Mass testing for COVID-19 using lateral flow tests

It is thought that at least 20% of people with SARS-CoV-2 infection do not have any symptoms. While the evidence suggests they are likely to be less infectious than people with symptoms, they can still transmit the virus. Therefore, approaches to detect this group of people are important. Testing large groups of largely healthy (asymptomatic) people to see if they have COVID-19 is being rolled out nationally by the UK Government.

This requires accurate tests that are easy to use and that give a result in a matter of minutes. Until an effective vaccine is licensed for use and widely available, the Government has proposed that mass population testing is an approach that could limit the need for restrictions on daily life. This article explains how reliable these tests are, what mass testing pilots are being undertaken, and the strengths and weaknesses of using them in this context.

Testing for COVID-19

Different COVID-19 tests detect either a previous infection (an antibody test) or an active infection (a molecular test). Lateral flow tests are a type of molecular test to detect an active infection. The tests contain antibodies that bind to proteins (antigens) on the surface of the virus if it is present in a sample. A positive result is seen as a dark band or a fluorescent glow on the test kit.

These tests do not need laboratory processing or other equipment and can give results in under an hour. Such antigen tests are a cheap technology that can be made in large quantities and so they are well suited to mass testing.

The trade-off is that their lower sensitivity means that they are less accurate at identifying people who have an infection. However, some researchers suggest that test sensitivity should come second to the ability to test frequently and obtain results quickly, if they are used to screen large numbers of people.

How is accuracy measured?

As with any diagnostic test, data about accuracy and reliability are crucial; no test can claim 100% accuracy. The most important terms used in describing accuracy are:

• Sensitivity: the proportion of people with SARS-CoV-2 infection who test positive. For example, a test with sensitivity of 95% would mean that 5 in 100 people who have COVID-19 would test negative (false negative). They have an infection, but the test says that they don’t.
• Specificity: the proportion of people without SARS-CoV-2 infection who test negative. For example, a specificity of 90% means that 10 in 100 people who are not infected still test positive (false positive). They do not have an infection, but the test says that they do.

These figures alone can be misleading because other factors influence overall accuracy when tests are used in real-world settings. When levels of infection in the population being tested are high, a test with a high level of sensitivity will be very good at identifying people with an infection but less good at detecting people that do not.

Conversely if the level of infection in a population is low, then a higher specificity becomes more important because it will be better at identifying the people that don’t have the infection than it is at detecting the people that do. The World Health Organization recommends that the acceptable minimum standards for tests should be 80% for sensitivity and 97% for specificity.

You can read more about this in this article on interpreting COVID-19 test accuracy and the implications for how they are used.

What test features are most important for mass testing?

No one test is suited to all the purposes they can be used for. A test for mass screening would need to be sufficiently sensitive to detect infected people; but very high specificity becomes much more important if tests are used to screen lots of healthy people.

While there may be a greater degree of tolerance for false positive and negative results in mass screening compared with diagnostic testing in a hospital, the implications of incorrect results are significant because, even for a test with high sensitivity and specificity, large numbers of people would get a false positive (and be required to self-isolate) or a false negative (and potentially go on to infect other people).

The amount of virus a person has in their body is also relevant – this is called the viral load. Tests vary in how sensitive they are at detecting different viral loads and this viral load changes over the course of an infection. Typically, a nose and throat swab from someone very recently infected will contain very little virus, but this increases until it peaks a couple of days later and then declines as the person recovers. This is simply depicted in this diagram.

An important question is how sensitive does a test need to be in order for it to be used on largely asymptomatic people in mass testing. A recent evidence review of multiple studies examined whether there are differences between people who have symptoms and those who do not. It found evidence that that asymptomatic people clear the virus more quickly.

Some studies in the review reported no difference in the viral load between asymptomatic and symptomatic people but others did report a difference. Viral loads can be expressed by either stating what concentration of viral genetic material (called RNA) is required in a sample in order to detect it, or how many cycles of replication in the PCR test have to be done before the RNA is detected (Ct value).

The evidence review found that viable virus cannot be isolated from a sample if the level of virus is below 1,000,000 RNA copies in 1 ml of sample. However, using this or a Ct value as a threshold are not necessarily reliable ways to determine if someone is infectious.

One research study outlines a view that test sensitivity should be considered as secondary to the speed with which results can be obtained and if tests can be repeated frequently. This is based on modelling studies of different testing scenarios. The use of repeated and frequent testing to overcome the limitations of using tests with poor sensitivity was noted by SAGE in a consensus paper on mass testing dated 31 August.

