To some degree, the winding down of Covid-19 restrictions at Kendal depends on the availability of testing. We don’t have any internal testing capability so far. In Lisa’s latest message to residents, she discussed “our future desire to be prioritized for additional testing for staff, health center residents, and eventually residents who live in our independent homes.” She also mentioned that Chester County has received 20,000 “antibody tests”, but that they haven’t been distributed because of “bureaucratic disputes and regulatory issues”.

In anticipation of our testing, I have been reading up on the nature of the tests available on the US market so far.

Two kinds of tests. Much of the Covid-19 news these days discusses the results of tests, primarily the counts of “confirmed cases”. Confirmed cases are those in which a sample from the patient, usually a nasal swab, shows the presence of the virus. That’s one type of test.

There is a second type of test, a test for antibodies to the virus in a person’s blood. The presence of antibodies indicates that the person had Covid-19 previously. These tests, even if antibodies are present, are not counted when “confirmed cases” are reported.

It all sounds pretty precise, but it’s not. There are problems with both types of tests. The antibody tests are particularly problematic. Let’s look at each type in turn.

Tests for the virus itself. These tests are used with sick people, to see if the Covid-19 virus is present and causing their disease. A nasal swab is used to collect material from their nasal passages and upper throat. The genetic material is extracted from the virus particles, amplified, and chemically matched with known viral gene sequences. Test results can be available anywhere from 15 minutes to a day later. A more detailed scientific description of the testing process is available here.

The fastest of the tests, called “ID NOW” from Abbott Labs, returns results in under 15 minutes. But NPR reports that there are problems with the test. Clinics report both false negatives (the test does not find Covid-19 in people who actually have it) and false positives (the test reports that Covid-19 is present in people who don’t have it). One doctor at the University of Pittsburgh is reported as saying that he has stopped using ID NOW because of these inaccuracies.

The Cleveland Clinic also ran trials of various tests and has also stopped using ID NOW and another test, DiaSorin Simplexa, both of which identified less than 90% of actual Covid-19 cases. They continue using two other tests (from Roche and Cepheid) that correctly identified the presence of virus in over 96% of cases. Only a lab test devised by the CDC correctly identified 100% of Covid-19 cases.

This means that even when we get virus testing, it may not be be 100% accurate. Testing everyone, isolating those who test positive, and tracing and testing all their contacts would catch the vast majority of cases, but a few might be missed. To prevent spreading, we may still need to retain some measures besides testing.

Tests for antibodies. Compared to the virus tests, the situation for antibody tests is much murkier. The CDC is basically allowing any company that claims to have an antibody test to sell it, without verification. Doctors and clinics have to try to figure out how much faith to put in the results of a given test. A handful of the tests now have CDC endorsement, but most don’t.

Some research facilities have been trying to compare the main antibody tests to see which ones really work. One such series of tests, conducted at the University of San Francisco and at Harvard Medical School, evaluated 14 antibody tests. The results were written up as a publicly available “preprint” (not yet peer reviewed) and summarized by the New York Times.

All 14 of the tests were able to detect antibodies to some degree, but some were very disappointing, especially in their reporting of “false positives”—reporting that people had antibodies when they actually didn’t.

Some tests had false positive rates of up to 14%. That’s not very useful. What I want to know is whether I’m immune to Covid-19. It’s a lot less helpful to know that I’m probably immune, but there’s a 1 in 7 chance that I’m not. Given those odds, I would probably conclude I couldn’t afford to change my behavior at all, and I imagine my friends still wouldn’t want me breathing on them.

Seven of the 14 tests had false positive rates of less than 5%, which is a level that some doctors interviewed by the Times found somewhat useful. Three had less than 1% false positives. Only one test, from a company called Sure-Bio, had no false positives at all.

This group of researchers, and others, will continue to evaluate the rest of the available antibody tests. Soon, there will begin to be an understanding of which tests are worth continuing to use.

One notable result of the testing was that no antibody tests are reliable until about three weeks after infection. It takes that long for your body to produce antibodies in large enough quantities to be certain of detection.

Do antibodies guarantee immunity? Suppose you are somehow able to confirm that you have antibodies to Covid-19. Does that mean you are safe from further infection and can resume “normal life”? Unfortunately, it doesn’t.

The World Health Organization is warning that there is no proof that you can count on immunity to Covid-19 even after you have recovered from it. Korea and China are reporting that a few percent of people who had recovered from Covid-19 have tested positive again. Is this because they weren’t actually immune after recovering from Covid-19? Or could it be that the original diagnosis or test results were wrong in these cases? Or can you get a relapse of your original Covid-19 case after a period of recovery? Or is there immunity, but it just doesn’t last very long? So far, no one knows. There are still major questions about immunity.

What would be the implications if antibodies do not provide immunity to this virus? It would dash our hopes for a vaccine (which would work by causing the production of antibodies), and it could mean that many of the “temporary” restrictions we are coping with would have to last much longer, until other treatments or preventative measures are developed, or until the virus evolves into a less virulent form.  

Much depends on the answers to the questions surrounding immunity. Hopefully, we will get some of those answers soon.