testing (convenient) - Does not require laboratory process or specialist knowledge to interpret results and can use self-collected samples, decentralising testing - Possible to conduct at very large scale - Lower cost than RT-LAMP (for example, for asymptomatic testing) - Can indicate infectiousness	- Lower clinical and analytical sensitivity and specificity than PCR - Results can be misinterpreted by inexperienced users
Serology	- Enables retrospective analyses of outbreaks (for example, highlighting asymptomatic disease) - Enables surveillance of seroprevalence	- Results only available when antibodies detectable, which may be outside window for informing intervention

Molecular tests

RT-PCR tests for SARS-CoV-2 were developed early in the pandemic in the UK, with tests available in small numbers from January 2020. The workup of a new PCR diagnostic test for SARS-CoV-2 took between 3 and 6 weeks as it was dependent on:

• knowledge of exact viral sequence and viral diversity in target area (see chapter 1, question 2: ‘What information could be gathered about the pathogen that could help develop an initial diagnostic test?’)
• ability to source key reagents (primers and probes)
• ability to source appropriate clinical material for assay validation purposes (establishing analytical sensitivity and specificity)
• ability to source control material (known template)
• ability to deploy the new test against relevant clinical material acquired from cases of new virus infection (which may be difficult to acquire)

RT-PCR tests did not easily distinguish between whole viable virus and viral fragments, and so repeat PCR tests were not advised within 90 days of infection. Initial PCR turnaround was slow because of limited supply and number of testing sites. Over the course of 2020 turnaround times reduced but this varied significantly during 2020 to 2022 according to:

• the level of demand
• wider testing processes and infrastructure
• whether the testing was performed in 4 nations’ NHS and public health laboratories (often referred to as pillar 1) or in the cross-UK laboratory network of pillar 2

The samples were initially taken by healthcare professionals, then by trained individuals operating in the established test sites for the collection of community samples, and finally, with guidance and communications, by the general population in self-collected samples. There was an option to self-test at home using posted tests (though delivery of tests to individuals extended the turnaround time), or to self-test at testing centres.

The analytical sensitivity (ability within the lab to detect a SARS-CoV-2 positive sample of RT-PCR) was very high, and as is typical for PCR specificity for virus was also high. Clinical sensitivity, taking into account not just the accuracy within the lab but also all other factors in collecting a relevant sample and processing it, was also high. The long tail of positive tests after infection, however, had an impact on clinical specificity as well as on individuals needing to comply with restrictions for those testing positive, as PCR could be detecting viral fragments from a previous infection in an individual who was no longer infectious. Clinical sensitivity and specificity, as well as positive and negative predictive values, can be impacted by many different factors such as the anatomical location of viral replication, which clinical sample was taken for diagnosis or, in the case of positive and negative predictive values, background prevalence.

It was important to have more than one diagnostic target to provide assurance of accuracy, and in the very early days of the expansion of testing for COVID-19 it was important to have a detection and confirmation strategy as separate steps. A confirmation strategy involved the use of a second viral target or a partial genome sequence to give confidence in accuracy.

There was a need to adjust test design or critical reagents during the early stages of this pandemic, as a result of emerging knowledge about optimum clinical sample or information about viral diversity.

Modest re-configurations of the same underlying test technology as PCR testing, and of a technology used in agriculture settings, led to development of high throughput endpoint PCR (ePCR) testing for SARs-CoV-2. This offered an efficient means to expand test capacity in the community-based testing programme.^{[footnote 10]} The technology incorporated high throughput sample handling into a large-batch-size, continuous manufacturing process.

RT-PCR was a core technology in the UK’s testing system and has provided the vast majority of molecular symptomatic diagnostic testing to date. It was also used for asymptomatic testing with weekly PCR tests supplemented by further LFD testing as part of the care home staff testing regime until March 2022. As the pandemic progressed and more variants circulated, the national testing programme (in conjunction with the regulator) requested that manufacturers reported the ability of the PCR technology to detect the circulating variants at the time, in order to ensure that viral identification was maintained.

It was important to recognise that the testing system needed to involve more than the ability to detect the virus in a laboratory and it needed to link to an end-to-end informatics system from individual through to clinical care and public health needs.

