COVID-19 Antibody Testing: What We (Think We) Know and Don’t Know

IMG_3750

Friends, my practice has sent this letter to our patients.  It is a summary of what we think we know about COVID-19 antibody testing, as of today.  This is our collective opinion and expression as a practice.  We do not speak for our employer or colleagues at large.  It is our best interpretation of the available evidence to date, and we present it in good faith.  Information evolves rapidly, and we expect to update our position and practices accordingly.  Please vet your information sources well, and make any and all medical decisions in collaboration with your primary care physician.

Many states are likely to extend shelter in place orders, albeit with some slow loosening of restrictions.  I worry that this will incite further unrest and divisions along ideological lines.

We all must now call forth our highest practices of patience, generosity, and love, so as to pull together better rather than separate further in body, mind, and spirit.  Hang in there, my peeps.  We can do this.

* * * * *

COVID-19 Antibody Testing

Covid-19 antibody testing is being promoted as a way to tell who has already had COVID-19 and therefore may be immune. We are getting many inquiries about the availability of antibody testing and wanted to let you know about the current state of antibody testing.

If you have had COVID-19, are you immune to it?

We don’t know the answer to this.  What we do know:

  • Most other viral infections stimulate your body to make antibodies against the virus which provide some degree of protection from re-infection for some period of time.
  • Antibodies to the common cold, which is sometimes caused by another type of coronavirus, seem to last only 1-3 years. Antibodies to measles generally last a lifetime.
  • There have been a few reports of seeming re-infections with COVID-19 in China and South Korea, but this is thought to more likely be due to a testing problem rather than true re-infection.
  • It’s very likely that people who have had COVID-19 should have some immunity for some period of time, but nobody knows for sure.

Does the severity of COVID-19 illness or the levels of antibody matter for presumed immunity?

When you are infected with a virus, your body makes many different types of antibodies against many different parts of the virus, in differing quantities. The IgG class of antibodies is the one that tends to provide long term immunity. Scientists are currently looking at all of the different antibodies present in the plasma of people who have had COVID-19, to see which antibodies seem to be the most numerous and react most strongly against the virus.

We currently don’t know which antibodies are most protective, how many of the different antibodies you need to have to be protected, or whether the levels of the antibodies matter for either degree or longevity of protection.

What do we need to know about a test before we call it a good test?

After the onset of the pandemic, the FDA allowed institutions and companies produce their own tests, provided they used an FDA-approved procedure to validate the tests. It’s not at all difficult to make an antibody test . . .but it’s very difficult to make a GOOD antibody test. You have to know:

  • Which antibodies to look for
  • How many antibodies to look for
  • What it means if some antibodies are present but not others (what if 2 of 4 tested antibodies are found -does that mean you are immune?)
  • How good your test is at picking up the people who have a positive test and are truly immune (the positive predictive value of the test)
  • If your test correctly tests negative when people have NOT been infected (the negative predictive value)
  • If your test is specific to COVID-19, or if it shows a positive result by detecting antibodies from infection with a different virus in the last few years (many common colds are caused by other types of coronaviruses).
  • That your test is valid -meaning you have reliably answered the questions above in as many people as possible. Generally this requires hundreds of people known to be positive and negative, as well as some who had other upper respiratory illnesses.

Finally, it is VERY important to remember what we DON’T KNOW:

  • Whether having antibodies and which type of antibodies actually provide immunity
  • If it does, how long the immunity lasts
  • If you have antibodies and have a new virus exposure, whether reinfection can occur
  • If you have antibodies and have a new virus exposure, whether you could still potentially transmit it to non-infected people

If scientists are still studying the antibodies, why are there hundreds of antibody tests already on the market and one being done by a drive-through facility in Chicago?

Many of the tests currently on the market are imported from China or Europe, and some have been made by small US companies who have rushed to produce a test (again, it’s easy to make any test; hard to make a good one).

We have investigated a number of these tests to see how they have been validated and how reliable they are, and the answer is that all of these tests are remarkably poor. (For those who want the scientific details, see below.)

NONE OF THESE TESTS HAVE BEEN VALIDATED OR APPROVED BY ANYONE OTHER THAN THE PEOPLE WHO  MADE THEM, AND THEY CAN DO ANY KIND OF TESTING THEY WANT, WHICH IS LARGELY GROSSLY INADEQUATE.

OK, so if what’s available now is terrible, will there be good antibody tests, and when?

YES! There will be good tests, likely in several weeks. Abbott has a test they are working to validate, as does Roche. When a RELIABLY GOOD, adequately validated antibody test is available to the general public, we will let you know. We anticipate that the first tests will be used to test healthcare and other essential workers, and then as production increases and reliability confirmed, extended to test the public.

Just a reminder . . .

Continue to wash your hands well and frequently, especially if you have been out in public.

Continue to stay 6 feet away from anyone if you leave your house.

We recommend wearing a mask if you are out in public. Remember that the mask protects others from you, and does not necessarily protect you from others . . .so the 6-feet distancing remains very important!

The nitty gritty scientific details, for those who may be interested . . .

Many of these tests look for antibodies to the coronavirus ‘spike protein’, the part that attaches the virus to human cells. The spike protein is very similar across all coronavirus species, so the risk for false positives is high in people who have had the common cold in the last few years (which is all of us).

The test currently being offered by a drive-through in Chicago is made by a German company, Euroimmun.  A recent paper examining its performance found that its sensitivity (meaning the test both accurately found positive and negatives) was only 67%. Put another way, a full third of people had test results that weren’t accurate. The positive predictive value -meaning if you test positive, the likelihood that you actually had the disease -is only 82% (so 18% of people think they are immune to COVID when they are not), and the negative predictive value is 87%.

Another test that is being marked by Vibrant America for $149, was ‘validated’ in a total of only 20-30 patients, which is far too few to claim reliable test performance. It tests several antibodies, each with a sensitivity of only 65-80%. The company doesn’t say how they interpret  a mixed positive/negative result (indeed no one knows how to interpret this right now). Finally, in the small print, the company notes that their test may be positive in people who had common colds in the past.

4/23/2020 11:10pm CDT–  Updated to add:  Please click/tap to find the formal statement on SARS-CoV-2 antibody testing from the Infectious Diseases Society of America (IDSA).  Bottom line:  There are no reliable tests at this time, and none of them should be used to make individual diagnostic or screening decisions.  Also, answers to myriad questions about antibody response are required for vaccine development and testing, so that will likely take many, many months (I expect closer to 18 months or longer, than the 12-18 we have all heard).