Coronavirus (COVID-19) asymptomatic testing programme: evaluation - November 2020 to June 2021

Annex B: Asymptomatic testing. Evidence review of existing literature and current evaluations

Executive Summary

Introduction

In recent months, testing has increasingly been seen as a fundamental tool to detect pre- and asymptomatic transmission and contain the spread of Covid-19. Asymptomatic testing is a key part of the Scottish Government's testing strategy, as it is estimated that around 1 in 3 people have Covid-19 without displaying any symptoms. This asymptomatic testing strategy has mostly made use of lateral flow antigen tests (LFDs).

The purpose of this review is to provide an account of models of delivery, benefits, barriers, costs and impacts of asymptomatic testing regimes implemented both in the UK and the rest of the world, and to explore the lessons learned from these experiences.

Methods

This evidence review is based on the analysis of evaluation work on testing pilots conducted in the UK, and international scholarly research selected through a standard literature review. The body of evidence consists of 73 studies. A search was conducted using Google Scholar, ScienceDirect, PubMed, and KandE, a Scottish Government online search engine covering several databases. Priority has been given to sources published in 2021, due to the rapid changes in scientific knowledge and availability of different kinds of test during the Covid-19 pandemic. A major limitation of the material is the difficulty in clearly distinguishing the effects of the testing strategies implemented from those of the other non-pharmaceutical interventions introduced at a similar time. There are also no control groups for robust comparison in the evidence reviewed. Finally, a substantial number of studies are models, hence relying on assumptions that may not play out in real life. At the time of the last search (29^th June 2021), not all the scholarly research presented here had been peer-reviewed. Yet that has been included in this review as the process of formal publication can be lengthy and there is a need to examine the available findings immediately, given the limited time for providing advice and the rapidly changing area of research.

Research questions

The examination of the available evidence on testing strategies' outcomes is based on the following key research questions:

• What is the value/impact of the different testing models?
• What are the benefits of asymptomatic testing?
• What obstacles have testing regimes met?
• What are the financial and human costs of testing implementation?
• What are the practical and psychological impacts for the population tested?

Key findings

The different types of testing regimes can be categorised as follows:

• Universal testing offer. From 26^th April in Scotland and 9^th April in England, free lateral flow test kits could be ordered online or by phone for home delivery from government websites, or picked up from many local walk-in or drive-through test sites. These kits are meant for people who do not already have access to asymptomatic testing in their workplace, school or community.
• Surge testing. This strategy offers targeted testing to anyone in a given small population of high prevalence, knocking door-to-door or testing whole settings in response to an outbreak.
• Self-Collect model. This regime has been implemented in a number of workplaces, schools and universities. Kits can either be collected from a specific location or received at home.
• LFD testing with release for 24 hours. This model, part of the 'test-to-enable' strategy being evaluated in England, presents an alternative to self-isolation for contact cases and aims at maintaining essential services. Staff members test themselves using a LFD each day for seven days and, if the result is negative, they are released from the requirement to isolate and allowed to undertake essential activities for the following 24 hours when the next test is due. Testing through Mobile Testing Units (MTUs). This regime relies on MTU fleets to expand community testing provision and target hard-to-reach populations.
• Pooled testing. This strategy is mostly used in populations with low prevalence where pooling can be used to increase capacity and lower costs. By combining a number of samples into a pool and testing this pool using a single PCR test, laboratories can test more samples, at the same time, with fewer resources.
• Mass testing. This strategy involves the testing of the whole national population or the population of a specific extended area within a short period of time. It has been adopted both in high prevalence contexts to reduce transmission to more manageable levels, and in low prevalence settings to stop community transmission through early identification of cases.

Almost all the sources examined in this review mention the same benefits of asymptomatic testing, which can be summed up as:

• Rapid detection of hidden cases. Asymptomatic testing serves the aim of identifying cases that would not otherwise have been detected in the absence of symptoms, or identifying them earlier than they otherwise would have been via PCR-based testing once symptomatic. By increasing the frequency of repeat testing over short periods of time, LFD tests have the potential to detect a higher number of cases. They also provide near instantaneous results, hence facilitating the timely isolation of the most infectious cases who may otherwise transmit infection while waiting for a PCR result (this can take up to 6 days).
• Participant wellbeing. LFD home testing reduces access barriers to testing (for instance, for older people or those with mobility concerns) and health risks associated with venue-based testing due to viral exposure from/to others. Asymptomatic testing based on the 'test-to-enable' model allows participants to benefit from being able to continue working and from reduced chances of having to self-isolate. Finally, negative test results can provide reassurance.
• Time and cost savings. LFD test kits are relatively inexpensive and provide results rapidly. Depending on the model of delivery, asymptomatic testing also has the potential to reduce unnecessary self-isolation when contact-traced, hence reducing work absences and the costs of sick pay associated with them. Furthermore, it can reduce the strain on laboratories that conduct PCR–based testing and cut healthcare costs by reducing the need for trained providers. Finally, it can positively impact on infections and hospitalizations.

The sources examined in this evidence review report some obstacles met by testing regimes that can impede their success. These are:

