Molecular and antibody point-of-care tests to support the screening, diagnosis and monitoring of COVID-19
April 7, 2020
Kile Green1, Sara Graziadio2, Philip Turner3,4, Thomas Fanshawe3,4, Joy Allen1
On behalf of the Oxford COVID-19 Evidence Service Team
Centre for Evidence-Based Medicine, Nuffield Department of Primary Care Health Sciences
University of Oxford
1NIHR Newcastle In Vitro Diagnostics Co-operative
Newcastle University, Newcastle upon Tyne, NE2 4HH
2NIHR Newcastle In Vitro Diagnostics Co-operative
Newcastle upon Tyne NHS Hospitals Foundation Trust, Newcastle upon Tyne, NE1 4LP
3NIHR Community Healthcare MedTech and In Vitro Diagnostics Co-operative, Oxford Health NHS Foundation Trust
4Nuffield Department of Primary Care Health Sciences, University of Oxford, OX2 6GG
Correspondence to email@example.com
Molecular & antibody POCT to support the screening, diagnosis & monitoring of COVID-19 PDF
POCT COVID-19 Tables PDF
Moving diagnostic testing for COVID-19 from laboratory settings to the point of care is potentially transformative in the rate and quantity of testing that could be performed. Eleven diagnostic tests that are potentially suitable for testing for COVID-19 at the point-of-care are described: six molecular tests, and five antibody-based tests. Some devices show high diagnostic accuracy during controlled testing, but performance data from clinical settings, and a clear understanding of the optimal population and role for these tests in the care pathway, are currently lacking.
The need for increasing levels of testing for COVID-19 has been identified by both the World Health Organisation and by the UK government.
Currently, most COVID-19 testing is performed in the laboratory environment. Guidance for virus testing in NHS laboratories is available here, and the WHO also provides technical guidance for laboratory testing. A comparison of oropharyngeal and nasopharyngeal swabs for laboratory diagnosis has been previously reported.
Accurate and scalable point-of-care (POC) tests for the diagnosis of COVID-19 would increase the scope for diagnosis to be made in the community and outside the laboratory setting (Wang et al (1), Nguyen et al (2)). They would have the potential to reduce the time to obtaining an actionable result, could support early identification of those with COVID-19 and could also support appropriate use of isolation resources, infection control measures, and recruitment into clinical trials of treatments.
In this report, we summarise the characteristics of current molecular and antibody diagnostic tests available to support the diagnosis and management of patients with suspected COVID-19. We consider assays that could run on analysers near to the patient, rather than those that would typically be placed within a laboratory. Many of these POC tests are molecular-based PCR-type tests, but others are serological assays, which detect the presence of antibodies in a blood sample.
The current reference test for diagnosis of active infection by SARS-CoV-2 is a real time reverse transcriptase polymerase chain reaction (rRT-PCR) assay (Corman et al (3)). The rRT-PCR assay utilises viral RNA extracted from patient samples (e.g. material collected by NP/OP swab), synthesises complementary DNA (cDNA) through the action of the reverse transcriptase enzyme, and amplifies target sequences of the viral genome from the cDNA template. rRT-PCR can be interpreted in a semi-quantitative manner, with the speed of target amplification dependent on the concentration and quality of viral RNA in the initial sample, and thus amplification rate can be used as a proxy for sample viral load.
Failure to amplify can be interpreted as a negative result, but could also be attributable to poor quality of the clinical sample or to early disease status. These assays can be run on standard rRT-PCR thermocyclers or large automated or semi-automated diagnostic platforms. Testing in patients suspected of having COVID-19 involves sending a respiratory sample (e.g. oro/nasopharyngeal swab, sputum or bronchoalveolar lavage in seriously unwell patients) to a reference laboratory for rRT-PCR testing. The time between sample collection and generation of results can range from 24 to 72 hours, but could be much faster with a streamlined approach from sample to answer for urgent clinical scenarios.
Molecular point-of-care tests utilise the same basic methodology as the laboratory assay, but essentially automate a varying number of the steps required. As they could be operated in near-patient settings rather than on the laboratory bench, they might be expected to provide a shorter time to result.
