The LAVA (Lateral Flow Antigen Validation and Applicability) Study for COVID-19

During the ongoing COVID-19 pandemic, testing of children for COVID-19 has become an area of
substantial need and intense scrutiny. The current gold-standard method for SARS-CoV-2
detection is the real time reverse transcriptase polymerase chain reaction (RT-PCR),
performed on a sample from the respiratory system. This diagnostic test identifies fragments
of viral RNA which are specific to SARS-CoV-2 and amplifies them. It can detect even
relatively low levels of RNA in people who have contracted the virus but have not yet
developed a high viral load or symptoms. It can also detect RNA after the live virus has
cleared, but fragments of the virus' RNA remain. This test has a high analytical sensitivity
and specificity. However, obtaining an adequate specimen is more difficult in children as the
test is uncomfortable and not always well tolerated. This means that the diagnostic
sensitivity is approximately 80%(1). Nevertheless, RT-PCR currently remains the most accurate
method of detection of SARS-CoV-2. However, logistical issues around RT-PCR mean that the
availability and usability of the test is reduced. This is due to a combination of factors
including testing site capacity, processing capacity and the time taken to process and report
the results by centralised laboratories. The net result is that the high sensitivity and
specificity of PCR is offset by the time lag for result reporting. In acute hospitals, these
delays can lead to poor patient flow through clinical areas, overuse of personal-protective
equipment (PPE) or late recognition of nosocomial transmission. Thus, rapid testing is of
particular importance in identifying highly infective individuals, for example people who are
about to undergo a high-risk aerosol generating procedure, enter a crowded emergency room, or
in rapid investigation of localised outbreaks.

Alternative methods and modalities of testing have therefore been explored, in particular the
use of anterior nasal swabs to detect the SARS-CoV-2. Salivary RT-PCR is also under
investigation(2, 3) but is affected by some of the limitations of performing RT-PCR on any
sample, as described above. An alternative method of detection of SARS-CoV-2 is the use of
antigen testing using a lateral flow assay (LFA)(4). Antigen-detection diagnostic test are
designed to directly detect SARS-CoV-2 proteins produced by replicating virus in respiratory
secretions. Most antigen rapid diagnostic tests use a sandwich immunodetection method using a
simple-to-use lateral flow test format similar to a pregnancy test. Antigen detection tests
have several advantages compared to RT-PCR; 1) it can be performed on an anterior nasal
swabs- a specimen which is much more acceptably obtained in children-, 2) the results can be
rapidly processed using a commercial kit at the bedside, with a result being available within
30 minutes of the sample being taken, and 3) it is less expensive to perform an antigen test
compared to RT-PCR. The trade-off for ease-of-use and rapid turnaround time of antigen tests
is a decrease in sensitivity compared to RT-PCR, particularly if the person has a low viral
load (e.g. they are early or late in the disease) because the levels of antigens present in
the upper respiratory tract might fall below the threshold for detection. The analytical
sensitivity appears to be approximately 50% overall, but this increases during periods of
anticipated infectivity when the viral load is >60,000 (RT-PCR cycle time threshold <27) to
approximately 90%. The specificity of antigen tests has consistently been reported to be very
high (>97%). Thus, if the antigen test is positive then it is very likely that it is a true
result, and the person does have COVID-19. Antigen-detection tests are therefore primarily
being used to detect infectivity of an individual, rather than being used as a clinical
diagnostic test for COVID-19.

The potential application of LFAs to defeat Covid-19 are evident, with several clinical
scenarios below demonstrating areas where comparing the performance of RT-PCR to LFA could
improve the care of children and improve the safety of staff in hospital. However, evaluation
of the use of these tests in clinical settings is essential before full implementation in
routine clinical practice is advised.

At present, all children are tested for COVID-19 with RT-PCR when they are admitted to
hospital, regardless of their symptoms. During the winter months, when more children with
respiratory symptoms will seek medical care, current guidance recommends isolation or
cohorting of these patients within hospital until their SARS-CoV-2 test is back(5). However,
the majority of children are likely to have an alternative cause for their symptoms, such as
respiratory syncytial virus, adenovirus or influenza. Early identification of infectious
children with COVID-19 using a point of care test would allow for more effective cohorting or
isolation to occur and potentially reduce the spread of COVID-19 in hospital settings, not
just from children but their parents too who are likely to be infected as a household
contact. Besides, early identification of children with COVID-19 using LFA will aid
compliance with stringent use of PPE for healthcare workers caring for these children.

Children being admitted for elective procedures are currently also routinely tested for
COVID-19 with a nasopharyngeal swab prior to admission. This is performed to reduce the
chance of them harbouring SARS-CoV-2 at the time of having a high-risk aerosol generating
procedure (AGP), where the risk of transmission of SARS-CoV-2 to staff is increased.
Generally, these children are asymptomatic, as the presence of ILI symptoms often precludes
an anaesthetic being performed. In this population, LFAs could be used to identify children
who were incubating SARS-CoV-2 at the time of swabbing but were not shedding the virus who
have subsequently progress to start shedding the virus (Table 1). Performing LFA potentially
enables infectious children to be identified and can be used to guide decisions around
proceeding with a procedure and the use of PPE in children when the procedure goes ahead.

Finally, LFAs could be used in 'high risk' hospital areas, particularly intensive care and
high dependency units where high-risk AGPs are commonly performed and where children are more
likely to have symptoms which are in-keeping with COVID-19. Surveillance of SARS-CoV-2 using
RT-PCR is being performed routinely in some units, but a 24-48 hour delay in the result has
minimal impact on practice around AGPs. LFAs may be utilised as a screening tool for
infectious patients in these areas to support the use of appropriate PPE.

Further use of these tests could also be considered within health care worker (HCW) screening
for infectivity after known exposures to Covid-19 cases, to limit the number of HCW absences
due to potential exposure both outside and inside the hospital. They could potentially also
be used in the community, in settings such as schools and sports clubs, to identify
infectious individuals. Validation and usability in a controlled clinical setting is
recommended prior to use in the wider community. Mathematical analysis of LFAs shows that,
due to their high specificity, the negative predictive value is good in times of both high
and low prevalence, even when the sensitivity of the test is low.