SARS-CoV-2 antibody rapid tests: best practices to minimize false positives
Practical, not medical, advice.
[Originally written as a sidebar for my article on the UCSF “white stick” evaluation, 29 April 2020. Here.]
This is practical, not medical advice. The usual warnings apply.
1. If you (a) want to use a finger-stick rapid test to determine if someone has blood antibodies to SARS-Cov-2, and (b) want to reduce the risk getting a false positive, this is what you do:
- Start with a reputable test. There’s no Consumer Reports buying guide yet, but I have one coming soon. It is extremely stupid to dismiss all Chinese tests as “junk.” One (Wondfo, a well-known and respected pharma company) tested in the top tier. If you must buy American, there is BioMedomics, manufactured in North Carolina. There are good tests from South Korea (SD Biosensor), Singapore (Biolidics), Ireland (Biopanda) and, coming soon, Germany (PharmACT AG).
- If you’re going to prick a finger for the blood sample, make sure you know how to do it right. Like poor throat-swabbing can be for the RT-PCT tests, simple and stupid things can trip you up. I watched a “how to” video aimed at nurses, and learning all kinds of things, like why you discard the the first drop of blood to avoid collecting skin cells.
- After putting in the blood and solution, if the control line fails to appear as expected, trying adding more diluent. Some brands “run dry.”
- Forget about IgM. The IgG test line on the stick the one you want. If you happen to also get IgM, think about recent infection (not necessarily SARS-Cov-2).
- The line for IgG has to be real dark, not faint. This is the “Mass General” trick, named after the Boston hospital, and appears to be a good one.
- Don’t give the test sooner than 28 days after the onset the Covid symptoms. Maybe 21. As I said in one of my earlier posts, the question to ask about these tests is not so much how sensitive they are, but when then are sensitive.
- To confirm with a second test, consider the (better) option of using venous blood and (probably better) lab-based ELISA tests. Mix it up.
- Ignore any of these rules if you’ve studied up and think you know better. I’m not a doctor, just a guy who reads carefully.
2. If you are, say, a health- or elder-care worker and need to make absolutely sure you are IgG antibody positive:
- Pick two different of the most specific — but not necessarily the most sensitive tests — you can find, and take both of them. Positive on one OR the other = positive.
- This is how Institute Pasteur does it, only they take 1 out of 3!
3. List of things that might (possibly, but not very likely) mess up your test result:
- You’ve had a cold recently
- You’re pregnant
- You have rheumatoid arthritis. (You would need a very high rheumatoid factor. High levels of rheumatoid factor in the blood are most often associated with autoimmune diseases, such as rheumatoid arthritis and Sjogren’s syndrome.)
I will happily delete this post when the CDC or FDA steps up with better advice.
— Will Bates
Comments to knowledgeengineering AT me DOT com
The “We’re not making this up” department
In a study posted on-line 5 May 2020, Dr. Ania Wajnberg and others from the Icahn School of Medicine at Mount Sinai, New York NY:
28 days
This suggests that the optimal time frame for widespread antibody testing is at least three to four weeks after symptom onset and at least two weeks after symptom resolution.
No need to rush to get antibody tested
In our survey, we did not find evidence for a decrease in IgG antibody titer levels on repeat sampling.
If you’re only suspicious you had Covid-19, you may be wrong
They tested a group of people who were either (a) told they might have it but were not PCR tested; (b) lived with someone with confirmed infection, or ( c ) were healthcare workers.
we found 36% of this population had IgG antibodies to SARS-CoV-2
