ABOUT THE STANDARD

This database lists point of care COVID-19 diagnostic tests – i.e. portable tests of the type used on productions.

Tests marked are suitable for CCC screening under the Production Restart Scheme.

The standard used is the MHRA ‘desirable’ performance characteristics for point of care COVID-19 diagnostics. If a test doesn’t meet this (admittedly high) standard, that doesn’t mean there is anything wrong with it – it just means it doesn’t meet the new standard for CCC testing. This list is not exhaustive so please contact us if a test is missing and we’ll add it.

NB The MHRA standard for laboratory PCR is different and will have been audited as part of the lab’s ISO accreditation with UKAS so this doesn’t apply to tests you send away to a laboratory for processing.

On November 1st 2021 The Medical Devices (Coronavirus Test Device Approvals) (Amendment) Regulations 2021 come fully into force. This SI creates a requirement for the mandatory approval of diagnostic tests for COVID-19. It is intended to ensure that diagnostic tests for COVID-19 sold in the UK meet minimum standards in their quality and accuracy.

This legislation was rather rushed and the time for the MHRA to review data submitted by manufacturers was short. So there are two lists: there is the ‘full’ list the Government committed to publishing by October 31st and a supplemental list of tests that can also be used. The second list seems to be based on tests evaluated by the TVG on behalf of the NHS. The intention is those tests will migrate to the ‘full’ list in due course once they’ve been checked.

Productions are advised to ensure their testing suppliers warrant that any test being used after November 1st is on one of the approved lists. 

More Information

This list is not exhaustive. Evaluations are based on comparing suppliers’ data and any independent external validations with the ‘desirable’ performance characteristics set out in the MHRA Target Product Profile1 (TPP). This is the performance standard agreed with DCMS for testing Close Contact Cohorts. Compliant tests are highlighted in yellow.

NB If a test fails any of the basics such as Sensitivity, Specificity or Limit of Detection, we may not include the less critical characteristics (such as time to result) unless those data are readily available. We don’t waste time on the minor aspects if a test has already failed one or more of the critical ones.

Please also refer to the Disclaimer below.

The first thing to remember is that no test is actually detecting the virus itself. They all detect various bits of the virus – which isn’t the same thing as detecting live, infectious virus.

PCR

PCR amplifies tiny fragments of viral genetic material over and over again until there is enough there for the machine to detect. There are many types of PCR and to be super-accurate SARS-CoV-2 is an RNA virus and PCR amplifies DNA, so you need to convert the RNA to DNA using an enzyme called reverse transcriptase, hence ‘RT-PCR’.

The benefit is that PCR can detect minute amounts – but viral RNA can still be detected long after an infection has cleared so people will still test positive when they are not a danger. On average, people are infectious for 8 days but will be PCR-positive for 22-33 days – sometimes much longer.

You’ll often hear about ‘Ct’ or ‘cycle threshold’ values. Each time PCR doubles the genetic material present, that’s a ‘cycle’. So if there is lots of viral genetic material it doesn’t take many cycles for there to be enough for the machine to detect. So people with a high viral load will have a low Ct value. For a sample with a low viral load the PCR needs to work harder (more cycles) so the Ct value will be higher. It’s generally reckoned that someone with a Ct≥34 will not be infectious.

Care: there are many ‘mini’ PCRs now on the market, many of which lack the exquisite accuracy of their lab-based big brothers.

LAMP

LAMP also amplifies viral RNA like PCR but using a slightly different technique. You don’t get a Ct with LAMP (because it doesn’t cycle) but the longer the reaction takes, the lower the viral load (again because the machine needs to work harder because there’s less RNA in the sample).

Immunofluorescence

This detects viral antigens (proteins) rather than RNA. These tests use fluorescent monoclonal antibodies that attach to SARS-CoV-2 antigens; these are then captured and the reader device shines a special kind of UV light at the sample. The fluorescent antibodies then – well, fluoresce – and are read by the machine. These tests use a similar process to visually-read lateral flow tests but can detect far less virus in a sample because the machine is so accurate and needs far less signal than the human eye needs to detect a ‘test’ line colour change on an LFT.