• Reluctance in taking LFD tests. Participants in a number of trials reported concerns regarding the tests, which they considered invasive and difficult to self-administer. Moreover, swabbing younger children and people with disabilities has been seen as an emotional and physical challenge for which many felt unprepared.
• Failure to upload test results. Testing participants may not upload their results or may upload them incorrectly.
• Risk that positive LFD tests are not followed by PCR tests. This could be due to a number of reasons, ranging from lack of time to lack of faith in test reliability.
• Failure to perform the test correctly. Self-testing requires skill, and swabs taken by untrained individuals are more likely to give false negative results.
• Lack of knowledge and misconceptions about testing. Some pilots report that individuals find it difficult to understand differences between PCR and LFD tests, and why a test is needed in the absence of symptoms. Lateral flow tests are more likely to detect positive cases when viral loads are highest and patients are most infectious. However, as this window is narrow, they are most suitable when testing is frequent. Informing the public about the reasons for frequent testing (every 2–4 days) may improve testing uptake.
• Perception of being low risk. For those groups who already feel their risk is low, regardless of their objective risk, negative results could falsely reassure people there is no risk of being infectious, reduce adherence to the guidelines, and increase the spread of Covid-19 as a result. In particular, perceiving oneself to be low risk can apply to those who have received their vaccine, have had Covid-19 in the past or to younger age groups.
• Language and digital access barriers. Testing participants might find it hard to interpret information on testing due to both language and digital access barriers. Some pilots suggest this could account for the lower uptake rate for potential participants from BAME and other backgrounds.
• Financial and psychological barriers to receiving a positive test. People may have concerns at having to self-isolate, being unable to work, the impacts on their household life, and being stigmatised if they test positive. The risk of false-positive findings may also lead to needless isolation and unnecessary psychological distress. For individuals within deprived areas, self-isolation tends to have a direct impact on remuneration and employment security.
• Lack of trust. A lack of trust in government has been identified as one of the main reasons some choose not to take part in testing programmes. Government intentions about use of personal data have been identified as a cause of concern, especially in deprived areas.
• Test access. Reluctance to get tested might also pertain to concerns about testing centre location, time lost taking the test, concerns over queuing and the logistics of home testing.
• Inadequate training or excessive work burden. A number of studies have reported how the testing programmes in the workplace have resulted in added responsibilities and increased workload for members of staff. A lack of guidance and potential loss of knowledge through cascade training have also been identified as barriers.

As for the costs involved in sustained testing regimes, these include:

• Costs of test kits. LFD devices are an affordable alternative to PCR tests. However, some participants might find it difficult to do the test properly. This means multiple swabs might be needed and unopened test kits wasted. The provision of a stock of swabs together with test kits could solve the issue.
• Costs to run the model. This includes the costs of setting up and administering the programme, including personnel training and equipment, and use of facilities and services. Estimates of the cost-benefit ratio of testing programmes vary per country. An article based on estimates published in the British Medical Journal calculated in March 2021 that, if tests delivery costs in England were £10-20 and only one test in 1500 comes back positive, that would amount to £15 000-£30 000 to detect one case, with a risk that this could be a false positive.^[20] However, the evaluation of the Welsh pilot in Merthyr Tydfil and the lower Cynon Valley found the intervention cost effective, with a central estimate of £2292 per QALY (quality-adjusted life years) gained.^[21]
• Time and human costs. Regular asymptomatic testing has the potential to avert infections, hence reduce workdays lost due to sickness. On the other hand, a false positive test could result in staff being removed from the workforce, unnecessary tests and possible isolation for colleagues linked to them. This could exacerbate staffing shortages and require further resources and time to manage suspected outbreaks.

The existing literature offers important reflections on attitudes towards and experiences of testing in relevant testing populations:

• Willingness to participate was high. Participants reported multiple reasons for wanting to take part in testing regimes, such as the desire to know whether they were infected, a sense of duty to society as a whole, and a feeling of obligation to keep working and help tackle the virus.
• Testing uptake and compliance with recording results is mixed. Some of the pilots show encouraging uptake and engagement, but others find barriers remain.
• Personal experiences of testing are complex. The sources analysed report a number of challenges: booking errors, a lack of slots in specific areas, and difficulties in reaching testing centres. Moreover, people seem to see testing as a process rather than a discrete technical event, entailing a significant burden of time, energy, and resources for the individual and their relatives/friends.

Lessons learned

Some of the lessons learned from the experiences within the sources examined are:

• Testing communications should be clear about the nature and benefits of testing. It is key that everyone understands that asymptomatic cases can still spread the virus, and that the confusion around the need to continue testing after being vaccinated or having Covid-19 is addressed. Concerns about test accuracy should also be targeted, and more information about different kinds of tests provided. Informing the public of the rationale behind testing frequently (for instance, to every 2–4 days) would address this concern and might improve testing uptake. Communication should also promote awareness of what test results mean, and particular attention given to negative results. It is important to avoid conflicting messaging that may confuse the public. In England, people who received a negative result were told it was "great news" in some settings, while in others that "you were not infectious when the test was done", with varying time periods suggested for continued testing.
• Involving community leaders and stakeholder organisations in the development and implementation of testing programmes could help build trust, share goals, and bridge cultural and language gaps. Some pilots found that local community leaders and stakeholders had a fundamental role in determining behaviour change. Local organisations were perceived as answerable to local people, hence more trustworthy than national, more 'faceless' organisations such as NHS Test and Trace in England.
• Less invasive sampling techniques could increase uptake. Willingness to participate in testing could be higher if less invasive sampling techniques (such as saliva sampling) are provided.
• The context of use requires careful consideration. The disconnect between the prescribed testing regime and the actual context of use should be addressed. Some testing regimes requiring employees or employers to get tested multiple times a week are not compatible with their working schedule and pose a high risk of increased staff dissatisfaction, and consequently staff turnover and burnout. In some settings, on-site testing can be organisationally complex and resource-intensive.
• Equitable access to tests should be promoted. Individuals who face additional barriers to testing, such as language barriers and/or digital exclusion, should be addressed by tailored campaigns. People should be told how to collect, book or perform a test by means of a range of media channels and formats. In low-literacy populations that might have trouble understanding written or graphic instruction materials, the use of online videos could be an option. On the other hand, as internet-based dissemination may limit access for some segments of the population and require a proactive individual seeking information, it is key to explore other routes, such as promotion of testing through commercial sites or community-based organisations.
• Psychological and behavioural consequences of test results may impact uptake. Testing programmes rely on members of the public undertaking a substantial burden of responsibility across the testing stages. Using a language that acknowledges the challenges people face and emphasising the contributions of individual actions to a societal response may increases uptake.
• Emphasising civic duty and the altruistic motivations of engaging in testing may be successful strategies to promote testing. A communication strategy that focuses on protecting others and reducing the impact of the pandemic for society as a whole has already demonstrated its value in other areas, such as vaccination and compliance with regulations.

