Thread by @COVID19Tracking

We want to share a new resource with followers of this project: a knowledge base for SARS-CoV-2 antibody testing: airtable.com/shrGVEvpWp12Ry….

Thanks to our friends at @airtable for the free Pro plan. Your product has been indispensable for building this knowledge base!

In our Airtable base you can find several tables:
- Vendor lists tracks lists of vendors compiled by regulatory bodies like the @US_FDA as well as non-profits like @FINDdx.
- Vendors tracks a subset of the many vendors who have products available.

- Papers, Press, and Tweets gather references to vendors from the scientific literature and popular press, as well as a few Tweets that have useful content.

Please note this knowledge base is very incomplete! Our notes on papers are very thin, we have only documented a small fraction of the vendors listed by the FDA and FIND, and we have yet to document individual products made available by each vendor.

Nevertheless we've decided to share this base publicly now as we have discovered many people in this rapidly evolving market are not aware of its scope.

Please do your own investigations before acting upon the data in this knowledge base.

In the process of putting together this knowledge base, we have discovered a few things about the antibody test market.

The FDA has only issued a single EUA to a vendor in this market. fda.gov/medical-device….

Many more vendors have notified the FDA that they have validated their tests; expand the "Q: What serology tests..." header at fda.gov/medical-device… to see the full list.

The market to purchase these tests in bulk is crowded and information asymmetry is extreme. We've heard several stories of hospitals and other smaller government entities being outbid by larger government buyers. One vendor tripled their price overnight.

There's also intense competition for samples from confirmed COVID-19 cases so that new vendors can validate their tests.

We need a nationwide clearinghouse for samples and tests.

We also need a nationwide effort to validate these tests.

Spain learned that tests provided by Bioeasy had very low sensitivity and had to return their purchase.

theguardian.com/world/2020/mar…

An Italian emergency department found very low sensitivity for tests from VivaChek onlinelibrary.wiley.com/doi/abs/10.100….

We also need more information from the test makers. Many do not disclose which antigen(s) their antibody test uses, nor the means of production or isolation of these antigens.

There's also a debate over which isotype(s) to target. We've seen total antibodies, IgM, IgG, and IgA. Each has a different time course and optimal type of sample. Which is best for which purpose?

A major concern is that specificity is not tested using samples known to have antibodies against antigens from one or more of the 4 coronaviruses that cause the common cold.

It's also important to note that antibody tests that can distinguish between vaccine-acquired immunity and naturally acquired immunity will be critical for the design of vaccine clinical trials:

https://twitter.com/mlipsitch/status/1245198825825984512

If you are interested in helping grow our antibody test knowledge base, please respond to this thread and we will try to be in touch.