Some immunofluorescence tests have accuracy similar to PCR but take care – some do not. Also the magic happens in the test cassette, not so much the reader so test cassettes from different manufacturers that will run on the same machine can give markedly different results. It’s a bit like a Nespresso machine; if you use the expensive pods from the manufacturer you’ll likely get a better coffee than from the knockoff ones that sell for half the price – but both run in the same machine.

Suitability of tests to enable close-contact work is now based on meeting a technical standard set by the MHRA and not test type – formerly only PCR was allowed. Clearly this opens up new options for productions; the only requirement is that tests need to meet the ‘desirable’ performance criteria set out by the MHRA.

The TPP was established by the MHRA to guide manufacturers regarding MHRA expectations for diagnostic test performance being offered to agencies such as the NHS. It lists an ‘acceptable’ and a ‘desirable’ performance for multiple aspects of a test– e.g. handheld and battery-operated devices are ‘desirable’, desktop and mains powered is ‘acceptable’. In all cases ‘desirable’ is the higher specification.

And while the TPP covers a lot of characteristics the CCC guidance only references a test’s performance characteristics which speak to its ability to pick up an infection.

The main features for COVID-19 diagnostics defined in the MHRA Target Product Profile (TPP)1 are Sensitivity, Specificity and Limit of Detection.

One of the challenges of PCR is the ‘long tail’ of detectability of viral RNA. On average people are contagious for 4-8 days4 5 but ‘PCR positive’ for a median of 22–33 days and often longer6. This, combined with current test, trace and isolation rules can, as we have seen, cause significant interruptions to productions.

Some of these newer tests can give far more rapid results at a much lower cost – and it is well established that time to result is far more critical in preventing an outbreak than exquisite sensitivity7.

The law changed on September 1st 2021 requiring manufacturers (or their ‘UK Responsible Person’) to submit a dossier proving a test is of sufficient quality. There is a grace period for existing tests provided the manufacturer submitted the dossier and paid the fees before September 1st. This period ends on 31st October.

By then the Government will publish a list of all qualifying tests and after that date anyone supplying a test not on that list will be committing an offence.

NB This is not the same as the existing DHSC technical validation for tests intended for supply to the NHS.

Clinical or diagnostic sensitivity is the test’s ability to correctly identify a positive. The lower the number the greater chance of false negatives.

Clinical or diagnostic specificity is the test’s ability to correctly identify a negative. The lower the number the greater chance of false positives. False positives or negatives are bad – for different reasons.

A 95% confidence interval is a measure of the precision of the sensitivity and specificity estimates. The narrower the CI the better the more precise the estimate.

This is also known as Limit of Detection (LoD) and is a key performance indicator. The LoD represents the lowest concentration of analyte (whatever you’re looking for) needed in a sample that can be reliably detected 95% of the time.

A test used for screening CCCs is a very different animal to a test used to support a clinical diagnosis of a symptomatic patient because the patient will have a much higher viral load than an asymptomatic person early in an infection.

PCR’s ability to detect low levels of analyte is one of the principal reasons PCR has always been the go-to assay for screening CCCs. It was also the first available test and for a while was the only one.

Lab PCR will generally have a LoD of under 100 copies per ml of sample. This is also the ‘desirable’ standard set in the MHRA TPP.

A viral load of 100 RNA copies per ml is equivalent to an average Ct on RT-PCR of 34.52 3 and it’s generally accepted that no infectious virus can be recovered from a test subject with a Ct of 34 or higher9.

A challenge with the TPP is it was written when NAATs were the only real option. LoDs are sometimes reported in units other than copies of viral genomic per ml, especially in assays not using nucleic acid amplification. LoDs can be given in units such as PFU, TCID50, copies/microliter, copies per reaction volume, or molarity of assay target, making easy comparisons difficult. But not impossible.

In Prof Tim Peto / Oxford / Porton Down’s evaluation of a large-scale trial2 3 the authors give a correlation between actual virus present (counted in something called Plaque Forming Units (PFU)/ml, RNA copies per ml (as measured by PCR) and approximate Ct values these equate to.

This is a slightly tricky correlation as you are comparing different ways of measuring how much virus might be in a sample plus Ct isn’t set in stone; it depends on the particular reagents and machine the PCR is using; the same sample run on two different PCRs will give two different Ct values. These likely won’t be hugely divergent but will be different.

So, if the TPP says the standard is 100 copies per ml, how do we apply that to tests that measure LoD in other units?

