Test, test, test is becoming the new mantra.
It has been identified as one of the five key steps the government wants in place before it starts to ease lockdown and they are as follows:
1) The NHS has the capacity to provide critical care right across the UK
2) A sustained and consistent fall in daily deaths from coronavirus
3) The rate of infection decreased to manageable levels across the board
4) Operational challenges including testing and PPE are in hand, with supply able to meet future demand
5) Confident that any adjustments to the current measures will not risk a second peak of infections that overwhelms the NHS (1)
Testing for COVID-19 has become a sticking point in this pandemic. I wrote here about why some countries, Germany and South Korea in particular, have managed to test far greater numbers of people from the outset.
The UK government had promised to ramp up testing to 100,000 tests a day by the end of April and did meet this target with 122,000 done on April 30th although it has dropped below the target again since.
There is a lot of confusion over the actual testing process. News footage shows people having their nose and throat swabbed “looking for the virus” but then another news item talks about the need for blood testing to establish if a person has already had COVID-19. Terms such as "antigen", "antibodies" and "gamma globulin levels" can be confusing and are not always used correctly in the media. Then the terms “sensitivity” and “specificity” of a test are used interchangeably and, wrongly, thought to reflect the accuracy of a test.
This piece aims to explain these terms and how they apply to testing for any disease and where we are at with testing for COVID-19.
Antigens and Antibodies.
Antigens are complicated protein molecules that are attached to the outside surface of viruses. They are rod or spike-like structures that protrude from the surface and, in the case of the coronavirus causing the current pandemic, they are the red club-like structures sticking up from the grey orb.
The antigen spikes allow the virus to attach itself to human cells and invade them. The better the spikes are at attaching themselves to human cells the more infectious the disease is.
(Note that “more infectious” means how easily something spreads person to person not “more serious”. The common cold is highly infectious because its antigen spikes attach very easily to human cells but it is not serious. Hepatitis B is less infectious but far more serious, the antigen spikes have difficulty attaching to human cells but when they do they cause a lot more harm.)
Antigens are unique to each virus, those on the measles virus are different to those on the coronavirus for example. This is why vaccination against one disease does not protect from other diseases. Sometimes the antigen is capable of changing its shape and structure which makes it much harder for the body to mount an antibody response or for scientists to make a vaccine. The HIV virus is a good example of this.
Antibodies are produced in the blood by the body’s immune system in response to an antigen. They are also made of protein and lock onto the antigen spikes thereby neutralising them. This either kills the virus directly or immobilises it to allow specialist white blood cells to come and destroy it.
Vaccines aim to mimic this antibody response to an antigen and you can read more about that here.
Antibodies are referred to scientifically as “Immunoglobulins” (abbreviated as Ig) and are sub-divided into five main classes - A, D, E, G and M which are notated as IgA, IgD, IgE, IgG, and IgM respectively. The antibodies produced in response to viruses are usually a mixture of IgA, IgG and IgM. The amount of each produced varies from virus to virus and the stage of the illness the person is in.
IgM is always produced first and then starts to tail off. As it does so IgA and IgG are produced. IgA is found on mucosal surfaces such as the mouth and in saliva. IgG is found in the blood and it is IgG antibodies that are the backbone to the immune response.
IgM begins the fight against infection kicking in immediately whilst the body is making IgG. After approximately four weeks levels of IgM have dropped off and the IgG takes over in fighting the infection. It is IgG that provides long term immunity after a person has had the disease or been vaccinated against it.
To bring all this back to the COVID-19 pandemic and testing, decisions have to be made as to exactly what to test for.
The whole virus? The antigen spikes? The antibodies? And which antibodies...IgM or IgG?
Antigen Testing - Done with Nose and Throat Swabs
It is extremely difficult to grow viruses in culture in a laboratory but with recent scientific advances it is now possible to test for a virus’s genetic make-up, its DNA or RNA, rather than the virus itself. Scientists have been able to identify the coronavirus’s RNA by analysing the antigen spikes.
The COVID-19 tests currently being used in the UK are looking for this RNA. Swabs are taken from the back of the nose and throat and then go through a labour intensive process as they are analysed to see if the coronavirus RNA is present.