Research on the underlying motivations that lead people to get a test when asymptomatic would aid the design of effective health communication and successful implementation of testing strategies. Exploring the reasons for use of testing could shed further light on how individuals in Scotland are using the testing pathways open to them (see Asymptomatic Testing Evaluation for more details on testing behaviours and motivations).

Asymptomatic Testing – An Evidence Review of the Existing Literature and Current Evaluations

Introduction

In recent months, testing has increasingly been seen as a fundamental tool to detect pre- and asymptomatic transmission and contain the spread of Covid-19. The Scottish Government has set out its approach to testing in Scotland's Testing Strategy, as part of the wider set of public health measures for the management of the pandemic^[22]. With the implementation of a range of testing regimes in different settings, a number of studies have assessed the barriers to intended outcomes from these regimes and suggested strategies to improve effectiveness. The purpose of this review is to provide an account of these studies carried out both in the UK and the rest of the world. This review also aims to explore the lessons that have been learned from previous and current experiences, and to examine what recommendations have been made to promote the adoption of successful testing interventions.

Background

Asymptomatic testing has been implemented as an infection control measure in a number of countries as part of the response to Covid-19. It has mostly made use of lateral flow antigen tests (LFDs) which can provide results within 30 minutes and significantly reduce costs and waiting times associated with polymerase chain reaction (PCR) testing. LFDs have also been adopted with the purpose of detecting and containing hidden infections, as it is estimated that around 1 in 3 people have Covid-19 without displaying any symptoms. The detection of asymptomatic individuals is meant to ensure prompt isolation of infectious cases, while also protecting others from being infected and maintaining vital services. As a negative result does not mean that there is no risk of being infectious, people should continue to follow the guidelines and restrictions in place.

Methods

This evidence review is based on a process of searching for and assessing material which can be divided into two main categories: evaluation work conducted in the UK and scholarly research.

The evaluation material has been provided by a variety of government sources and analyses just recently concluded or ongoing testing pilots across the UK. The majority of these evaluations are process evaluations, namely they include ongoing insight to help reassess the work and improve the intervention being delivered, rather than summative evaluations conducted at the end of delivery. These have been conducted mostly using quantitative methods (such as surveys or data analysis), although some work has also relied on qualitative methods (such as interviews and focus groups).

As for the scholarly research, a standard literature review has been undertaken, searching for relevant literature from across the globe. A first search was conducted using KandE, a Scottish Government online search engine covering several databases. Subsequently, a search was carried out on Google Scholar, ScienceDirect and PubMed. Search terms included "testing", "mass testing", "Covid", "coronavirus", "asymptomatic", and "evaluation". Further references have been added by means of the snowballing technique, where references in relevant studies are reviewed for additional evidence. Only sources pertaining to Covid-19 (as opposed to testing regimes for other disease) have been included.

Priority has been given to sources published in 2021, due to the rapid changes in scientific knowledge and availability of different kinds of test during the Covid-19 pandemic, though some relevant material from 2020 is also included. Evaluations were prioritised for review to assess practical models and their implementation. For the purposes of the current project, assessment of reliability and scientific comparison between different kinds of test are also out of scope.

The body of evidence selected in this reviews consists of 73 studies. Most of the scholarly research uses robust research methods. Yet, a substantial number of the evaluations included in this evidence review are models, rather than evaluations of implemented testing regimes, hence relying on assumptions that may not occur in real life. Together with this, another limit of the material remains the difficulty in clearly distinguishing the effects of the testing strategies implemented from those of the other non-pharmaceutical interventions introduced at a similar time. There are also no control groups for robust comparison in the evidence reviewed^[23]. Finally, not all the scholarly research presented here has been peer-reviewed: some was in the form of preprints at the time of the last search (29^th June 2021). Nonetheless, that has been included as the process of formal publication in a scholarly journal can be lengthy and there is a need to see and discuss the available findings immediately, given the limited time for providing advice and the rapidly changing area of research.

Although the geographical coverage of the scholarly research material was international, this review was limited to resources in English. This partly justifies the higher number of UK based studies included in this review, the other reason being the wider use of LFD tests in the UK compared to other countries.

The evidence discussed in this review should be considered in the context of these limitations.

Terminology

Effective Covid-19 control is increasingly relying on testing of pre- and asymptomatic cases. Generally, regular testing has been adopted in specific settings, for instance in schools and care homes. More recently, free lateral flow test kits have been made available for everyone in Scotland who does not have Covid-19 symptoms as the "universally accessible offer". It has to be noted that there is not an agreed definition for "universal testing" in the existing literature and in some studies this is used to identify testing regimes targeting specific groups of people (e.g. students, healthcare workers, or prisoners)^[24]. These ambiguities also extend to the variability in the use of the definition of "mass testing", sometimes used to refer to the national population and again also sometimes to specific categories of people. In this review the term "universal" is used to refer to the universally accessible offer, or equivalent testing strategies, that give any individual in the community the opportunity to get free tests on a regular basis.

Research questions

The examination of the available evidence on testing strategies' outcomes, in terms of the successes and barriers to implementation and/or uptake is based on the following key research questions:

• What is the value/impact of the different testing models?
• What are the benefits of asymptomatic testing?
• What obstacles have testing regimes met?
• What are the costs of testing implementation (for example, costs for the government, impact on the population)?
• What are the practical and psychological impacts for the population tested?
• Are hard-to-reach (e.g. non-digital, geographically remote) individuals able to access the tests and how do we ensure equal accessibility?

Key findings

This section will report key findings both from the evaluation work conducted in the UK and the scholarly research on testing examined here. First, it will introduce the different kinds of testing regimes that have been adopted in a number of settings and countries, together with their models of operation and delivery. Then, it will focus on the benefits of asymptomatic testing, the barriers to it and the costs involved in its implementation. Finally, it will explore attitudes towards and experiences of testing in relevant populations.