Prof Peto’s data suggests ≤1 TCID50 is roughly equivalent to ≤100 copies per ml. There are other studies we could have used10 11 12 but they all end up in the same ballpark.

This is a different kind of specificity. This asks “are other organisms or substances you find in healthcare samples going to give a false positive”. There is a very long list but diagnostics do not need to be tested against all of them to pass. This is a suggested list from the MHRA that manufacturers are invited to ‘consider’.

The most important on there are the other coronaviruses – you need a test that’s specific for COVID-19 that’s not triggered by seasonal coronaviruses – and organisms with a similar clinical presentation such as influenza.

  1. Medicines and Healthcare products Regulatory Agency. Target Product Profile: Point of Care SARS-CoV-2 detection tests. GOV.UK. [Online] 1.0, 15 June 2020. [Cited: 22 May 2021.] https://www.gov.uk/government/publications/how-tests-and-testing-kits-for-coronavirus-covid-19-work/target-product-profile-point-of-care-sars-cov-2-detection-tests.
  2. PHE Porton Down & University of Oxford. Preliminary report from the Joint PHE Porton Down & University of Oxford SARS-CoV-2 test development and validation cell: Rapid evaluation of Lateral Flow Viral Antigen detection devices (LFDs) for mass community testing. 2020.
  3. PHE Porton and Oxford University. Rapid evaluation of Lateral Flow Viral Antigen detection devices (LFDs) for mass community testing. 8 : November, 2020.
  4. SARS-CoV-2, SARS-CoV, and MERS-CoV viral load dynamics, duration of viral shedding, and infectiousness: a systematic review and meta-analysis. Cevik, M, Tate, M and Lloyd, O. 2021, Lancet Microbe, Vol. 2, pp. e13-e22.
  5. Viral cultures for COVID-19 infectious potential assessment – a systematic review. Jefferson, T, et al. s.l. : IDSA, 3 December 2020, Clinical Infectious Diseases.
  6. Prolonged persistence of SARS-CoV-2 RNA in body fluids. Sun, J, Xiao, J and Sun, R. s.l. : NIH, 2020, Emerg Infect Dis, Vol. 26, pp. 1834-1838.
  7. Test sensitivity is secondary to frequency and turnaround time for COVID-19 surveillance. Larramore, Daniel B, et al. 27 June 2020, medRxiv.
  8. SARS-CoV-2 detection, viral load and infectivity over the course of an infection. Walsh, Kieran A, et al. s.l. : The British Infection Association, 26 June 2020, Journal of Infection, pp. 357-371.
  9. —.Walsh, Kieran A, et al. s.l. : The British Infection Association, 26 June 2020, Journal of Infection, pp. 357-371.
  10. The total number and mass of SARS-CoV-2 virions in an infected person. Sender, Ron, et al. 17 November 2020, medxiv.
  11. Rapid Inactivation of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) by Tungsten Trioxide-Based (WO3) Photocatalysis. Ghezzi, Silvia, et al. 2 August 2020, bioRxiv.

While we make every effort to ensure all the information in this document is accurate and up to date at the time of writing, much of the subject matter is changing rapidly. This document has been developed with the intention of providing information only. First Option Safety Group Ltd accepts no responsibility for use of the information provided. The information is as comprehensive and accurate as possible, but it can only be of a general nature and should not be used as a substitute for a consultation with a medical professional.

First Option Safety Group Ltd, its subsidiaries, directors, employees and agents cannot accept responsibility for the references referred to, or the information found there. All the references are provided for information and convenience only. A reference does not imply an endorsement of its information. Reference to any specific product or entity does not constitute an endorsement or recommendation by First Option Safety Group Ltd , its subsidiaries, directors, employees and agents; likewise, not referring to a particular product or entity does not imply lack of endorsement.

First Option Group Ltd, its subsidiaries, directors, employees and agents do not make any representation or warranty of any kind, express or implied, as to the accuracy, completeness, suitability or validity of this information herein and do not accept any liability for any loss or damage howsoever caused arising from any errors, omissions or reliance on any information or views contained herein to the maximum extent permitted by applicable law.

Nothing in this document constitutes medical advice. We are not a health service provider or laboratory, nor do we provide any medical tests, or equipment, therefore we are not responsible for any result or information on the Services carried out by any third-party Testing Providers.