Active Virus versus Antibody Testing
There are two types of COVID-19 testing. Molecular testing verifies that a person has an active virus infection. Antibody testing verifies that a person has been exposed to the virus at some time in the past. These are two different things.
Molecular Testing: Detecting the Active Virus
The so-called molecular tests that rely on nasal swabs detect virus RNA and are relatively accurate. A positive generally means the patient has COVID-19, right now. This test type will only establish the current infection status of an individual.
To get a better handle on the pandemic, we are trying to increase the number of molecular tests. Unfortunately, until testing is broad and uniform, positive tests will be misleading. The more people we test, the more positives we will get. For now, hospitalizations and daily death rates, the so-called lagging indicators, will be more valuable. Uniform testing of symptomatic people is quite valuable if the sampling method is balanced and biases are removed. These metrics allow us to accurately gauge the present state of the infection in a given area.
Trying to track a low level infection rate with random sample testing will require unrealistically high number tests, for example, if the actual infection rate in a population is 0.01%, that means one person in 10,000 is infected. For a population of 1 million, we would have only 100 people actively infected. To get an accurate measurement resolution for this infection rate we would need to test many thousands of people every week.
Antibody Serology Tests: Detecting Who Has Had the Virus
The other types of tests are called serology tests; they detect antibodies in the blood. Here, we are in wild, wild west territory. Because of the emergency associated with the pandemic, the FDA has greatly loosened the restrictions on the tests. There are scores, if not hundreds of tests available. To be legal, the test maker only must submit an application to the FDA. Once it is in the mail, they are free to go at it.
The FDA may issue an emergency use authorization (EUA), which it has done for about 10-15 antibody
tests. The EUA is not a full approval and does not involve the usual regulatory scrutiny. However, it is better than nothing.
Besides the lack of full vetting of the antibody tests, there are other concerns:
First, we do not know if the antibodies that are measured can prevent viral particles from infecting. That’s right, all antibodies are not equally effective. Imagine that the virus is a big, strong running back, running down field. A burly linebacker with a good hit will take the running back down—that would be like a neutralizing antibody. However, a corner-back, while also a defender, might hit the running back but not stop him—that would be like a non neutralizing antibody.
Second, even if a patient develops neutralizing antibodies and immunity to infection, we know that with some other coronaviruses, immunity is not lifelong. Next year, the patient might be susceptible, again.
Finally, there is concern that some of the antibody tests will register “false positives” because they are fooled by antibodies to other coronaviruses. If this gives you a corona sized headache, you are not alone. Doctors and scientists desperately want reliable, meaningful antibody tests.
Antibody testing is valuable for the purpose of establishing general exposure history, helping us to understand how deeply the virus has penetrated a population. This data is useful to calibrate our models and determine the percentage of susceptible individuals in a population. The errors associated with individual results are less of a concern for these purposes.