Hacker News Reader

Comment by alkonaut | Hacker News Reader

> Still, I hear from a lot of people that they indeed show (mild) symptoms, but are not tested - even upon request. How can that work?

Why would you test everyone with mild symptoms? 90% or more of those people are going to have the flu or a common cold. And they’d use a test that can be used on someone where it makes a difference: someone admitted to hospital, or someone working at a hospital.

Everyone else with symptoms can just act as if it is covid and isolate for 14 days.

When you have exponential growing, testing capability can never keep up.

View on HN

Given the long incubation period, as well as the fact that many people who are carriers are completely asymptomatic, blanket testing has been tentatively shown to be an effective way to halt the spread of the virus. See: the whole of South Korea's approach, as well as https://www.independent.co.uk/news/world/europe/coronavirus-...

My point wasn’t that fewer than possible should be tested, my point was that given the constraints (staff, equipment, time) it’s best to test people that it makes a difference for, such as those who must be in contact with people - in healthcare.

If you could test everyone with symptoms (or everyone full stop) then go for it. That’s not possible though.

> That’s not possible though

With less foot-dragging it can be. Germany tests 20,000 people a day, Korea tests 15,000 a day, the US (4-6x more populous) only test a few thousand...

The tests have an approx 20% false negative rate.

You test someone who has symptoms. The test returns negative. What do you tell them to do, given that you know the false negative rate?

This doctor says it better than me.

https://twitter.com/keirshiels/status/1241042696560619523?s=...

You tell them the same thing regardless of the outcome of the test.

The 80% of the time it returns positive, you tell the epidemiologist, and they prevent a few further infections by investigating close personal contacts.

>blanket testing has been tentatively shown to be an effective way to halt the spread of the virus.

Testing does not halt the virus. A test is just an indicator to see if you have it or not. Even if you could test for Corona as easily as you can for drunk driving, it'd still do us no good if you don't actually take some concrete steps to halt the virus.

If you have mild symptoms, you don't need a test. You need to stay the fuck home. No symptoms? Avoid going out if you don't need to.

I saw a paper that said 80% of positive tests in a Chinese population sample of people without symptoms were false positives. Is there some reason that wouldn't happen anywhere else?

Berlin for instance tested over 20.000 people last week alone. Less than 3% results were positive.

> Berlin for instance tested over 20.000 people last week alone.

20000 per week seems to be an absurdly low number to me.

> Less than 3% results were positive.

Given high number of VIPs that are already infected (members of the parliament, professional soccer, actors), I suspect that 0.5% to 1% of the population is already infected.

Edit to add numbers: The German parliament of size 709 had 4 member already infected last weekend. On March 19 we were at 6.

If 3% test positive and 0.5% are infected, then most of the positives are false, right?

But they test only 20000 per week(!).

Because if you're tested and it's positive you can tell everyone you came into contact that they need to isolate themselves too.

We assume you at least have temporary immunity so testing everybody with symptoms will make their anxiety less and a lot of more people could be mobile and working.

Reply on news.ycombinator.com