Loop-mediated isothermal amplification

RT-LAMP is a rapid nucleic acid (molecular) amplification technique that takes less than 20 minutes to provide a result. Two assay formats were developed and deployed to detect SARS CoV-2: first, using extracted RNA, and second, direct using saliva samples. RT-LAMP assays amplify larger genomic regions than RT-qPCR, and therefore performed well during the pre-infectious and infectious phase when there is freshly produced RNA. RNA RT-LAMP has comparable performance (sensitivity and specificity) to RT-qPCR on swabs and can detect virus in a wide clinical window. The first multicentre pilot deploying a LAMP CE-marked assay to detect SARS CoV-2 began in August 2020, and the assay was validated by the Technical Validation Group in December 2020.^{[footnote 11]}

For COVID-19 detection the direct RT-LAMP assay was predominantly deployed for asymptomatic testing in the NHS in staff members, with smaller use cases in school and social care. Twenty-nine mobile processing units making use of RNA RT-LAMP were also deployed, which were able to respond to outbreak areas in care homes, hospitals, schools, prisons and town centres as well as providing testing at events such as the G7 Summit, and could use the same swabs as the PCR infrastructure.

Direct RT-LAMP testing was not widely used in this pandemic, because the machines to process tests were large and needed space to sit and required regular maintenance. They were also a relatively new technology that had not been used at scale for other pathogens and so required staff training for use. For these reasons they were not as easy to scale and manage in a national end-to-end pathway as PCR. Similarly, for widespread deployment of asymptomatic testing, another testing technology (LFDs) provided a lower cost option that did not require healthcare professionals, maintenance of machinery or dedicated areas to store machines.

Antigen tests

LFDs enabled rapid point-of-care or self-test for current infection and when people are likely most infectious, with results appearing on the device in 10 to 30 minutes.^{[footnote 12]} LFDs did not require sophisticated laboratory infrastructure or skilled personnel and therefore provided decentralised testing. A digital process was set up to enable ordering and distribution of LFDs and reporting of the outcome by individuals.

LFDs were increasingly used by individuals as the pandemic progressed to conduct routine asymptomatic testing and manage their own risk (for example, by testing before high-risk activities or contact with a clinically vulnerable person, daily contact testing or case testing for infectiousness to determine appropriate end date for isolation). They were also used in research – for example, in the SARS-CoV2 immunity and reinfection evaluation (SIREN) study, which was an important source of evidence on duration of immunity. By the spring of 2022, they also took an increasingly central role in assessing infectiousness to guide isolation timelines (with 2 negative tests on sequential days after day 5 enabling confirmed COVID-19 cases to end isolation).

LFDs are less sensitive than molecular tests. However, they have been shown to be effective in indicating high viral load, and so were used as an indication of likely infectiousness.^{[footnote 13]} COVID-19 Testing Delivery Programme data indicates that the LFDs in use detect between 83.0% (95% confidence interval 82.8% to 83.1%) and 89.5% (95% confidence interval 89.4% to 89.6%) of cases.^{[footnote 14]} Their ease of use and speed of results therefore had to be balanced with careful interpretation of results, and public messaging stressing this – for example, recommending repeat testing on sequential days to increase sensitivity.

LFD quality was initially highly variable and this is an area that required strong regulatory processes (see ‘Quality evaluation, improvement and validation of testing technologies’ below). In order to identify those LFDs that displayed high specificity and high sensitivity against viral loads associated with infectiousness, in August 2020 ministers commissioned the UKHSA laboratories at Porton Down to evaluate LFDs, with UKHSA and Oxford University setting evaluation protocols and providing oversight and UK government labs used to rapidly evaluate the tests.^{[footnote 15]} Importantly, there was a need to monitor real-world data alongside these lab evaluations to understand true effectiveness in use.

Finally, for highly accurate data collation from testing modalities, data from LFDs relied on people self-reporting their results and this could not always be relied upon, particularly where there might be little incentive to do so. However, by giving individuals information on their likely infection status, the tests still supported early access to treatments for those most at risk of severe outcomes and changes in behaviour to reduce transmission following a positive test.

In 2021, UKHSA refined the types of LFD it evaluated, focusing on more usable devices that had regulatory approval for self-testing and used less invasive nasal swabs.

There were other important antigen tests besides LFDs, such as microfluidic immunofluorescence assay point-of-care antigen tests using nasal and nasopharyngeal swab samples which were used for rapid admissions testing in clinical settings.