The different types of testing regimes can be categorised as follows:

• Universal testing offer. From 26^th April in Scotland and 9^th April in England, free lateral flow test kits could be ordered online or by phone for home delivery from government websites, or picked up from many local walk-in or drive-through test sites. These kits are meant for people without Covid-19 symptoms who do not already have access to asymptomatic testing in their workplace, school or community^{[25] [26]}. Each kit contains 7 LFDs. The NHS recommends taking a test twice a week (every 3 to 4 days) to check if one has the virus^[27].
• Surge testing. This strategy offers targeted testing to anyone in a given small population of high prevalence, knocking door-to-door or testing whole settings in response to an outbreak. It has the potential to find cases early and reduce onward transmission and spill over into the wider community, hence reducing overall community transmission^[28].
• Self-Collect model. This regime has been implemented in a number of workplaces, schools and universities. Kits are offered by the employer or the organisation and can be either collected from a specific location or received at home.
• LFD testing with release for 24 hours. This model, part of the 'test-to-enable' strategy being evaluated in England, presents an alternative to self-isolation for contact cases and aims at maintaining essential services. In this model, staff members test themselves using a LFD each day for seven days and, if the result is negative, they are released from the requirement to isolate and allowed to undertake essential activities for the following 24 hours when the next test is due. A two-arm randomised control trial of a sample of contacts of confirmed cases of Covid-19 has also been run in England to test the hypothesis that daily testing with lateral flow devices with release for 24 hours is non-inferior to a single PCR test and isolation for 10 days^[29].
• Testing through Mobile Testing Units (MTUs). This regime relies on MTU fleets to expand community testing provision and target hard-to-reach populations. The Targeted Community Testing pathway in Scotland, for example, widely utilised mobile testing units, buses and lorries, and other temporary locations set up as asymptomatic testing sites to provide an adaptable and rapid response in specific circumstances (e.g., with their deployment to places or organisations with an outbreak or in locations lacking suitable venues).
• Pooled testing. This testing regime is mostly used in populations with low Covid-19 prevalence where pooling can be used to increase capacity and lower costs^[30]. By combining a number of samples into a pool and testing this pool using a single PCR test, laboratories can test more samples, at the same time, with fewer resources (reagents and personnel time)^{[31] [32]}. This testing model has been used in the past for the detection of the human immunodeficiency virus and hepatitis B/C viruses in blood products. It has the capability of providing a good estimate of the actual incidence of a virus, and the heterogeneity of this incidence in terms of geography and age^{[33] [34]}.

Pooled testing of asymptomatic individuals has been adopted in England, where a pilot is currently exploring its use in university student households^[35], and in Scotland at the end of 2020 for a 'proof-of-concept' pilot. Pooled testing is currently being evaluated within NHS Scotland for its potential use in large-scale testing.

Although pooled testing can contribute to informing public health policy and resource allocation, it presents a particular barrier when applied to the current pandemic: as each individual belonging to a positive pool is asked to self-isolate even if they are negative, this might result in lower compliance rates^[36].

• Mass testing. This strategy involves the testing of the whole national population or the population of a specific extended area within a short period of time. It has been adopted both in high prevalence contexts to reduce transmission to more manageable levels, and in low prevalence settings to stop community transmission through early identification of cases. Examples of mass testing are the campaigns implemented in Slovakia^[37], the Faroe Islands^[38], in Liverpool^[39], in Merthyr Tydfil and in the lower Cynon Valley (Wales)^[40] and the Italian region of South-Tyrol^[41].

Mass testing presents limitations: first, when the screening begins, approximately half of the infected individuals will be in the latent phase so won't test positive^[42]; secondly, the programme would have to be run on a regular basis and relatively frequently if a sustained mitigation of the pandemic were to be achieved, posing a number of issues in terms of feasibility and cost-effectiveness^{[43] [44]}. As shown by the Slovakian case, the immediate benefits can be quite substantial: the intervention resulted in a 60% decline in infection prevalence within one week (or 80% in two weeks), while primary schools and workplaces were mostly open, suggesting that such a rapid drop in numbers was linked to the mass testing campaign^[45]. However, critics pointed out that the daily positivity rate for PCR tests 2-3 weeks after the intervention was roughly the same as it was before it, and questioned the possibility of further rounds that could exhaust the already stretched capacity of medical workers and cause a significant and unnecessary interruption to society^[46].

Almost all the sources examined in this review mention the same benefits of asymptomatic testing, which can be summed up as:

• Rapid detection of hidden cases. Testing asymptomatic individuals serves the aim of identifying cases that would not otherwise have been detected in the absence of symptoms (current estimates indicate these to be about a third of infections^[47]), or identifying them earlier than they otherwise would have been via PCR-based testing once symptomatic. As a result, transmission could be rapidly reduced and quicker easing of lockdown measures allowed. A number of studies have explored the benefits of asymptomatic testing, highlighting how strategies based on symptom screening could miss between 40%-100% of infected people depending on setting and disease prevalence^[48].

LFD tests can provide near instantaneous results, hence facilitating the timely isolation of the most infectious cases who may otherwise transmit infection while waiting for a PCR result^{[49] [50]}. Indeed, it can take up to 4-6 days to receive PCR test results, as the process requires several steps such as ordering the test, carrying it out, mailing it to the laboratory and waiting for the testing procedures to be carried out^[51]. By increasing the frequency of repeat testing over short periods of time, LFD tests have the potential to detect a higher number of cases^[52]. Furthermore, due to its high sensitivity, PCR testing tends to detect viral shedding long after the infectious period (a mean of 17 days), potentially affecting return to work or school^{[53] [54]}.

• Participant wellbeing. LFD home testing reduces access barriers to testing (for instance, for older people or those with mobility concerns) and health risks associated with venue-based testing due to viral exposure from/to others. Asymptomatic testing based on the 'test-to-enable' model allows participants to benefit from being able to continue working and from reduced chances of having to self-isolate. Finally, negative test results can provide reassurance.

Qualitative research on a pilot programme implemented in Southampton using saliva samples collected at home shows that participants expressed a sense of relief and reduced feelings of anxiety when they tested negative. They felt that this enabled a "near normal" life to continue and were more confident going to school or work, or visiting vulnerable family and friends, knowing they were not spreading the virus^[55].

