Menu
Sign In Search Podcasts Charts People & Topics Add Podcast API Pricing

Vinay Prasad

👤 Person
142 total appearances

Appearances Over Time

Podcast Appearances

Talk Evidence
Talk Evidence covid-19 update - lack of testing transparency, how to give good debate

You've put me on the spot, haven't you?

Talk Evidence
Talk Evidence covid-19 update - lack of testing transparency, how to give good debate

So there are these tests, aren't they?

Talk Evidence
Talk Evidence covid-19 update - lack of testing transparency, how to give good debate

They're called lateral flow immunoassay tests.

Talk Evidence
Talk Evidence covid-19 update - lack of testing transparency, how to give good debate

What's good about them is they're easy to use and they're quick.

Talk Evidence
Talk Evidence covid-19 update - lack of testing transparency, how to give good debate

And

Talk Evidence
Talk Evidence covid-19 update - lack of testing transparency, how to give good debate

Just to say this is a conflict of mine because this work was done in Oxford where I'm based, but I have had nothing to do with it whatsoever.

Talk Evidence
Talk Evidence covid-19 update - lack of testing transparency, how to give good debate

So they looked at a number of different tests and tested them for SARS-CoV-2, IgM and IgG.

Talk Evidence
Talk Evidence covid-19 update - lack of testing transparency, how to give good debate

And I think they had about eight tests in total.

Talk Evidence
Talk Evidence covid-19 update - lack of testing transparency, how to give good debate

And then they looked at them in a small, compared them to the gold standard.

Talk Evidence
Talk Evidence covid-19 update - lack of testing transparency, how to give good debate

And what they came out with was the results.

Talk Evidence
Talk Evidence covid-19 update - lack of testing transparency, how to give good debate

And they have two results.

Talk Evidence
Talk Evidence covid-19 update - lack of testing transparency, how to give good debate

One was the sensitivity of LFII devices, that's what they're short for, ranged from 55 to 70%.

Talk Evidence
Talk Evidence covid-19 update - lack of testing transparency, how to give good debate

And the specificity was 95 to 100%, depending on the test.

Talk Evidence
Talk Evidence covid-19 update - lack of testing transparency, how to give good debate

OK, well, let's try and work this out.

Talk Evidence
Talk Evidence covid-19 update - lack of testing transparency, how to give good debate

So let's take the best case scenario.

Talk Evidence
Talk Evidence covid-19 update - lack of testing transparency, how to give good debate

We assume the sensitivity is 70% and the specificity is 98%, OK?

Talk Evidence
Talk Evidence covid-19 update - lack of testing transparency, how to give good debate

Now, we're starting to use words and I'll try and talk you through them because everybody's totally forgot what they mean.

Talk Evidence
Talk Evidence covid-19 update - lack of testing transparency, how to give good debate

And that's one of the problems in this language school.

Talk Evidence
Talk Evidence covid-19 update - lack of testing transparency, how to give good debate

It's like, oh, what do they mean?

Talk Evidence
Talk Evidence covid-19 update - lack of testing transparency, how to give good debate

But let's take first a population with a 5% serial prevalence.