Serology tests

Antibody tests were available from February 2020 and were initially considered to guide interventions on an individual level (for example, to enable those with previous infection to return to work). Many healthcare workers were offered antibody testing to judge potential immunity status, particularly in the first wave when there were no known effective medical countermeasures and risk to staff and patients was potentially at its highest. However, using antibody testing to guide individual interventions requires extensive understanding of reinfection and immunity across different individuals. Therefore potential use cases for antibody tests – such as prioritising antivirals for those with immunosuppression – needed to be treated with care.

Antibody testing has been an important tool for research throughout the pandemic to date – for example, in the SIREN study on healthcare staff reinfection rates.^{[footnote 16]} It has also been key in understanding population seroprevalence – for example, through the ONS CIS, Public Health Scotland’s seroprevalence survey or the Real-time Assessment of Community Transmission 2 (REACT-2) study.^{[footnote 17]}, ^{[footnote 18]}, ^{[footnote 19]} In COVID-19 it was possible to differentiate between antibodies due to prior infection and to vaccination; this is not always possible.

Genomic sequencing and genotyping

In the UK, the first COVID-19 sequence was generated by UKHSA laboratories. Joint work across these laboratories, the public health services of the UK, the Wellcome Trust Institute and a network of NHS clinical laboratories and universities through the COG-UK consortia enabled the collective establishment of a national sequencing and analysis capability that tracked several VOCs. COG-UK was instrumental in getting genomic sequencing established and scaled in the UK. It was also successful in integrating the skills and expertise of academic experts with public health specialists to understand the genomic variation within the virus, how this was evolving and which mutations might be responsible for severe disease or increased transmission.

Rapid and scalable whole genome sequencing capacity was needed to underpin efforts to control transmission. Genomic sequencing was transitioned from research into a sustained service linking genomic sequencing with serological and biological analysis to understand the impact of the emergence of variants on the trajectory of the pandemic.

In March 2021, the UK provided close to 50% of the world’s registered output in genomic sequencing. By September 2022, over 2.8 million cases had been sequenced in the UK with the Wellcome Sanger Institute leading the sequencing of community cases. The institute also provided an important early interpretation signal throughout the pandemic for genomic surveillance of viral evolution and for VOCs once these were identified. UKHSA laboratories in 2021 expanded sequencing capacity to 25,000 genomes per week and since then led the sequencing of the virus from hospitalised patients across the NHS.

New variants made it an imperative to build an integrated capability to rapidly test, diagnose and sequence samples, and to continue to undertake baseline surveillance for emerging variants. To speed the detection of known VOCs, genotyping was introduced alongside PCR testing into both hospital and community testing laboratory networks in order to test for specific mutations in known variants. This enabled an early warning system (usually within 24 hours of a positive PCR result) ahead of definite results from whole genome sequencing.

Genomic surveillance has supported research on vaccine and therapeutic effectiveness for new variants by tracking the growth and distribution of variants in circulation alongside changes in rates of hospital admissions and severe disease. It also enabled monitoring of the virus for genetic mutations which could cause it to be more easily transmitted or to escape vaccines, and for the public health response to be guided accordingly. Sequencing has also supported timely assessment of the efficacy of diagnostic tests for different variants. An early warning system was introduced for laboratories performing genotyping to report and refer concerns in assay performance, and to identify assays that did not work for particular variants at earliest opportunity. This was linked with a Medicines and Healthcare products Regulatory Agency (MHRA) regulatory requirement for manufacturers to report ongoing evaluation of deployed technologies in relation to variants.

The ability to track variants and monitor emergence of variants also supported international collaboration and planning, both directly with other countries and with the World Health Organization (WHO). When border testing was introduced, the COVID-19 Testing Delivery Programme set the standards for private sector providers to widen the performance of whole genome sequencing on positive samples and track the introduction of new variants into the UK. Across many countries during this pandemic, genomic surveillance was expanded and results were shared early with other nations. This was important to track variants and highlight potential risks early.