Serological and antigen tests
Serological tests, using enzyme-linked immunosorbent assays, detect the presence of antibodies to coronavirus in a whole blood, plasma or serum sample (Xiao et al (4)). These tests detect immunoglobulins M and G (IgM and IgG). IgM is the largest immunoglobulin, and is the first to appear after initial exposure to an antigen. IgG is the most common antibody found in the body, which will appear later but will be generated in abundance. These tests can determine whether a patient has previously been infected with coronavirus, as they will stay positive after active infection has gone.
Currently, serological testing is not routinely offered as part of the screening or diagnosis of COVID-19, as no validated assays are available. These tests will not be positive until the body has started to make antibodies to fight the virus, typically 5-10 days post-infection. The widespread use of such a test could reveal what percentage of the population has had the virus, but these tests are less likely to detect cases in the early stages of disease. In cases where the molecular test is negative but there is a strong clinical suspicion of COVID-19 disease, serological testing could support a diagnosis once validated assays become available.
Antigen tests (Khan et al (5)) may also offer additional information before or at the time of taking a sample for molecular screening, but there are no commercially available antigen tests for COVID-19 available at the time of writing.
For a more detailed overview of relevant laboratory methods, see Loeffelholz & Tang (6).
We accessed the websites listed in the Search Strategy (below) on 26/03/2020 and extracted the list of POC tests available.
We recorded the following information as recorded on manufacturers’ websites and assay package inserts. We also attempted to obtain information about diagnostic performance by contacting the manufacturers directly, but as little extra was obtained we report here publicly-available information only.
- Device type
- Target (e.g. SARS-CoV-2 or immunoglobins)
- Sample type required for testing
- Whether CE marked and/or having emergency FDA approval
- Time required for sample preparation and to obtain diagnostic result
- Throughput (e.g. number of cartridges that could be processed at any one time)
- Storage requirements
- Diagnostic performance (e.g. sensitivity and specificity, and whether using laboratory or clinical samples)
We found six commercially available molecular POC tests, five antibody-based tests and no antigen tests at the time of conducting the search. A comparative summary is shown in Tables 1 and 2.
Molecular POC diagnostics
Most of the six molecular POC tests have either gained CE marking or emergency FDA approval. As at time of writing we could not find clinical evaluations of these assays in the literature, the information summarised below is extracted from the manufacturer package inserts or from their websites.
Almost all are portable, benchtop-sized analysers, apart from the MicrosensDx RapiPrep©COVID-19 test and the MesaBioTech Accula Test, which are smaller, handheld devices.
Typical validated sample types include nasal, throat, oral or nasopharyngeal swabs. The MicrosensDx also supports sputum samples.
All tests require sample preparation, which involves placing the swab sample into a viral transport media and pipetting a proportion of the sample into a single-use cartridge. This preparation step is typically quoted to take approximately two minutes but may take 5-10 minutes for some devices. The Abbot ID Now kit indicates a 1-2 minute preparation time, as the swab is mixed with the viral transport media within the cartridge in the analyser.
Most POC devices are single-access and operate with single-use cartridges. The Cepheid Xpert SARS-CoV-2 can run 2-4 samples per run in a random access manner, and the GenMark EPlex can run 3 samples per run in a random access manner.
Storage of most cartridges requires refrigeration plus some time to equilibriate to room temperature, apart from the Cepheid Xpert SARS-CoV-2, Mesa BioTech Accula SARS-CoV-2 and Abbott ID NOW COVID-19 tests, which can be stored at room temperature prior to use.
Time to result varies from 13 minutes (Abbott ID NOW) to 45 minutes (Cephied Xpert Xpress).
For these six devices there was no evidence of clinical diagnostic accuracy from prospective clinical evaluations. Preliminary evidence extracted from the package inserts of the cartridges showed validation data restricted to small numbers of spiked samples in a laboratory setting (typically 20-50 positive samples). Most compared positive agreement on a range of limits of detection and, where available, reported perfect diagnostic performance in this controlled setting. Validation information for each device is provided within Table 1, and a more concise summary appears in Table 3 for comparative purposes.
Antibody POC diagnostics
Of the five antibody-based tests, two are lateral flow immunoassays (BioMedomics rapid test and Surescreen rapid test cassette), one is a time-resolved fluorescence immunoassay (Goldsite diagnostics kit) and two are colloidal gold immunoassays (Assay Genie rapid POC kit and VivaDiag COVID-19 IgG-IgM test).