UK Government evaluation of new point-of-care lateral flow tests

The Government is evaluating multiple rapid point-of-care tests, of which lateral flow tests are one type. Lateral flow tests are disposable cartridge style tests. A nose and throat swab fluid sample is applied to an absorbent pad and reacts with substances that show if the virus is present.

The Prime Minister has previously outlined how such tests could be used in numerous settings such as schools, prisons, travel hubs, and cultural and sporting venues as a way of screening people without symptoms to check if they have a current infection. In a statement on 23 November the Prime Minister outlined how mass testing with lateral flow tests will be rolled out in a range of settings over the next few months as part of the Government’s COVID-19 Winter Plan.

Public Health England (PHE) Porton Down and the University of Oxford are evaluating these tests for the UK Government. There is a 4-stage process, with tests progressing according to how well they meet specified standards. The fourth stage is not detailed on the Government website but has been described in a recently published test evaluation.

• Phase 1 – “desktop reviews” of products with potential (Department of Health and Social Care)
• Phase 2 – laboratory tests on prepared solutions containing controlled amounts of SARS-CoV-2 to establish test sensitivity and specificity and test kit failure rates (PHE Porton Down)
• Phase 3 – tests to establish how well the test works on real samples from people (research sites across England)
• Phase 4 – evaluation of how tests perform in real-world settings (including schools, universities, regional COVID-19 testing centres)

From 130 tests put forward by manufacturers, 40 progressed to stage 2. Of these nine met the criteria, with six tests progressing to more detailed evaluations in phases 3 and 4. Data so far show that four tests show moderate sensitivity (about 70%) when compared with highly sensitive PCR tests, and are better at picking up infections in people who have high viral loads. One test has been rolled out nationally, on the basis of an incomplete evaluation, and is discussed next.

The Innova SARS-CoV-2 Antigen Rapid Qualitative Test

The Innova SARS-CoV-2 Antigen Rapid Qualitative Test is in the latter stages of an independent evaluation, but some preliminary data were published on 11 November. Other data about the test comes from the manufacturer.

Innova claims that that test has 100% specificity and 96% sensitivity. This is based on their evaluation of its performance on 25 samples from symptomatic people infected with SARS-CoV-2 and 25 samples without the virus. PHE reports that it has 99.68% specificity. This would mean that for every 1,000 people tested, about 3 people would get a false positive result and subsequently be required to self-isolate. The test’s overall sensitivity was 76.8% but this varied according to who was performing the test:

• laboratory scientists: 79%
• healthcare workers: 73%
• self-trained members of the public following instructions: 58%

This means that the test will not pick up about a quarter of infected people – leading to false negative results.

Overall, these data show that test performance is dictated not only by a test’s physical features, but also how it is used in real life. The level of infection in a community is also important, this is explained later.

The test’s sensitivity diminishes significantly according to how much virus someone has (viral load). The data show that the test is more sensitive in people who have a higher viral load. There is at least a 90% chance that infections will be detected when viral loads are high (over 100,000 copies of viral RNA) but this decreases to less than 60% for lower levels of virus. The Innova test evaluation report states that there is little difference in the test’s ability to pick up virus in people whether or not they have symptoms but the sample of people tested was too small to draw meaningful conclusions that are statistically valid.

While tests like this can be a useful tool in community surveillance such as identifying areas with high disease prevalence, they may be inappropriate as a “test and release” tool for individuals because they are not accurate enough. The Innova manufacturer test kit information sheet sets out that a negative result does not rule out a COVID-19 infection.

Understanding COVID-19 test accuracy in mass testing

The accuracy of a test in the real-world use can be very different to the reliability reported by a test manufacturer. One factor influencing test accuracy is how common the infection is in the population. This is because we do not know what proportion of the population are infected. This can influence the likelihood of both false negative and false positive results.

The table below sets out the scale of false negatives and positives using a test with the sensitivity and specificity of the Innova test in a testing population similar in scale to the serial mass testing pilot in Liverpool (100,000) using COVID-19 prevalence rates reported for the city when mass testing began on 6 November (300 people per 100,000). Using these figures means that the chance of being infected is 0.3%.

This table shows the overall predictive accuracy if 100,000 people are tested with test with 76.8% sensitivity and 99.68 % specificity:

The test is very good at identifying almost all the people who do not have the infection (99.9%) but is less good at identifying the people who do (42%). This is why follow-up PCR tests are important to confirm positive results, but this will not catch the 70 people who test negative even though they are infected.

