Are self-testing rapid antigen (lateral flow) tests for Covid-19 accurate enough

The dramatic rise in the number of cases of the Omicron Covid-19 variant has led to an increased use of self-testing kits for the virus.  The results from these tests are used to guide personal behaviour.  Many people are confused about how their own results should be interpreted and acted on, with varying guidelines and differences in expert opinions on what to do following a positive test.  I hope this blogpost answers these questions and helps families and groups make informed decisions.  Please share and do send me your feedback through the Comments section below together with any questions 

What is the purpose of these tests?

  • This might seem a somewhat obvious question, as both lateral flow (LFT) and PCR tests are designed to diagnose Covid-19 infection
  • But, as is well known, the purpose of diagnosing Covid-19 is not primarily to guide  what treatment people should have: the more practical role is deciding if someone is infectious and hence could pass the virus on to other people
  • Thus, the accuracy of these tests is more important in determining whether someone is infectious*  than whether they are just infected with Covid-19
  • This is important as one can be infectious without being  ill and be ill without being infectious  

*A comment on words: the terms contagious and infectious I have seen interchangeably in articles on Covid-19, some argue that the Latin origin of the former strictly speaking means transfer of infection by touch which is not predominantly the case with Covid-19 

What is the time course of Covid-19 infection? 

  • I am sure most readers will be familiar with these details, but a refresher can often help!
  • In the above chart, assuming that the  contact with an infectious person is at ‘Day 0’, then there is a short period of time – called the ‘latent period’ – during which  the virus gets established  
    • This is probably around 2-3 days
    • This may be a little shorter with Omicron, but is at least a day and probably longer
  • This is then followed by a period called the ‘incubation period’ – where the virus start multiplying and invading cells in the body
    • At the beginning of this period everyone is asymptomatic
    • Typically around 3 to 7 days, symptomatic infection can start
  • The light blue bar shows that you can be asymptomatic for up to 10 days before first symptoms occur, but if symptoms are going to happen then they will normally have developed within 7 days 
  • Conversely, as shown in the red bar, symptoms can continue beyond the 10th day – ‘long Covid’ being used if symptoms persist for more than 28 days   
  • Now look at the diagram below in terms of when people are infectious
  • People typically start being infectious during the period of incubation, whether or not they are going to get  symptoms
  • But 90% of people stop being infectious after a week
  • As the red bar shows it is exceptionally rare to be infectious after 10 days – including people who still have symptoms

Is there a totally foolproof  way of determining if someone is still infectious?

  • LFT and PCR both assess if someone is still infectious but they are both only indirect tests
  • The only true way of determining if someone is still infectious is to culture (ie attempt to grow) the live virus from a swab
  • If the virus grows in the laboratory, we assume that it could grow ‘in life’ and spread to another person
  • Such tests are not done routinely because of their costs, time to complete and complexity
  • It is also not a definite that even if you can culture a virus from a nose swab, that there is enough virus present to pass on the infection
  • Summary: despite PCR and LFT both being indirect ways of measuring infectivity, they are  entirely appropriate, especially as public health tools, for large scale testing

What is the difference between a PCR and LFT?

  • (For those who are interested, I have given some more technical detail about these tests in an Appendix at the end of this post)
  • Both tests use the same sample from a nose swab 
  • The PCR test aims to discover the presence of evidence of Covid-19 RNA in the sample
  • The LFT test aims to discover the  presence of proteins made by Covid-19 virus in the sample
  • Why this difference is important is shown in the chart below:
  • The PCR test becomes positive very soon after the latent period but remains positive sometimes for a week or even longer after people stop being infectious
  • The LFT needs the infection to be established and the virus to be producing enough proteins before that test becomes positive
  • With this background, I will attempt to answer the key questions about these tests in terms of their performance to accurately detect if someone is infectious
  1. Do these tests work differently depending on the variant?
    • The short answer is no
    • Most of the research about these tests does not come from Omicron, but the conclusions should  apply to this variant for both PCR and  LFT 
  2. Is it possible to be infectious but be negative on both PCR and LFT?
    • The answer is yes and possibly as high as 5% of infected people could be negative on both PCR/LFT
    • There are a number of ‘technical reasons’ for a false negative given all the steps where there could be a problem (see below)
    • As one example, there are many LFT kits on the market which when they have been tested on the same swab, do not always give the same result
    • But in general, it is very unlikely (less than 5%) that someone is shedding virus that could infect another person and yet has a negative PCR 

