Trying to determine whether someone has antibodies against the new corona virus is important for health authorities, but this approach faces many hurdles.
Technology is imperfect. Anyone who has ever used a computer knows that. You’re looking for a file that you know is on your hard drive, and somehow the search function can’t find it. If computers are faulty, is it any wonder that diagnostic and screening tests are also done?
There’s a lot of hope with serological COVID-19 testing. “Serological” means that it is the serum, the liquid component of our blood that has been freed from clotting factors but contains proteins such as antibodies: It is these antibodies that are destined for our salvation. Antibodies are produced by a subset of our blood cells in response to infection. If you have antibodies against the new corona virus in your blood, this must mean that you had the virus in your body and managed to fight it off. The presence of these antibodies protects this corona virus from disease a second time (although this is not yet guaranteed, but small studies in chimpanzees are encouraging) and shows whose body is now “shielded” from COVID-19 by those who are immunologically naive could prove critical to health authorities.
However, tests are technological in nature and the technology is incomplete. For example, a COVID-19 antibody test result could be wrong.
A new gold rush
The first limitation of antibody tests is that they depend on our immune system. They cannot detect something that is not there, and our blood cells do not make antibodies against the corona virus once we are infected. In fact, it takes between two and three weeks after the onset of symptoms for enough antibodies to be formed to be detected (although in some cases it may only take four days). Someone in their first week of infection would most likely get a negative result from an antibody test. Therefore, unlike diagnostic tests, where the virus is collected with a long Q-tip and its genetic material is detected, these special tests are best used to Identify people who had the virus but no longer do so It is also important to note that not everyone who encounters the virus develops antibodies against the virus or enough antibodies to be detected by the test.
Another major problem is incompetence. The universal demand for these antibody kits has created a new gold rush, and just as an inexperienced buyer may believe pyrite is gold, unqualified companies are developing tests of dubious accuracy to profit from the demand. Some of the COVID-19 antibody kits on the market were made by companies with no experience in medical device manufacturing (one such company usually sells Chinese herbal medicines, while another sells solar panels and tires), and independent tests showed that their accuracy was no better than flipping a coin (and a coin is much cheaper, after all). Health Canada was very cautious and only allowed two kits at the time of writing. However, the US has opened the floodgates to companies wishing to sell their serological tests and informed them that they do not have to go through a lengthy approval process: Just let the FDA know what your kit can and cannot do, and we will allow you to sell it. Since then, they have corrected their course and expressed concerns about fraudulent kits. Pyrite may not do much damage, but a false sense of immunity in the middle of a pandemic can be fatal.
A final problem with these antibody kits – and with any diagnostic or screening test – is that they are never 100% sensitive and 100% specific. Using the analogy of a home pregnancy test, you want the test to tell you that you are not pregnant, only if you are not pregnant, and that you are only pregnant if you are pregnant. You want all positive test results to be true positives and all negative test results to be true negatives. But at the same time you can ask for world peace, because these ideals simply cannot be achieved in the real world. For example, a kit might tell you that you have antibodies against the corona virus, but it turns out that your antibodies were mistakenly detected against another type of corona virus, such as one of the many that cause colds.
Health Canada asks kit manufacturers to specifically test this and ensure that this does not happen. However, these pre-approval tests are limited in scope and the technology is fallible, so this is still possible.
Even if an antibody test kit has a very high sensitivity (i.e. it detects antibodies when they are present) and a very high specificity (meaning it gives a negative result when antibodies are not present), there is another problem that needs to be addressed and it involves some eye-opening math.
One number changes everything
Pregnancy tests at home are not accurate, but they are quite reliable. This is partly because the people who use them are more likely to become pregnant. They often use them because they have had sex and their period has not happened, so they suspect that they are pregnant. However, if you were to give this test to any person who could give birth to a child on earth, the accuracy of the test would decrease. And it boils down to one important concept: how often what you are looking for is in the population you are testing.
Let’s take one of the two COVID-19 antibody kits approved by Health Canada, with a sensitivity of 97.4% (this almost always indicates that you have antibodies to the corona virus) and a specificity of 98.9% (almost always) Say you don’t have antibodies if you don’t. Sounds like an almost perfect kit.
Imagine a scenario where only 1% of the Canadian population has antibodies against the corona virus. If each Canadian’s blood is tested for antibodies and you personally get a positive result (meaning that you are told that you have antibodies and may now be immune), there is a 47.2% chance that your result is accurate (a coin flip). However, if 5% of the population has these antibodies and you test positive, there is an 82.3% chance that your result is correct. If the percentage of the population exposed to and developing antibodies against the corona virus increases from 10% to 25% to 50%, the probability that your positive result is actually similar increases from 90.8% to 96.7% to 98.9%. . The more people have antibodies against the coronavirus, the more likely it is that your positive result from this one test is correct. The accuracy of a negative result is similarly affected by this phenomenon.
Even with a robust kit, the results you are looking for are rather unreliable if what you are looking for is rare. According to this logic, a random test with an antibody kit at the beginning of a pandemic leads to results that are not as reliable as those generated towards the end of a pandemic. For this reason, the Centres for Disease Control in the United States have recently recognised that the use of antibody tests – even more reliable – in populations where very few people are thought to be infected with the coronavirus leads to inaccurate results.
This is also the reason why mass screening for cancer is not as beneficial as it sounds: the more people without cancer you screen, the more false positives you get, leading to anxiety and more invasive testing. Since COVID 19 serological tests are used to screen the population, we need to consider the many ways in which they can be mistaken. They have their place, but as imperfect technologies they have their shortcomings, both in their design and in their application.