Author Topic: Tweets, Pictures, And Videos That Are Coronavirus Related (Read 573526 times)

S209 · « **Reply #2080 on:** December 22, 2020, 08:39:21 PM »

Quote from: farmbochur on December 22, 2020, 07:18:03 PM

When it comes to medical testing, high accuracy should not be conflated with predictive value.

A positive result on a test that's 90%+ accurate may imply only a ~50% chance of actual infection. Given a positive test result, the likelihood of actual infection is highly dependent on the prevalence of the disease.

Or you can make it simpler: Where there is zero prevalence (say in New Zealand) even a positive test that’s 99.5% accurate implies a 0% chance of infection.

skyguy918 · « **Reply #2081 on:** December 23, 2020, 12:51:11 AM »

Quote from: S209 on December 22, 2020, 08:39:21 PM

Or you can make it simpler: Where there is zero prevalence (say in New Zealand) even a positive test that’s 99.5% accurate implies a 0% chance of infection.

And how is zero prevalence determined?

PlatinumGuy · « **Reply #2082 on:** December 23, 2020, 12:54:03 AM »

Quote from: skyguy918 on December 23, 2020, 12:51:11 AM

And how is zero prevalence determined?

The PCR tests don't have false positives. There can be an error in the machine but if it's truly positive, there's covid RNA there. You can also confirm it with lung CT.

skyguy918 · « **Reply #2083 on:** December 23, 2020, 12:59:04 AM »

Quote from: PlatinumGuy on December 23, 2020, 12:54:03 AM

The PCR tests don't have false positives. There can be an error in the machine but if it's truly positive, there's covid RNA there. You can also confirm it with lung CT.

Read my question in context. You're response is more of a contradiction to the post I replied to than to mine.

PlatinumGuy · « **Reply #2084 on:** December 23, 2020, 01:00:29 AM »

Quote from: skyguy918 on December 23, 2020, 12:59:04 AM

Read my question in context. You're response is more of a contradiction to the post I replied to than to mine.

I'm aware, but that post can be explained with antigen test that do have false positives.

skyguy918 · « **Reply #2085 on:** December 23, 2020, 08:14:24 AM »

Quote from: PlatinumGuy on December 23, 2020, 01:00:29 AM

I'm aware, but that post can be explained with antigen test that do have false positives.

You'd have to PCR the whole population at once to establish that scenario of zero prevalence firmly enough to then say with absolute certainty that a positive on a rapid means nothing.