Samizdata

Simon Gibbs of Libertarian Home renown has some interesting information about Wuhan Coronavirus testing.

By the end of this article, I want you to be writing to someone and sharing this link. That person could be your MP, your employer, or if you are in a position of power, someone in your organisation who you will want to start making a plan. Preferably, all three.

There is a lot of drama in comment sections and on the street about false positives and tests for The Coronavirus producing positive results for other coronaviruses, like the common cold. Usually this is understood to mean you should ignore the Government. I have some sympathy for that, obviously, but what I have come to understand about tests, however, is that they are not binary. They are not limited to producing yes and no answers – if the person using them allows for it.
Usually the people doing tests don’t allow it, for reasons that are basically economical, but actually the missing subtlety is interesting. Tests can tell you how much virus is present – the “viral load”.

There are two other interesting properties of tests:

Specificity – a test with good specificity only detects SARS-COVID-2 i.e The Coronavirus, and not older or similar bugs

Sensitivity – a test with good sensitivity will tell you if your body has ever come into contact with SARS-COVID-2 even if the “viral load” is small.

There were, apparently, some issues with specificity at one point. The major issue at the moment is that most tests are far too sensitive.

If you just caught the virus yesterday then you are probably not spreading it yet, but you’ll test “positive” because there are already a few thousand bits of virus DNA in the sample. If you had it in January then you will have stopped spreading at the very latest by mid-February but would still be testing “positive” in March, because there are still thousands of bits of virus DNA in the sample. Either of these results makes it unclear what you, as an individual, should do about it.

What researchers have learned is that SARS-COVID-2, specifically, is mainly transmissible in a period between 1 to 6 days when your body is producing trillions of copies of the virus. If the test was able to tell you that you had trillions of copies in your body, then you would know what to do. You would know that tomorrow will be a bad day for you, and that going to a party today would make next Saturday a bad day for everyone at the party.

The brevity of this period is easy to overstate. It might be up to six days long, but by day two of that period you will probably want to be in bed anyway, and limited transmission will actually happen after that.

Testing as a means to stop transmission has only one day to do anything useful.

→ Continue reading: Test results are not binary. Good tests are not all the same