3. How often is someone negative on LFT and positive on PCR?

  • This is the question that most worries people:  indeed, it is well recognised that an LFT may be negative when a PCR would be positive 
  • From the previous diagram above there are two main periods of times when this occurs:
    • At the end of the incubation period, and especially if people do not have symptoms
    • After  around 7-10 days, when the PCR continues to be positive but the virus has stopped producing new proteins in the nose
  • There have been a very large number of studies to answer how often a PCR is positive and the LFT is negative with very differing results 
  • The answer is indeed not simple as it depends on:
    • Whether symptoms are present or not: LFT is much more likely to be negative if there are no symptoms 
    • The time since the onset of infection: LFT is much more likely to be positive in the first five days of symptoms 
    • The threshold for a PCR positive
      • The result from an LFT is just either negative or positive.    
      • By contrast, PCR is very sensitive and can pick up very tiny concentrations of virus RNA even if only a very small amount is present in the swab.  
      • In the lab the PCR process can be tweaked to increase the amount of RNA.  
      • The likelihood that viral RNA is  detected by the PCR test can then depend on how much tweaking (so-called ‘amplification’) is needed to get enough RNA for a positive result (this is the nature of the science behind PCR, see Appendix)
      • LFT is less likely to be positive the more the PCR test needs to be amplified 
  • I have summarised the results from a publication that looked at 48 studies in the chart below
  • The percent refers to what proportion of the samples that were positive for PCR were detected as positive on LFT
Taylor et al, Cochrane Database System Rev 2021 (3) CD013705
  • Thus, for people with symptoms and a positive PCR (the first left hand blue column), around 70% would have a positive LFT, compared to under 60% without symptoms
  • Similarly, in those with a positive PCR,  LFT is more likely to be positive in the first week of infection than in the second week 
  • Thirdly LFT is much more likely to be positive if the PCR is positive without the need for lots of amplification 
  • A completely separate point relates to the fact above that the PCR test can be positive long after someone stopped being infectious 
    • This is especially a problem for interpreting the result of a ‘routine’ LFT in someone without symptoms for say a work or social engagement
    • A PCR done at the same time could be positive and the LFT negative
    • The problem is we cannot know with an asymptomatic person like this, we cannot know whether their PCR positive means an old or a new infection
    • If the latter is the case, then the negative LFT, is giving a ‘correct result’ about not  being  infectious


  • Apologies if I have made a simple question seem very complex!
  • Given the impossibility of population wide PCR testing, LFT are a very reasonable substitute but the interpretation of a negative result needs to take account of the timing since the possible contact and the onset of any symptoms
  • LFT needs to detect people who are infectious so failure to detect people who are positive on PCR after the infectious phase has passed is not a concern – although the timing may not be known
  • If a very sensitive approach is used for the PCR to pick up tiny amounts of virus RNA, then such people, although positive, may not pass on the infection and hence a negative LFT may also not be a problem
  • For sure after a definite contact with an infected person, compared with LFT, PCR is more likely to be positive sooner and before symptoms develop
  • Maybe the bottom line is there is no perfect test of infectivity and to control the spread of a highly infectious variant such as Omicron, that widespread LFT testing is no substitute for mitigation measures (mask wearing etc)  


How does a PCR test work?

  • The Covid-19 PCR test aims to discover the presence of evidence of RNA from the virus in a swab from the nose or mouth
  • PCR stands for polymerase chain reaction and is a laboratory method for making very large amounts of DNA or RNA from tiny samples (we use the term ‘amplification’ to describe this)
  • Thus ‘in the test tube’ what started off as a very small quantity of RNA is increased to produce a much larger amount
  • This process is used in police investigations to identify suspects from minute samples dissolved from blood stains – as you will know from any detective TV drama!
  • For Covid-19 the PCR technology can also  identify very specific strains – which is how Omicron was identified in the first place
  • The other aspect of the PCR test to be aware of is that the lab can also measure how easy it is to find RNA ie how much ‘amplification’ is needed.  The more amplification, needed to get enough RNA the lower the amount of virus genetic material in the original sample

How does the rapid antigen (lateral flow) test work?

  • The LFT test aims to discover the  presence of proteins made by the virus, also from nose and throat swabs
  • (We now know that Covid-19 is much more likely to grow in the nose, that a throat swab is not really necessary) 
  • The presence of the virus protein is detected by seeing if a sample of nose fluid reacts with a specific laboratory-made antibody
  • The test works by the absorbent paper strip inside the plastic cover being impregnated at the ‘T’ level with this antibody 
  • If the sample has such proteins, the antibodies on the strip react with them, produces a reaction and a line appears as on the right below
  • Whether or not there are  viral proteins in the sample, the sample drops will also dissolve the antibodies at the T mark, and then the drops spread to the C mark
  • The antibodies react with a dye at the C mark to produce the red line as shown
  • A positive LFT will produce both a line at T and at C, indeed if there is no line at either, the test has not worked or been done properly 

If you would like to receive email notifications of new posts on this blog just click this link and enter your email at the bottom of the home page:

3 replies on “Are self-testing rapid antigen (lateral flow) tests for Covid-19 accurate enough”

There is a lot of anecdotal evidence that LFTs are less effective or inaccurate at correctly giving positives for Omicron. The supposition is that Omicron is growing more in the throat than the nose and current LFTs are nasal only. It has now been backed up by a study from UCL.


The reason for the anecdotal evidence of a lower sensitivity is the impact of vaccine coverage on the occurrence and severity of symptoms with Omicron. Biochemically there is no difference (or any that I have found) to suggest that the antibody in the test strip of most LFT’s would not detect viral antigens produced by Omicron to the same extent.

I have not seen any studies showing that it is harder t culture Omicron from the nose in people with Omicron than people with Delta. Coronaviruses do have a predilection for nasal mucosa, and indeed nasal symptoms do seem to predominate with Omicron


The post is very informative and helpful. Whatever procedures are used to conduct the current testing frenzy, the results of it are dire. Widespread subsequent isolation is crippling normal life, business and the economy. The tests are akin to conducting an opinion poll to check if we are being invaded when the streets are already full of enemy troops!


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s