Further technical innovations

Testing without a way to create information for the individual and for the pandemic monitoring process is of limited use. Information technology was vital to:

• delivering tests
• processing tests
• reporting, storing and sharing the data
• communicating the results and action to be taken

It was also key to bringing results together at scale for rapid analysis and assessment of the situation, with testing data supporting policymakers, public health professionals (nationally, regionally and locally), health and care professionals, academics and the public to understand the course of the epidemic. Localised testing data also supported people to take informed decisions on day-to-day activities in their local area. For the first time within a national system, testing results were returned into individuals’ healthcare records, giving them a permanent healthcare record of their test result. This included self-reported LFDs, though there were questions around the robustness of these results as there were limited incentives for individuals to upload results, and in some instances incentives to falsely upload negative results – for example, to enter venues or events where certification was in place, or to attend work.

Electronic contact tracing also generated real-time data on the rate of transmission and the geographical distribution of cases and enabled us to track the effectiveness of control strategies over time.

Digital readers for LFD results improved the accuracy of the tests while making reporting of results easier. Digital readers had to go through a full development process including regulatory steps, and this took time. Once regulatory approval was in place, an initial pilot and evaluation was undertaken followed by wider rollout of the technology.^{[footnote 20]} Other innovations included quick response (QR) codes – for example, to identify individual LFDs, the development of mobile laboratories for use in outbreaks and specific settings, and improvements in accessibility across a number of areas (for example, apps providing visual support for home PCR test sample collection and self-test LFDs).^{[footnote 21]}

Table 2: summary of testing technology uses

Strategy	Use	Deployed testing technologies
Symptomatic Testing	- Diagnostic testing for clinical care - Diagnostic testing for public health purposed to stop onward transmission - Confirmatory testing - Surge/outbreak testing - Directing therapies eg AV - Testing to determine infectiousness (and therefore guide isolation timelines) following infection	- qRT-PCR - ePCR - Viral sequencing - RNA LAMP - Point of Care (PoC) - Genotyping – reflex assay - LFDs (at a later stage of the pandemic)
Asymptomatic testing	- Universal offer - School/university testing - Daily contact testing - Regular testing of staff members in high-risk workplaces - Borders testing - Certification/COVID-pass to access events - One-off testing for closed environments such as elective care testing, testing of visitors to care homes - Discharge/transfer testing - Outbreak testing - Testing for outbound international travel (which private suppliers provided) - Testing to determine infectiousness (and therefore guide isolation timelines) following infection	- LFDs - Direct LAMP – saliva - PoC tests - qRT-PCR/ ePCR (in limited situations)
Surveillance	- Detection of existing VOCs - Detection of new VOCs - Pandemic trajectory - Research studies - Borders testing for surveillance	- qRT-PCR - Viral sequencing - Antibodies - Genotyping – reflex assay - Wastewater sequencing

Quality evaluation, improvement and validation of testing technologies

Throughout the pandemic there was a pressing need to:

• act swiftly
• review a wide range of technologies across the testing pathway
• encourage and enable innovation, in particular from the private sector

At the same time, technologies needed to be high quality and effective, and entry to the market needed to be properly managed by a regulator at a time of high demand and rapid innovation. The United Kingdom Accreditation Service (UKAS) was of course in place, but there was considerable pressure on the organisation due to a rapid increase in inspection demands during the pandemic. This need for rapidly scaled-up and independent accreditation is an important lesson from this pandemic, and we set out here the steps taken to set this up at speed.

An evaluation process for diagnostic technologies was established in the national testing programme in mid-2020 to support procurement strategies, including performance testing in real-world settings. Early engagement and support from the MHRA facilitated entry into the regulatory system and entry to market for effective technologies. In lab-based testing technologies this included the full pathway, from validation of sample collection methods to quality assurance in laboratories and behavioural insight review of results messages. In LFDs this included pre and post-deployment evaluation to ensure technologies continued to perform as expected once deployed.

To evaluate and improve quality, product and laboratory validation oversight was set up alongside assurance of regulatory compliance both pre and post-deployment in the market. A number of frameworks and processes supported this, including:

• measurement of key performance indicators against service standards
• compliance with relevant standards and regulatory requirements across all organisation functions
• continuous quality improvement
• review and monitoring of any risks associated with the delivery of products and services

Public health agencies, alongside the UKAS and regulators including the MHRA, worked jointly to manage these processes, but the pandemic exposed a gap in the regulatory process for diagnostic testing both in the UK and globally. For example, data packs provided for self-certification were variable and often limited so it was difficult to judge real-world performance from these.