• Time and cost savings. Asymptomatic testing may be a cost-effective strategy. LFD test kits are relatively inexpensive, do not require laboratories and provide results rapidly. Their ability to provide near instantaneous results avoids the delays associated with PCR tests and facilitates the timely isolation of infectious cases by shortening the length of time between initial infection, test results, and preventive action^[56]. Depending on the model of delivery, asymptomatic testing also has the potential to reduce unnecessary self-isolation when contact-traced, hence also reducing work absences and the costs of sick pay associated with them. Furthermore, it can reduce the strain on laboratories that must conduct PCR–based diagnostic testing and cut healthcare costs by reducing the need for trained providers^[57]. Finally, it can positively impact on infections, mortality and hospitalizations. Modelling from the Slovakian case shows that, without the nationwide testing, the ICU bed occupancy could have almost doubled in the worst-case scenario^[58].

The sources examined in this evidence review report some obstacles met by testing regimes that can impede their success. These are:

• Reluctance in taking LFD and PCR tests. Participants in a number of trials reported concerns regarding LFD and PCR tests, which they considered invasive and difficult to self-administer. Furthermore, swabbing younger children and people with disabilities has been seen as an emotional and physical challenge for which many felt unprepared^[59].

Saliva tests have been indicated as a preferable alternative. Qualitative evaluation of a Southampton-based pilot showed higher compliance when swab tests were replaced by saliva sampling. Making participation as convenient and easy as possible was key to increasing uptake, with parents reporting that the test was simple enough for children to take responsibility for carrying out the tests themselves^[60]. Saliva tests are also a stable way of testing: the tube for the sample contains preservatives and additives that kill the virus while preserving its RNA. This means the sample doesn't retain infectious particles unlike a swab^[61].

• Failure to upload test results. Testing participants might not upload their results or may upload them incorrectly. A pilot conducted in care homes in Liverpool identified a lack of void LFD tests recorded online. This suggests that void tests were not uploaded or incorrectly uploaded as negative results to the testing system^[62]. This barrier raises questions about the validity of collected data and the risk of spreading the virus, especially if positive results are not uploaded. It also invites question about the reasons why people fail to upload their results and whether this happens more frequently with negative results.
• Risk that positive LFD tests are not followed by PCR tests. Testing participants might fail to take a PCR test following a positive LFD test for a number of reasons, ranging from lack of time to lack of faith in test reliability.
• Failure to perform the test correctly. Self-administering the test has the potential to reduce demand on trained personnel, transmission risk in the process of sample collection, reduce strain on laboratories performing PCR tests and guarantee increased access to frequent testing. However, these benefits need to be weighed against the potential loss of sensitivity. Self-testing requires skill, and swabs taken by untrained individuals are more likely to give false negative results.

Although data from Germany suggests that self-administered tests have very similar sensitivities to those achieved by professionals, even when people deviated from the instructions^[63], other studies show a different picture. When the UK Government implemented a pilot with LFDs to support mass population testing and to open care homes to visitors in November 2020, a drop in sensitivity was recorded: test sensitivity was 48.89% when self-administered, compared to 73% when carried out by trained healthcare workers^[64]. This loss of sensitivity depending on the person performing the test had already been observed by Public Health England in its evaluation of the Innova test: a sensitivity of 79.2% when used by trained laboratory scientists, 73% when used by trained healthcare staff, but only 57.5% when used by Test and Trace centre staff employed by the pharmacy chain Boots^[65]. Performance may improve with experience, as people become more familiar with tests over time.

• Lack of knowledge and misconceptions about testing. Some pilots report that individuals find it difficult to understand differences between the PCR and LFD tests, why a test is needed in the absence of symptoms, and how and where to get tested^{[66] [67] [68]}. Qualitative research conducted among contacts of confirmed Covid-19 cases who were offered the option of daily testing highlighted how for some participants, doubts about the accuracy of LFD tests led them to prefer self-isolation. Participants, especially those living with a positive index case, were also concerned about the possibility of contracting or transmitting the virus within the 24 hour window between tests^[69]. These concerns regarding asymptomatic testing often originated from media coverage and ongoing debates about their accuracy.

The use of LFDs has divided the medical and scientific community, between those who believe that the tests may miss so many infections that they could cause more harm than good, and those advising that using rapid antigen (LFD) tests frequently would make them effective at stemming the tide of a pandemic^{[70] [71]}. Analysis using mathematical modelling suggested that a strategy using tests with lower sensitivity could be as effective as relying on more sensitive tests (PCR) used less frequently^[72]. Lateral flow tests are more likely to detect positive cases when viral loads are highest and patients are most infectious, usually one to three days before the onset of symptoms and during the first five to seven days after their onset. This means that false negative results can arise in people tested before the viral antigen shed in the nose and throat is sufficient to be detected. It needs to be noted that despite a higher risk of false negative results, the rapid increase in viral shedding after the incubation period should leave only a short period when there will be a substantial difference between the point when you get a first positive result from a PCR test compared with a LFD^[73]. As the window for using lateral flow tests to detect infectious cases is narrow, they are most suitable when testing is frequent^{[74] [75]}. Informing the public that increasing the frequency of testing (for instance, to every 2–4 days) helps discover false negative cases (i.e. genuine positives) might improve testing uptake.

• Perception of being low risk. For those who already perceive their risk as low, regardless of their objective risk, negative results could falsely reassure there is no risk of being infectious, reduce adherence to the guidelines, such as becoming less vigilant in applying social-distancing and hygiene measures, and increase the spread of Covid-19 as a result^{[76] [77] [78] [79]}. In particular, perception of being low risk can apply to those who have received their vaccine or have had Covid-19 in the past.

When the University of Illinois implemented a campaign in August 2020 aiming at screening everybody on campus twice a week with saliva tests, they modelled that students were going to go to parties and that they probably weren't going to wear masks, but what they didn't expect was that people would choose to go to a party even if they knew they were positive^[80]. This suggests perception of being low risk in certain age groups, maybe driven by the awareness that younger people tend not to have such negative outcomes as older age groups from Covid-19. It could also be hypothesised that students perceive proximity to their peers as low risk, especially if they live in shared accommodation, and that social responsibility is less relevant to them while they are in a campus environment, hence explain the high prevalence of Covid-19 outbreaks in universities across the UK^[81].