319 people are wrongly told that they are infected and would have to self-isolate until a PCR test is done. Finally, a negative result does not necessarily exclude COVID-19 – the person tested may be very early into the infection or they may have recently had it and their viral load has now reduced.

Mass testing pilots

Pilots of rapid tests have taken places in communities and healthcare settings. This has been followed by a national rollout of lateral flow tests. On 10 November the Health Secretary announced that 67 Directors of Public Health in Local Authorities in England have been offered test capacity for 10% of their population per week, as have the devolved administrations. Public Health Directors have the discretion to use the tests as they see fit in their communities. Some of these community testing pilots and examples of lateral flow tests in other settings are detailed below. A full list of Local Authorities using the tests is available.

Liverpool – repeat mass testing in a city

The Innova test was chosen by the Government for a 2-week community testing pilot in Liverpool (the Liverpool Mass Asymptomatic Serial Testing). Liverpool was chosen as it has had high levels of infection, with Tier 3 restrictions in force from 14 October.

Trained operators took nose and throat swabs and interpreted the results. Positive results were confirmed with a laboratory PCR test. Testing capacity is reported to be large enough to allow for the testing of 500,000 people every 7–10 days.

On the 20 November the Local Authority reported that 90,429 Liverpool residents and 19,585 people from neighbouring areas were tested with lateral flow tests. There were 788 positive results from lateral flow tests. It is not specified whether all the positive results were confirmed with a follow-up PCR test or what proportion were from asymptomatic people.

The pilot may run beyond the original 2-week period until the end of November. The subsequent evaluation will outline how well tests work in this context but there are no published details about how the pilot will be evaluated and what constitutes success.

Stoke-on-Trent – mass testing targeted at different groups

Staffordshire County Council announced a pilot of mass testing using lateral flow tests on 10 November. Rather than the city-wide approach taken in Liverpool, about 10,000 tests per week will be offered to several groups including 16–18 year olds in educational settings, care providers, and communities in which the infection is prevalent. The Local Authority will be evaluating the practicalities of testing in these settings which will help to inform how wider rollout could work best.

Universities

Targeted mass testing is also in progress in the university sector, part of which involves lateral flow tests. Managed by Public Health England, the University of Oxford and the Department of Health and Social Care, university pilots will determine how best to use the technology and how it can be managed operationally.

An example profile of a university’s approach to participating can be seen on the Durham University website. Oxford University has detailed how it will evaluate the use of tests that are being used in three colleges.

The study is called the Feasibility and Acceptability of community COVID-19 rapid Testing Strategies (FACTS). It will track how many people take up the offer of testing, how many test regularly, and how many cases of COVID-19 are detected.

Healthcare settings

The Health Secretary also announced that the Innova lateral flow test will be rolled out as part of twice-weekly testing of NHS staff across England. Staff will swab and test themselves, with any positive result confirmed with a laboratory PCR test.

Evaluation of mass testing pilots

There are no published protocols on how these individual pilots will be evaluated. Ideally a protocol would outline how large the pilots need to be in order to draw meaningful conclusions and how randomisation will be used to control for study bias.

It will be difficult to assess the impact of tests on transmission because there are no data as to whether the tests can single out those people who are capable of transmitting the virus.

Scientific advice on mass testing

A sub-group of the Scientific Advisory Group on Emergencies (SAGE) published an analysis of mass testing in a consensus statement on 31 August, reflecting on the technological, epidemiological and behavioural aspects and how such a programme should be distinct from but linked to test, trace and isolate programmes.

SAGE’s expert sub-group SPI-M-O’s latest report (11 November) considers how rapid tests could be used as a tool to reduce the duration of self-isolation. They note that there is uncertainty as to how well tests can detect the virus at different points during the course of an infection.

Scientific opinion on mass testing

There are mixed views about the value of screening for SARS-CoV-2 using mass testing. Some researchers think that mass tests provide a useful additional tool with which to monitor the transmission of the virus and to inform where interventions are needed.

However there is criticism of the Government’s approach to mass testing in respect of rolling out programmes before test evaluations are complete, a lack of transparent standards for evaluation, and the extent to which such a programme diverts resources from other approaches for which the evidence base is stronger (prioritising testing of symptomatic people to break transmission chains). There are also ethical concerns regarding a lack of clarity about informed consent and data privacy and sharing.

Photo by Medakit Ltd on Unsplash