select search filters
briefings
roundups & rapid reactions
before the headlines
Fiona fox's blog
Earlier this week the government posted a policy paper on the community testing programme.
Dr Alexander Edwards, Associate Professor in Biomedical Technology, Reading School of Pharmacy, University of Reading, said:
“The accuracy of rapid antigen lateral flow tests is known to suffer somewhat from relatively low sensitivity compared to some RT-PCR methods. But there is a clear trade-off in speed, cost and convenience. It’s also important to note that not all RT-PCR methods are equal, and a major independent comparison run by the US FDA found some RT-PCR methods are far more sensitive than others. So the limitations in accuracy for all these different tests are well documented and well understood within the clinical diagnostics community. Part of these limitations come from the test; but a lot comes from the biology of a viral infection. There will always be the possibility of being infected yet not having enough virus present in the body to be detected. For example, during the incubation period from exposure to first starting symptoms, which might be on average 5 days.
“What is vital is to know what people have been told after they receive their test result. And to know how the mass testing program was promoted or publicised. Participants must be fully informed that there is still a possibility they are infected even if they did not test positive. Ideally, they should have this explained to them so that they can ask questions and be assured that a negative result does not make them “safe”.
“The key is in how different tests are used. For example, a rapid, portable, lower cost test could be deployed in a widespread way, aimed at picking up additional cases before symptomatic. The knowledge of less than perfect sensitivity must be built into the testing service. And it must be shared transparently with participants.
“The positives of lower accuracy mass testing: if you can pick up extra cases, and trace contacts, and isolate, you may be able to reduce transmission. The negative: false negatives might provide false reassurance, if this leads to increased spread the gains from identifying more real cases might be lost.
“Explaining the meaning of test results, and designing careful testing and screening programs, is not as easy as it appears at first sight. But we do now have plenty of data to ensure evidence based testing services are developed.”
Prof Rowland Kao, Royal (Dick) School of Veterinary Studies and Roslin Institute, University of Edinburgh, said:
“It has always been known that the lateral flow tests were likely to be less sensitive at picking up infected individuals. The Liverpool data provides a valuable direct comparison of this in the field. It is absolutely correct to say that a negative on a lateral flow test should not be taken as a strong indicator of not being infected. Where the mass testing results are most valuable, is in providing a broad, relatively unbiased picture of how COVID-19 infection is distributed across a large population. This information could then be used to better assess the assignment of tier status for example, or where it might be valuable to place increased effort into contact tracing.”
Dr Daniel Howdon, Senior Research Fellow, Health Economics, said:
“Previous validation of lateral flow testing carried out in England had suggested that, when tested by experienced staff, around 25% of individuals who would test positive using a PCR test would test negative using these new tests. For symptomatic individuals with a higher viral load, it had previously been suggested that only 5% who would test positive would test negative using a lateral flow test.
“This government statement suggests that, in practice, a substantially higher 50% of PCR positives are currently testing negative using a lateral flow test, with around 30% of those with a high viral load testing negative. Although pointing to a previous estimate of false positives, the statement says nothing new about the likely rate of these with the test as used in practice, an issue that becomes increasingly important with declining prevalence of the virus.
“All of this substantially calls further into question the utility of – and further raises the spectre of potential harm caused by – the government’s ongoing mass testing of asymptomatic individuals. This is particularly the case for the huge numbers of students in higher education who are currently being tested prior to returning home, often with advice from their institution that this test offers “peace of mind” that they are not infectious. Furthermore, especially without confirmatory testing, even a small proportion of false positives is likely to lead to unnecessary and potentially-harmful legally-mandated isolation of students.
“While we regrettably do not have a full set of results on the performance of this test beyond this single sentence in a statement published on the government website, it is clear that this test is potentially giving false reassurance to a large number of potentially-infectious individuals and delivering an unknown proportion of false positive results. Full results from the Liverpool pilot should be released so that researchers can assess in further detail the figures alluded to in this one sentence of the government statement.
“We should not assume that testing cannot cause harm, and given the paucity of information available about these tests and the near-zero information available about behavioural responses to their results, we certainly should not assume that in this case.
“While it is questionable as to whether this programme of mass testing should continue at all, at a very minimum it is a matter of the utmost importance that those undertaking the test are informed as to the likely performance of this test, and that government and higher education institutions make clear to students the meaning, insofar as this is possible, of their test result. Failure to do so means that those undertaking the test both cannot be reasonably said to be giving informed consent for testing, nor can they reasonably interpret their test result.”
Prof Sheila Bird, Formerly Programme Leader, MRC Biostatistics Unit, University of Cambridge, said:
“Other areas of society, such as universities and care homes, should follow Liverpool’s lead in analysing the performance of any rapid test that they deploy.
“I welcome the release by DHSC & Liverpool with the information about performance and limitations for the INNOVA lateral flow test when used for asymptomatic testing. However, it would be more useful if they published the raw data with the actual numbers, rather than just the percentages, of citizens in Liverpool who had the INNOVA lateral flow test and RT-PCR testing. More information is needed, such as – was the swab for the PCR test always used with the INNOVA test? If so, then swab-order and test-type were confused, which is not ideal . . . but perhaps the best that could be achieved in the race to begin mass screening.
“Liverpool also offered its citizens RT-PCR confirmation of INNOVA positive tests but the confirmation-swab was not necessarily taken on the same day. Initially, Liverpool was reliant on the delivery of home-test-kit which were self-administered. The risk that the PCR test wasn’t delivered, wasn’t used at all or was used by someone other than the intended individuals were all uncontrolled risks.
“However, the principle of using a PCR test to check a positive result from the lateral flow test is excellent and another important piece of the testing jigsaw that Liverpool put in place. It would be useful if the full results from this check could be published.
“Finally, a cautionary note: the performance of the INNOVA’s rapid test on asymptomatic Liverpool citizens is critical prior information that universities and care homes should now have in detail.
“Deployment of the same or a different rapid test for use with university students or for older relatives who wish to visit a family-member who is a care-home resident requires the same evaluations as in Liverpool. Only by doing so can they be confident about test-performance in different contexts.”
Dr Joshua Moon, Research Fellow in the Science Policy Research Unit (SPRU) at the University of Sussex Business School, said:
“There’s a few factors that need to be thought about here regarding the use UK’s programme of mass testing:
1) The lateral flow tests require much higher amounts of viral RNA in the sample to be detectable, so this was predictable that there would be a higher false-negative rate
2) In higher viral load cases, the false negative rate goes down, this is also predictable
3) This does, therefore, raise the question of how many cases the mass screening programme is missing
4) However, there is also the issue of whether those cases that are missed have high enough viral loads to be transmitting the virus anyway
5) Of course, it could be that they are early in the infection process and thus have low viral loads that will get higher and become transmissible
6) What this means is that among these options (asymptomatic-infectious, asymptomatic-safe, presymptomtic-infectious, presymptomatic-safe, not infected) is uncertain without additional testing.
“It is therefore questionable as to how safe this may be, particularly on large scales like the whole of the UK.
“The lack of data for the testing programme is extremely disappointing. The release from the government has a few statistics but no release of the overall data and how this was produced.
“This is doubly important when assessing the shortcomings of a mass screening programme like this because this data is both valuable insight into the pandemic but also data for evaluating the utility of the tests themselves in a real world setting.”
Declared interests
Prof Sheila Bird: “I am a member of the Royal Statistical Society’s COVID-19 Taskforce.”
None others received