Some research presents more positive data. A pilot study conducted in England on close contacts of index cases who were offered the option of daily testing found that only 13% of those who tested negative reported engaging in more high risk activity than prior to testing. Indeed, 58% reported having fewer risky contacts than they had before they were contact traced^[82]. The qualitative part of the study highlighted how a considerable number of participants were still reluctant to leave their homes despite consenting to take daily tests, and restricted their behaviour more than they had prior to testing^[83].

• Language and digital access barriers. Testing participants might find it hard to interpret information on testing due to both language and digital access barriers. These difficulties apply to the end-to-end journey, from ordering a test to the message containing the results^[84]. This could account for the lower uptake rate for potential participants from BAME and other backgrounds in some of the pilots, and suggests a need to develop materials and campaigns addressing them^{[85] [86]}. For instance, when contacts of index cases were offered LFD testing with release for 24 hours in England from December 2020 to January 2021, individuals of Asian ethnicity and residing in the two most deprived IMD deciles were more likely to decline the offer. Of those who participated in the programme, individuals from BAME groups were less likely to report a result^[87].
• Financial and psychological barriers to receiving a positive test. People may have concerns at having to self-isolate, being unable to work, the impacts on their household life, and being stigmatised if they test positive. Students from a higher education institution in England who participated in a qualitative study expressed a sense of guilt if their household had to self-isolate because of them and feared the interpersonal conflict this situation could bring. Therefore, despite seeing testing as an important national strategy to manage the pandemic, they were not always willing to participate in the testing campaign^[88].

For residents within deprived areas, self-isolation tends to have a direct impact on remuneration and employment security. Furthermore, although financial and practical support is available, people are not always aware of it. People might be reluctant to participate in a testing campaign if they find that their private costs outweigh any social benefits from not infecting others. Qualitative research conducted in England on the 'test-to-enable' model showed that a small number of participants declined testing when they had concerns that a positive test could potentially extend the standard 10-day isolation period^[89].

It is noteworthy that 52% of participants in a research study on contacts of index cases offered the option of daily testing in England reported being more likely to share details of people that they had been in contact with following a positive test result, if they knew their contacts would be offered the same option^[90]. This suggests potential higher compliance with the contact tracing programme and better case finding.

Many studies have also reported that individuals wilfully ignore medical diagnoses when they are torn between what they think they should do and what they want to do. Some might not want to know their health status regarding Covid-19^[91].

Reluctance to get tested might have increased with media coverage on the unreliability of LFD tests^[92]. Moreover, the risk of false-positive findings is not inconsequential, as they may lead to loss of work, needless isolation, separation from family members, and unnecessary psychological distress^{[93] [94] [95]}.

• Lack of trust. A lack of trust in government has been identified as one of the main reasons some choose not to take part in testing programmes. Government intentions about use of personal data have been identified as a cause of concern, especially in deprived areas, where social and economic inequalities often determine people's perceptions of the government and formal bodies^[96].

Researchers evaluating the pilot using saliva samples in Southampton reported that many of those who declined to take part in the testing regime were anxious about the risks associated with data transfer to NHS Test and Trace in the event of a positive test. On the other hand, the local NHS Foundation Trust was trusted, suggesting that local organisations would increase testing uptake if they run the programme^[97]. Hence, maintaining trust in government is key to promoting participation in testing.

• Test access. Reluctance to get tested might also pertain to concerns about testing centre location, time lost taking the test and the logistics of home testing. The voluntary mass testing scheme implemented in the Italian region of South-Tyrol in November 2020 revealed how individuals were more likely to get tested in communities where there were more centres and access to them was convenient^[98]. The evaluation of the asymptomatic testing pilot in Liverpool included a rapid thematic analysis of local narratives through local community media and social media and revealed that key barriers to participation were accessibility and concerns over queuing, sometimes despite advanced booking^[99].
• Inadequate training or excessive work burden. A number of studies have reported how the testing programmes in the workplace have resulted in added responsibilities and increased workload for members of staff. Senior representatives of the organisations involved in the Southampton pilot on saliva testing suggested that a 'toolkit' of instructions and tips for those implementing the programme could have helped manage the expectations of both staff and participants^[100]. The staff involved in a pilot for personnel and visitors of care homes implemented in Liverpool showed poor adherence to LFD testing protocols (the majority completed less than a third of the tests specified). A potential loss of knowledge through cascade training and test regimens complicating workflows of already over-burdened staff were identified as two main obstacles^[101]. Research conducted in October 2020 in Germany in care homes noted how a regular testing regime might require substantial additional staffing and resources. The existing healthcare workers in the care homes examined were already too stretched to engage in further tasks, such as testing visitors. Furthermore, getting tested before the beginning of a shift meant that staff needed to start work 30 minutes early and be paid for the extra time^[102].

Although testing has been recommended as a key public health strategy, several countries have reported the challenges posed in terms of the costs involved in sustained testing regimes^[103]. Most of the studies examined in this review highlighted:

• Costs of test kits. LFD devices cost governments about £3.50 per unit (converted from 5 euros/dollars in the articles), making them an affordable alternative to PCR tests^{[104] [105]}. However, some participants might find it difficult to do the test properly (for example, correctly following the requirement to test at the tonsil area without touching elsewhere). This means multiple swabs might be needed and unopened test kits wasted. The Under-Represented Groups Team (URGE) at NHS Test and Trace recommended the provision of a stock of swabs together with test kits to solve the issue^[106].
• Costs to run the model. This includes the costs of setting up and administering the programme, including personnel training and equipment, and use of facilities and services. The Operation Moonshot for mass testing implemented in England is said to cost £100 billion and represent the equivalent of 77% of the NHS annual revenue budget^[107]. Estimates of the cost-benefit ratio of testing programmes vary per country. In the US, one study estimated that the economic benefits of testing were about 30 times its cost^[108], while another concluded that the increase in GDP resulting from the testing programme ranged from 2 to 8 times the incremental cost of the tests^[109]. Analysis of asymptomatic testing conducted in Spain calculated a cost-benefit ratio of 7 to 19 in one study^[110] and a social return of €1.20 on the investment of €1 in another^[111]. For England, some more critical voices calculated (from estimates) in March 2021 that, if tests delivery costs £10-20 and only one test in 1500 comes back positive, that would amount to £15 000-£30 000 to detect one case, with the risk that it could be a false positive^[112]. The evaluation of the Welsh pilot in Merthyr Tydfil and the lower Cynon Valley found the intervention cost effective, with a central estimate of £2292 per QALY (quality-adjusted life years) gained^[113].
• Time and human costs. Regular asymptomatic testing has the potential to avert infections, hence reduce workdays lost due to sickness^[114]. On the other hand, a false positive test could result in staff being removed from the workforce, unnecessary tests and possible isolation for colleagues linked to them. A study on routine asymptomatic testing of long-term care staff in Ontario suggests this could exacerbate staffing shortages and require further resources and time to manage suspected outbreaks: anecdotal reports reveal staff intention to leave or thoughts about exiting the sector due to the stress caused by regular testing, especially as staff members are not consistently paid for time to get tested^[115]. This paper also reports that staff resources needed to operationalize the lateral flow testing strategy in long-term care facilities are much higher than what would be required with a PCR-based testing regime, with estimates of each home requiring an additional two full-time employees for this implementation^[116].

The existing literature offers important reflections on attitudes towards and experiences of testing in relevant testing populations:

• Willingness to participate was high. Participants reported multiple reasons for wanting to take part in testing regimes. Some mentioned the desire to know whether they were infected, and a feeling of obligation to keep working and help tackle the virus^{[117] [118]}. A rapid qualitative study conducted in Lothian showed how undergoing testing was often seen as a duty not only to loved ones, but to society as a whole. Joining a nationwide testing programme gave participants a sense of civic duty and of contributing to a collective pandemic response. Testing was also valued for providing personal reassurance and enabling social intimacy and freedom of movement^[119]. A mixed methods study exploring the experiences and perceptions of mass testing of students at the University of Edinburgh reported that the decision to participate in the programme was motivated by trust in the university's guidance and/or the desire to know they were not infectious and could travel and mix with family safely^[120].
• Testing uptake and compliance with recording results are encouraging in some of the evaluation pilots. Research from England reports that 51.1% of contact cases offered serial testing as an alternative to self-isolation accepted the offer, with high compliance with self-reporting LFD results^{[121] [122]}. The pilot targeting the population in Merthyr Tydfil and the lower Cynon Valley recorded an uptake of 49 and 56% respectively in the two areas^[123], while the Liverpool-based pilot reported a 57% uptake of testing^[124]. During mass testing in the Italian region of South Tyrol, 69.7% voluntarily decided to get an LFD test^[125]. However barriers to uptake and engagement remain, as shown throughout this review.
• Personal experiences of testing are complex. While testing is usually presented as a straightforward process in government documentation, personal experiences of testing report a number of challenges and issues: booking errors, a lack of slots in specific areas, and difficulties in reaching testing centres. Moreover, people seem to see testing as a process rather than a discrete technical event, entailing a significant burden of time, energy, and resources for the individual and their relatives/friends. This process includes different stages: weighing up information from multiple sources, interpreting ambiguities in testing criteria, navigating online bureaucratic systems, organising testing logistics, managing uncertainties around results, matching government guidelines to individual circumstances, and handling the repercussions of test results^[126].

Lessons learned

This review aimed at gaining an understanding of what testing strategies have been implemented and the benefits and obstacles to success they have met. It has explored evaluation work and other relevant literature to assess the economic and societal potential and costs of testing regimes. Some of the lessons learned from previous and current experiences within the literature are:

• Testing communications should be clear about the nature and benefits of testing. It is key that everyone understands that asymptomatic cases can still spread the virus, and that the confusion around the need to continue testing after being vaccinated or having Covid-19 is addressed. Concerns about test accuracy should also be targeted, and more information on different kinds of tests, their reliability, and the rationale and importance of testing regularly provided. Data from the pilot programme on the use of saliva samples developed by the University of Southampton and Southampton City Council show how participants emphasised the need for open communication of the reasons they should register for the programme in order to promote transparency and trust, and help dispel myths, particularly about the accuracy of the tests. The pilot also implemented educational engagement activities among students, whose increased knowledge appeared to make them more engaged with the programme and inclined to think they had more agency in controlling the spread of the virus and its damaging consequences^[127].

Communication should also promote awareness of what test results mean, and particular attention given to negative results: a drop in adherence to behaviours that reduce transmission due to lowered risk perception could affect the overall effectiveness of testing programmes. Experimental research carried out in Germany showed that testing negative decreased the likelihood of exhibiting protective behaviours such as wearing a mask or keeping the required distance from others; however, receiving information about the validity of negative results reduced this tendency^[128]. A study conducted in England also showed how intention to comply significantly increased when the behavioural implications of a negative test were communicated more explicitly^[129]. Reiterating that a negative lateral flow result does not mean "not infectious" is essential^[130]. It is also paramount to avoid conflicting messaging that may confuse the public. In some settings in England, people who received a negative result were told it was "great news", while in others that "you were not infectious when the test was done". There were also varying time periods suggested for continued testing: "regularly" in Lewisham, "once a week" in Bradford and "twice a week" in Havering^[131]. In January 2021, a search of the websites of the 114 English local authorities rolling out lateral flow testing showed that the advice given to the public about a negative test result ranged from "Don't let a negative Covid-19 test give you a false sense of security" to "It is good news that you don't have the coronavirus"^[132]. A mixed methods study exploring students' perceptions of testing at the University of Edinburgh reported how students had concerns about conflicting or unclear information on how to self-administer the test, as the guidance they were given differed between their first and second tests and the instructions they received from on-site staff differed from those received elsewhere^[133].