All assays detect the presence of IgG and IgM from whole blood, serum or plasma. They involve pipetting a few drops of blood from a fingerprick or vein onto the immunoassay, followed by a couple of drops of buffer solution, with the result displayed (as lines similar to a pregnancy test) within 10-15 minutes. All use single-use disposable cartridges, and most can be stored at room temperature.
The reference standard used for comparison in these studies was RT-PCR testing. Some diagnostic accuracy data was collected from clinical, rather than laboratory testing, the largest such study being the evaluation of the BioMedomics IgM-IgG rapid test (Li et al (2020) (7)), which estimates 89% sensitivity and 91% specificity among 525 patient samples (Table 3). Being based on published clinical data, this evaluation constitutes stronger evidence than the other evaluations reported in Table 3. We also found a registered clinical trial protocol for VivaDiag and anticipate that further clinical accuracy data will become available as the COVID-19 pandemic progresses.
An increasing number of diagnostic devices that are potentially suitable for the diagnosis of COVID-19 at point-of-care are in development. Different devices may be more suitable for diagnosing new cases on infection, while others, especially those that test for the presence of antibodies, are better suited to determining whether an individual has previously been infected. This latter scenario is likely to be of paramount importance in identifying healthcare workers who may have recovered from initial infection, to ascertain suitability to return to frontline health services. It may also help to inform public health strategies at the end of periods of lockdown or as social distancing restrictions are relaxed.
Importantly, we have found relatively little current information reporting the diagnostic performance of these POC devices using clinical samples taken from community settings. Relevant data may still be under collection in ongoing studies, or may not be published publically. Typically, diagnostic performance might be expected to be lower in clinical settings than when using spiked samples in a controlled laboratory environment.
It should also be noted that the laboratory rRt-PCR reference standard is subject to some misclassification error, and in particular false negative results may arise. This has relevance to the conduct of clinical evaluations as misclassification in the reference standard may affect the apparent diagnostic performance of the POC tests being evaluated. Other considerations that may influence performance include pre-analytical factors such as the quality of the respiratory sample collected, the time point during infection when the sample is collected, and the handling and storage of the sample prior to analysis.
In the event of large-scale rollout in the community, any decline in diagnostic performance is likely to have serious consequences, either in providing false reassurance to infected cases, or by overdiagnosing disease-negative individuals. There is also little evidence as to the psychological and behavioural consequences of knowing immunity status, whether or not correctly diagnosed. Sufficient clinical testing is therefore vital in determining suitability.
Disclaimer: The article has not been peer-reviewed; it should not replace individual clinical judgement and the sources cited should be checked. While this article contains information about the performance of diagnostic devices available online on the search date, this information is subject to change and may be superseded as new data become available, and should not be interpreted as an endorsement of any particular device. The views expressed in this commentary represent the views of the authors and not necessarily those of the host institution, the NHS, the NIHR, or the Department of Health and Social Care. The views are not a substitute for professional medical advice.
We accessed the following websites on 26/03/2020 and extracted the list of POC tests available:
1. Wang C et al. A novel coronavirus outbreak of global health concern. Lancet 2020; 395(10223): 470-3.
2. Nguyen T et al. 2019 Novel coronavirus disease (COVID-19): Paving the road for rapid detection and point-of-care diagnostics. Micromachines 2020; 11(3): 306
3. Corman VM et al. Detection of 2019 novel coronavirus (2019-nCoV) by real-time RT-PCR. Euro Surveill 2020; 25(3): 2000045.
4. Xiao SY et al. Evolving status of the 2019 novel coronavirus infection: Proposal of conventional serologic assays for disease diagnosis and infection monitoring. Journal of Medical Virology 2020; 92(5): 464-7.
5. Khan S et al. Analysis of serologic cross-reactivity between common human coronaviruses and SARS-CoV-2 using coronavirus antigen microarray. bioRxiv 2020. https://doi.org/10.1101/2020.03.24.006544.
6. Loeffelholz MJ & Tang YW. Laboratory diagnosis of emerging human coronavirus infections – the state of the art. Emerging Microbes & Infections 2020; 9(1): 747-5
7. Li Z et al. Development and clinical application of a rapid IgM‐IgG combined antibody test for SARS‐CoV‐2 infection diagnosis. Journal of Medical Virology 2020. https://doi.org/10.1002/jmv.25727.