There was also a need for validation processes to confirm manufacturers’ performance claims. Initially, manufacturers of SARS-CoV-2 detection devices could self-declare compliance to obtain their CE marking and there were no set processes or minimum evidence levels for test performance required when making this declaration. This allowed manufacturers to maximise performance claims, and validation for Department of Health and Social Care (DHSC) procurement of tests flagged that a significant number of these devices were failing to replicate their claimed performance when assessed in a technical lab validation.

For example, 75% of lateral flow test devices that applied for DHSC procurement and went through the validation process failed the validation standard. However, these devices remained available for sale on the UK market for anyone to buy, including NHS trusts and commercial providers of testing services. It is unclear what harm this may have caused. Without an independent validation process it was extremely difficult to know whether these test devices performed as claimed for a particular use case. As the private testing market grew, the issue of these poorly performing tests continuing to be available became more acute, both for testing processes and testing kit (such as swabs) both of which required quality assurance.

To address this issue, legislation came into force on 28 July 2021 requiring all antigen and molecular detection tests for COVID-19 to be validated. From 1 November 2021, it was unlawful to supply an unvalidated SARS-CoV-2 test (subject to limited exceptions).

As universal testing offers came to an end in different nations of the UK, the general public have been able to purchase tests from online and high street retailers. The Coronavirus Test Device Approvals regime has helped to drive up performance standards and should give greater confidence to UK consumers in the accuracy and performance of the tests they purchase.

Collaboration

A cross-UK approach to setting up community testing at scale meant that surge capacity could be offered more efficiently and to areas needing it most, and built resilience into the system (for example, if one lab was suddenly out of operation, samples could easily be diverted). This was particularly helpful to smaller nations but not exclusively: in 2021, for example, Public Health Wales supported surge testing for variants in Bristol by processing samples in its genomic sequencing facilities. It was also particularly important at points in the pandemic where there was a very sudden and acute need to ramp up capacity, encourage the population to test (symptomatic and asymptomatic), or change public health guidance on isolation. Although some elements of delivery and policies varied across the UK, approaches to testing were underpinned by the same evidence base and testing principles and a major lesson from this pandemic is the value of joint working across the 4 nations.

Data systems and health systems differ across the 4 nations, and there was a need to consider the full range of circumstances when designing a shared testing system. Testing policy and delivery is complex, with multiple interacting systems and needs, and so governance structures need to bring in and work collectively with colleagues from across the UK while remaining simple and understood across the relevant sectors or organisations involved, and while enabling appropriate governance within each nation for operational deliveries and devolved responsibilities. For future preparedness, early consideration, delineation and resource allocation to any necessary regulatory and assurance bodies should also be seen as necessary for rapid but robust innovative diagnostic provision.

Collaboration was essential to testing delivery – between government, the NHS, public health agencies, industry and academia, through the exchange of staff, equipment, knowledge, skills, data sharing and interoperable systems. It was a significant challenge, particularly during the first wave when multiple organisations were having to act outside their usual remit, an unprecedented volume of samples and data had to be gathered, stored and shared, and there were multiple competing demands on resources. There was widespread sharing of staff between universities and the cross-UK laboratory network, but interoperability has not yet been achieved – the UKHSA Rosalyn Franklin laboratory, for example, is not interoperable with NHS testing laboratories, and sample tracking and results sharing has had to be retrofitted rather than done using shared interoperable systems across laboratories. Having agreements and sleeping protocols for sharing information, equipment, samples and staff across the sector (including with private sector suppliers) is an important step to avoid this in the future, but realistically the move to scale up any testing system in such a short time will likely meet similar challenges in the future.

Information sharing across testing systems was important across the 4 nations – for example, when people travelled from one nation to another. It was also important globally, with global sharing of sequence and variant data informing the global response as well as allowing evaluation of the performance of diagnostic tools as the virus’s properties emerged.

The changing nature of the pathogen impacted testing strategies and operations. The ability to track variants and monitor emergence of variants informed and strengthened the ongoing response to the pandemic and informed the government response, especially at the borders and when working with governments in other countries and international agencies. Genomic surveillance enabled monitoring the virus for genetic mutations which could cause it to be more easily transmitted or to escape vaccines, and for the public health response to be guided accordingly.