• Involving community leaders and stakeholder organisations in the development and implementation of testing programmes could help build trust, share goals, and bridge cultural and language gaps. The UK Scientific Advisory Group for Emergencies (SAGE) has already emphasised the importance of high levels of engagement with communities and individuals for the successful outcomes of different testing strategies^[134]. A rapid review of the role of community engagement in testing uptake in Southampton found that local community leaders and stakeholders had a fundamental role in determining behaviour change. Participants in the pilot perceived local organisations, such as schools and universities, as answerable to local people, hence more trustworthy than national, more 'faceless' organisations such as NHS Test and Trace. They suggested that receiving information from the University of Southampton and Southampton City Council about the rationale for, the design and the progress of the programme could help testing uptake^[135]. Qualitative research conducted in the United States highlighted the importance of establishing trust within the community, understanding the issues impacting it and finding the most effective channels to promote testing^[136].
• Less invasive sampling techniques could increase uptake. A test that is seen not only as reliable, but also quick and easy to administer has higher chances of being done. Willingness to participate in testing could be higher if less invasive sampling techniques (such as saliva sampling) are provided^[137]. The use of less uncomfortable types of sampling methods may be an important factor for the success of testing programmes in clinical and community settings^[138].
• The context of use requires careful consideration. The disconnect between the prescribed testing regime and the actual context of use should be addressed. A pilot in England showed how the requirements for care home employees or employers to get tested multiple times a week were not compatible with their working schedule. These testing regimes posed a high risk of increased staff dissatisfaction, and consequently staff turnover and burnout^[139]. Similarly, qualitative research conducted with staff in English care homes reported a need for flexibility in a setting where swab-based testing can be organisationally complex and resource-intensive. Staff members often had to return to residents more than once to test at a time which was acceptable, and sometimes only those familiar with some residents managed to perform the test, such as in the case of people affected by dementia. This had a number of implications for staff time and wellbeing^[140].
• Equitable access to tests should be promoted. In order to achieve equitable uptake, individuals who face additional barriers to testing, such as language barriers and/or digital exclusion, should be addressed by tailored campaigns. People should be reminded how to collect or book a test for delivery, and how to perform a test, by means of a range of media channels and formats. In low-literacy populations that might have trouble understanding written or graphic instruction materials, the use of online videos could be an option. On the other hand, as internet-based dissemination strategies may limit access for some segments of the population and require a proactive individual seeking information, it is paramount to explore other routes to communication as well. Recommendations have included promotion of LFD testing through commercial sites (for example, pharmacies and local shops), community-based organisations, and MTUs that could allow dissemination of both venue-based tests and self-test kits^[141].
• Psychological and behavioural consequences of test results may impact uptake. Testing programmes rely on members of the public undertaking a substantial burden of responsibility across the testing stages. Willingness to participate can be affected by concerns about one's capability to self-administer a test, compliance fatigue, fear that positive test results may lead to stigma and worry that self-isolation will not be supported financially in the form of sick pay or other monetary payments. Research conducted in Southampton on testing uptake shows that participants were particularly worried about the personal consequences of a positive or false-positive result: they were concerned that if they had to isolate they would lose income, their employer would be unsympathetic and that a history of infection with the virus might affect their ability to get a mortgage and life-insurance^[142]. This suggests a need for reassurance and that ensuring knowledge of the support available in Scotland is key in dealing with test outcomes. Using a language that acknowledges the challenges people face and emphasising the contributions of individual actions to a societal response may increases testing uptake^[143].
• Emphasising civic duty and the altruistic motivations of engaging in testing may be successful strategies to promote testing. A communication strategy that focuses on protecting others and reducing the impact of the pandemic for society as a whole has already demonstrated its value in other areas, such as vaccination and compliance with regulations^[144]. Furthermore, believing that testing programmes specifically are contributing to tackling the pandemic and knowing that others are willing to participate may increase uptake^[145]. Data from the Southampton pilot show how participants related their decisions to engage in the programme to pride in knowing that they were contributing to the national effort to manage the pandemic and viewed it as a privilege. Southampton University students reported being envied by those from other universities. Those accessing the programme through schools and GP surgeries (smaller and more cohesive organisations where, for example, staff and school pupils saw one another every day) said they frequently spoke about the programme and encouraged one another to take part^[146]. The rapid thematic analysis of local narratives – part of the evaluation of the Liverpool-based pilot – revealed that the desire to protect the community and the belief that mass testing could help the city (and the country) return to normality were key drivers to getting tested^[147].
• Research on the underlying motivations that lead people to get a test when asymptomatic would aid the design of effective health communication and successful implementation of testing strategies. Data from four cross-sectional surveys suggest that respondents got a test mainly to answer the question of whether they could be infected, demonstrating that testing wasn't used as a screening endeavour but rather as a reassurance after risky situations^[148]. Exploring the reasons for use of testing could shed further light on how individuals in Scotland are using the testing pathways open to them (see Asymptomatic Testing Evaluation for more details on testing behaviours and motivations).

Conclusions

This evidence review sets out an analysis of evaluation material provided by a variety of government sources and scholarly research on asymptomatic testing. It has examined 73 studies and reports of varying quality, which have provided an insight into models of delivery, benefits, barriers, costs and impacts of asymptomatic testing regimes. It has also presented some of the lessons learned through the implementation of testing programmes, both in the UK and in the rest of the world.

This analysis should be considered in the context of the rapidly changing area of the investigation, and in light of the lack of complete and/or robust evaluations on asymptomatic testing. The need for accurate empirical data on the effects that asymptomatic testing is having on case finding and reduction, on test numbers and positivity rates has been highlighted by articles and editorials ^{[149] [150] [151]}.

How to access background or source data

The data collected for this social research publication:

☐ are available via an alternative route – see Public Health Scotland's published data on Covid statistics

☐ may be made available on request, subject to consideration of legal and ethical factors. Please contact socialresearch@gov.scot for further information.