As the coronavirus pandemic has progressed, headlines about testing have become less common, but, with over 420,000 tests conducted on the first day of 2021 alone, it has become one of the largest parts of the UK’s response. Someone who thinks about testing a lot is Dr. Lennard Lee. As an Academic Clinical Lecturer at the University of Oxford, Dr. Lee has his faiir share of medical experience and is now Project Delivery Lead on FALCON-MoonShot, the research programme that evaluated the effectiveness of the LFD tests that have been used extensively by University testing programmes, among others. Below is a written interview that Dr. Lee did for The Oxford Blue, in which he discusses the science behind coronavirus tests old and new, and how tests can best be used by a variety of institutions.

The Oxford Blue: Could you give a brief summary of how the different types work? Why should we use one type over another?

Dr. Lee: There are many different types of tests to detect whether an individual has a coronavirus infection. When the pandemic first was identified, Chinese scientists used an older form of testing called polymerase chain reaction (PCR). PCR detects ribonucleic acid (RNA). However, you can also identify individuals with a coronavirus infection many other ways. For example, by doing imaging tests (like CT scans or X-rays), blood tests that see what is happening to an individual’s immune response (antibody testing) or tests that detect viral protein, viral enzymes or even viral substrates. Identifying viral protein is what lateral flow tests do.

In the second half of 2020, as a result of the World Health Organisation making COVID-19 testing a global research priority, many different forms of testing have become available. Each has advantages and disadvantages. Clearly the right test must be performed in the right individual at the right time.

At a very basic level, tests must be accurate. What is unique about COVID-19 is that it is a disease of the population. Therefore, you also need a test that is acceptable to individuals, easy to implement and use and leads to reduction in COVID-19 cases and transmissions.

I don’t believe there is a perfect test. It has not yet been invented. What we need to do is continue to learn from using new tests, understand their strengths and weaknesses and continue to identify as many cases as possible.

You were personally involved in verifying the sensitivity of the new lateral flow tests. What did the process involve and how confident are you in the results?

The evaluation for lateral flow devices was led by Oxford University within the NIHR FALCON research study. People who had a PCR-proven SARS-CoV-2 infection were asked to return to a COVID-19 testing centre and take a second swab and see if the result was positive on an LFD test. The study tested 200 people who had PCR-proven SARS-CoV-2 infection and showed that 8 out of 10 people with infections were identified. The LFD test did really well at identifying individuals with lots of virus but performed less well when viral levels were low. The amount of virus you have during an infection is a measure of how infectious you are, and therefore LFDs have been better described as a test of infectiousness. Like all Oxford University studies, it was done with great attention to detail and with the highest standards. Each person doing the swab and test did it perfectly. 

When lateral flow tests were used in Oxford at the end of last term, we were required to confirm positive results with a conventional test. Does this mean false positives are a serious risk?

Our Public Health England/Oxford University study showed that if you did 1000 tests in people without an infection, approximately 3 individuals might see a line and therefore interpret the test as a positive, which wasn’t necessarily true. Put into perspective, if you tested every undergraduate in your college, you might expect 1 student to get a positive result which was not correct. Clearly that individual would be frustrated at having to do 10 days of isolation, and if you were affected you would see this as a bad thing. A confirmatory PCR test to double check if the result was true was therefore a smart strategy. It would then ensure that no one is wrongly put into isolation and give you full confidence in that result. Even one person in each college wrongly put into isolation would not be a good thing.

If the tests really do miss as many positive cases as the BMJ’s article suggests, do you think they were appropriate for the sort of “test and release” strategy we saw at the end of the autumn term?

I was not involved in the policy decisions for student testing in Oxford, however, when anything gets done, it is useful to think about what the alternative was and what could have been.

At the end of term, one option was to offer no testing. If you test no-one, then everyone can be released. However, potentially, some students might have put grandparents or parents/siblings with co-morbidities at risk of a terrible disease.

The other option is to consider whether PCR testing could have been done. Completely, aside from the challenges of building new PCR labs for this at last minute and how long it takes for results to come back, there is one other issue with PCR testing. Once you have COVID-19, your PCR test can be positive for several months afterwards. If you did a PCR “test and release” strategy, you hypothetically could end up in a scenario where large number of individuals were told they couldn’t go home for Christmas. Isolating large number of individuals unnecessarily is not a good thing for the student population.

Do you support mass testing programs like the one we saw in Liverpool? How good are programs like this for suppressing coronavirus transmission?

Mass testing is an interesting concept. The idea is that it would be possible to end/curtail COVID-19 if you knew everyone who had COVID-19 everywhere and could isolate them completely. However, it is based on one crucial premise which is that everyone is willing to get a test and is happy to know the result. In reality, not everyone wants to be tested and clearly you have to ensure that there is a carrot and stick for testing. It also depends on having the right test that can be rolled out to large numbers of people in a short space of time. Mass testing needs to be proven, but clearly just because something isn’t proven, doesn’t mean that it is not worth trying.

A report on the program in Liverpool said that it missed 60% of positive cases. Does this mean the program was a failure? How might we improve the effectiveness of future programs?

The Liverpool program was good, but also had areas to improve on. In terms of the good, it identified several thousand people in a short space of time who had no symptoms, but actually had COVID-19. If these people isolated, this would have reduced the COVID-19 burden in Liverpool.

However, the Liverpool program did not test the whole city. Many people chose not to test. In addition, the LFD test was used differently to how it was used in the original FALCON study. This probably highlights that there might be areas for improvement, particularly in ensuring that anyone performing the test gets good training, motivating people to take tests and ensuring the test was used how it was designed to be used.

What place could testing and mass testing have in suppressing the virus? Could it help bring an end to the pandemic?

If everyone who had COVID-19 could be identified and isolated, then testing and mass testing could suppress the virus. The more people identified, the better.

However, I do not believe you can just have one strategy to suppress or bring an end to the pandemic. Testing won’t work in isolation if people don’t isolate if they have COVID-19, don’t use facemasks and don’t socially distance. As a country, we probably have to use every tool available to us to control the virus.