Reflections and advice for a future CMO or GCSA

Point 1

There were 2 important questions at the outset:

1. What do we need?
2. How should we prioritise what we have while we build up to what we need?

Limitations in testing capacity and an end-to-end system to effectively use the output of testing were initially a major constraint. The magnitude and speed of scale-up required in the testing system for COVID-19 was unprecedented. The major efforts required to expand testing capacity highlighted the importance of building testing systems that maintain some form of contingency response, or at least retain some expertise on how to surge in the event of a new variant or an entirely new pandemic. The diagnostics industry should be included in planning as they may be a key partner (for example, in providing rapid surge capacity).

Point 2

It was important – and the UK did not always get this right – to align testing aims, use cases, technologies, data flows and communications in coherent testing strategies.

This can be challenging in the context of new systems and processes, new testing technologies and use cases, and inter-organisational working. An agreed plan for prioritising usage was also required – for example, targeted at high-risk settings (staff and patients in hospital and in care homes) and for outbreak management.

Point 3

Testing was deployed for a wide range of use cases in this pandemic, some of which may be required in future pandemics.

Some use cases were very similar to normal use of tests in infectious disease outbreaks, including for clinical diagnosis, infection control in hospitals, case finding, surveillance and research. Others, such as repeated testing using self-read and self-reported testing, were new at this scale.

Once reliable lateral flow tests were available it significantly improved people’s ability to manage their own risks and the risks for those they were meeting, as well as supporting surveillance at scale.

Point 4

Testing innovations came at speed and required a rapid, independent quality assurance and validation process.

Quality in the market was very variable and the regulatory approach globally was variable.

Point 5

Communication of the rationale and practical requirements of testing strategies and changes to testing policy was important, whether with the public or professionals.

Although better communications were developed throughout the pandemic, there are some specific interventions – such as translating testing instructions and advice from the very outset, and engaging through trusted community leaders – which could be delivered better in future responses.

Some elements of testing were, and will remain in a future pandemic, complex to communicate – such as the link between the positive or negative predictive value of a test and prevalence. Pilots were helpful in understanding how new strategies or policies might operate and how people might respond to them.

References

1. See, for example: Scottish Government: Coronavirus (COVID-19): Scotland’s testing strategy – adapting to the pandemic – supporting documents, Coronavirus (COVID-19): review of testing strategy - October 2020 and Coronavirus (COVID-19) – testing strategy: update – March 2021 – supporting documents ↩

2. Plessis LD and others. Establishment and lineage dynamics of the SARS-CoV-2 epidemic in the UK, Science 2021: volume 371, issue 6,530, pages 708 to 712 ↩

3. DHSC Policy Paper: ‘Coronavirus (COVID-19): scaling up testing programmes’, published 4 April 2020. Available at: https://www.gov.uk/government/publications/coronavirus-covid-19-scaling-up-testing-programmes ↩

4. UKHSA, ‘NHS Test and Trace laboratory and contact centre utilisation’, 7 February 2022. Available from: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/1053145/utilisation-ad-hoc.pdf ↩

5. DHSC Transparency Data: ‘Weekly statistics for NHS Test and Trace (England), 13-19 May 2021’. Available at: https://www.gov.uk/government/publications/weekly-statistics-for-nhs-test-and-trace-england-13-may-to-19-may-2021 ↩

6. DHSC, DCMS, BEIS Notice: ‘Information on the Events Research Programme’, updated 26 November 2021. Available at: https://www.gov.uk/government/publications/information-on-the-events-research-programme/information-on-the-events-research-programme ↩

7. Young B. C. et al, ‘Daily testing for contacts of individuals with SARS-CoV-2 infection and attendance and SARS-CoV-2 transmission in English secondary schools and colleges: an open-label, cluster-randomised trial’. The Lancet, 2021. V 398, Iss 10307, pp 1217-1229. DOI: https://doi.org/10.1016/S0140-6736(21)01908-5 ↩

8. DHSC Guidance: ‘Daily contact testing study’, published 29 April 2021. Available at: https://www.gov.uk/guidance/daily-contact-testing-study ↩