However, a positive swab test only indicates that a person is infected right now. It does not give any indication of whether or not a person may already have had COVID-19 and is now clear. Also, false negatives can occur in up to 30% of tests, for example if the swabs are not taken correctly, so they cannot be fully relied on. The WHO recommends a repeat swab if the first one is negative but the person has symptoms suggestive of COVID-19. (2)
It can be seen that this type of testing is limited, it cannot tell you who has already had the illness and its result, if negative, cannot be relied upon. Its main value has been identifying those who are significantly unwell with COVID-19 where a positive result has allowed the correct medical decisions to be made in hospital and enabled doctors to differentiate COVID-19 from other illnesses.
Antibody Testing – Done with Serology (blood sampling).
To know if a person has had a disease and is immune we need to look for antibody levels.
A positive IgM test means that the infection is recent and ongoing but a negative IgM does not tell you if the person has already had it. For this an IgG test tells you more, if positive it means that the person has had the disease, probably several weeks or even months ago.
If you only test for IgM and there is none present either the person doesn’t have the disease, or has had it but the IgM response is gone as it disappears from the blood after a few weeks.
If you test only for IgG and there is none present you can conclude the person has not had the disease in the past but, as there is no traceable IgG in the first few weeks of illness, it is possible for the person to actually have the disease at the time of testing.
This is why the ideal antibody test needs to look for both IgM and IgG.
And herein lies another problem. Devising accurate antibody tests is technically very difficult because you need to know what you are looking for. That sounds obvious but before any test can be produced to look for antibodies to COVID-19 infection, scientists first have to study blood samples from people who have tested positive for the infection (from the swab testing) to find those antibodies and understand their structure. Once the antibodies have been identified they need to be analysed and tested to be absolutely certain they are the ones that will fight the COVID-19 infection. Then scientists also want to know what levels of antibodies are needed to confer immunity to the disease and whether that immunity lasts for a long or short time.
This complex process is why antibody testing is not a quick and easy solution.
Sensitivity and Specificity.
Once a test has been devised scientists talk about its sensitivity and specificity. These terms can sound very good if a test is described as “90% sensitive” or “85% specific”. However these terms are not related to accuracy.
Sensitivity measures how often a true positive is found. In other words the person definitely has the disease and the test for it is positive. If a test has 90% sensitivity that means 90% of people with the disease get a positive result...but 10% of people who have it get a negative result, a false negative.
Specificity measures how often a true negative is found. In other words the person definitely does not have the disease and the test for it is negative. If a test has 90% specificity that means 90% of people who do not have the disease get a negative result...but 10% of people who also don’t have the disease get a positive result, a false positive. (3)
These over-estimates are important as they can heavily skew the numbers.
For example take a test with 90% sensitivity and 90% specificity.
Imagine a population of 1000 people has a 10% infection rate.
That means 100 people have the infection but 900 do not.
Of those 100 infected people a test with 90% sensitivity reads positive for 90 of them (true positives) but reads negative for 10 of them even though they have the disease (false negatives)
Of the 900 people who are not infected a test with 90% specificity records 810 of them as negative (true negatives) but reads positive for 90 people even though they do not have the disease (false positives).
So when you have tested all 1000 people you have 180 positive tests in total (90 true + 90 false) but only 90 people really have the disease the other 90 do not.
So, only half of the people with a positive test result have the disease, but unfortunately you don’t know which half.
These numbers hopefully show that the best tests have sensitivity and specificity rates of 99% or higher. Some of the tests that have been promoted have had rates of just 60% which renders them in effect useless but some countries, including the US, have allowed them to be marketed. (4)
There has been a whole flood of tests launched onto the market by different companies in recent weeks but many have not been scientifically validated and tested. The WHO maintain a list of the number of antibody tests currently available and there are in the region of 275 different tests. With so many it is hard to know which of them are good with both a high sensitivity and a high specificity. The WHO is working with laboratories around the world assessing them but this takes time.
Here in the UK the rapid tests where a pin prick of blood is place onto a stick (rather like how a pregnancy test works) that looks for IgM and IgG are being touted as ideal for the general public to use and many are for sale on line. My advice would be to avoid them, they do not appear to be working well with many false positives and false negatives.
When a scientifically validated test with high sensitivity and specificity rates is available we will know about it, not least because here in the UK the government and scientists alike will be wanting to test as many people as possible.
