False Positives

[BeElBeeBub]

The most interesting thing to come out of that study was that asymptomatic people did not infect others.

aside from the fact it blows the "it's all false positives!" Argument out of the water?

Whilst it's possible asymptomatic people are never infectious (bear in mind there are plenty of virus which asymptomatic people are infectious) the problem is that there is currently no way to tell between an asymptomatic person and a pre-symptomatic person.

The 300 cases in Wuhan were asymptomatic *at the time of the test* ie some might go on to develop symptoms.

The whole reason this virus is harder to stop than (say) SARS is because there is a definite phase of infectivity *before* any symptoms arise.

If all we had to to do was isolate as soon as we had a cough or fever it would be a doddle.(relatively).

There is good evidence now that the infectivity is heavily skewed. Most (60%+)  people don't infect anyone else at all.

A further 10%  only infect 1 other person.

The remaining 30% are the ones who infect more than 1 other person with a small (low single digits) number infecting alot more (dozens). The so called super spreaders


We don't know if it's something with the biology of the super spreaders (maybe they have a high viral load, or breathe alot) or simply their circumstances. A vicar who is infectious for 2.days before symptoms might infect nobody if those two days are mid week but if they're over a weekend with 3.weddings, 4 services and a choir practice ....

Still assuming the Chinese study is right and they haven't missed to many cases (the false negative rate is far higher than the false negative rate, especially for a single test) it is proof that you can beat the virus.

[Barry]

The most interesting thing to come out of that study was that asymptomatic people did not infect others.

[BeElBeeBub]

Wuhan conducted a mass screening program which tested just under 10 million residents (9.8m)

300 cases were found.

If we assume that every case is a false positive and there are no CV infections at all in Wuhan.

That means we have 300 false positives out of 10m true negatives.

Or a FPR of 0.003%

Obviously we could be suspicious of Chinese  data. Though simply saying "the Chinese lie" isn't a very strong argument.

And their tests are not going to be identical to UK tests, so there there is room there for differences. I'm particular it is likely they used a combination pool system to test multiple samples in one test which could reduce the sensitivity.

However it does show that very low positive rates are very achievable. The UK achieved a rate of less than 0.05% over the summer, again assuming all the positives were false.



https://www.nature.com/articles/s41467-020-19802-w