9. DHSC Research and Analysis: ‘Liverpool coronavirus (COVID-19) community testing pilot: full evaluation report summary’, published 7 July 2021. Available at: https://www.gov.uk/government/publications/liverpool-coronavirus-covid-19-community-testing-pilot-full-evaluation-report-summary/liverpool-coronavirus-covid-19-community-testing-pilot-full-evaluation-report-summary ↩

10. DHSC Research and Analysis: ‘Evaluation of endpoint PCR (EPCR) as a central laboratory based diagnostic test technology for SARS-CoV-2’, published 28 January 2021. Available at: https://www.gov.uk/government/publications/evaluation-of-endpoint-pcr-epcr-as-a-diagnostic-test-technology-for-sars-cov-2/evaluation-of-endpoint-pcr-epcr-as-a-central-laboratory-based-diagnostic-test-technology-for-sars-cov-2 ↩

11. DHSC Research and analysis: ‘Rapid evaluation of OptiGene RT-LAMP assay (direct and RNA formats)’, published 1 December 2020. Available at: https://www.gov.uk/government/publications/rapid-evaluation-of-optigene-rt-lamp-assay-direct-and-rna-formats/rapid-evaluation-of-optigene-rt-lamp-assay-direct-and-rna-formats ↩

12. Welsh Government Guidance: ‘How COVID-19 lateral flow tests work’, published 23 February 2021. Available at: https://gov.wales/how-covid-19-lateral-flow-tests-work ↩

13. DHSC Research and analysis: ‘Key points summary: asymptomatic testing for SARS-CoV-2 using antigen-detecting lateral flow devices (evidence from performance data October 2020 to May 2021)’. Published 7 July 2021. Available at: https://www.gov.uk/government/publications/lateral-flow-device-performance-data/key-points-summary-asymptomatic-testing-for-sars-cov-2-using-antigen-detecting-lateral-flow-devices-evidence-from-performance-data-october-2020-to-m ↩

14. Lee L. W. Y. et al, ‘Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Infectivity by Viral Load, S Gene Variants and Demographic Factors, and the Utility of Lateral Flow Devices to Prevent Transmission’. Clin Infect Dis, 2022 Feb 11;74(3):407-415. doi: 10.1093/cid/ciab421. ↩

15. UKHSA Guidance: ‘Protocol for evaluation of rapid diagnostic assays for specific SARS-CoV-2 antigens (lateral flow devices)’. Updated 11 July 2022. Available at: https://www.gov.uk/government/publications/assessment-and-procurement-of-coronavirus-covid-19-tests/protocol-for-evaluation-of-rapid-diagnostic-assays-for-specific-sars-cov-2-antigens-lateral-flow-devices ↩

16. UKHSA Guidance: ‘SIREN study: Providing vital research into coronavirus (COVID-19) immunity and vaccine effectiveness nationally.’ Published 20 June 2022. Available at: https://www.gov.uk/guidance/siren-study ↩

17. Example statistical bulletin from ONS Coronavirus (COVID-19) Infection Survey antibody data, UK. Antibody data by UK country and age in England from the Coronavirus (COVID-19) Infection Survey. Available at: https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/conditionsanddiseases/bulletins/coronaviruscovid19infectionsurveyantibodyandvaccinationdatafortheuk/24august2022 ↩

18. Imperial College London: Real-time Assessment of Community Transmission. Summary page of reports and findings, available at: https://www.imperial.ac.uk/medicine/research-and-impact/groups/react-study/real-time-assessment-of-community-transmission-findings/ ↩

19. COVID-19 weekly seroprevalence for Scotland. Available at: https://publichealthscotland.scot/our-areas-of-work/conditions-and-diseases/covid-19/covid-19-data-and-intelligence/covid-19-weekly-seroprevalence-for-scotland/covid-19-weekly-seroprevalence-for-scotland-overview/ ↩

20. Consortium, AI LFD and Banathy, R. et al, ‘Machine Learning for Determining Lateral Flow Device Results in Asymptomatic Population: A Diagnostic Accuracy Study.’ Available at SSRN: https://ssrn.com/abstract=3861638 or http://dx.doi.org/10.2139/ssrn.3861638 ↩

21. UKHSA Press release: ‘COVID-19 rapid testing made easier for partially sighted people’. Available at: https://www.gov.uk/government/news/covid-19-rapid-testing-made-easier-for-partially-sighted-people?msclkid=fef3dec3cfed11ec8d2cd361e8acca3a ↩