4 ms·
I don't think the quote you provided says anything about false positives; at least, I really hope it doesn't. A 10% false positive rate would be ridiculous. I
by zenexer 7y ago
I don't think the quote you provided says anything about false positives; at least, I really hope it doesn't. A 10% false positive rate would be ridiculous. If 0.1% of the population has SARS-CoV-2, a test has a 10% false positive rate, and you get a positive result on that test, you only have about a 1% chance of actually being infected.
That's a little unintuitive, but here's how it breaks down, without taking into account false negatives:
* 10% of the population will receive a false positive
* 0.1% of the population will receive a true positive
Which category are you more likely to fall into--10% of the population or 0.1%? You're 100 times more likely to fall into the false positive category.
- Moru 7y agoCan't find the article now but there was two fast tests that got forbidden in Sweden yesterday because they miss a large percentage of the cases making it meaningless in best case or dangerous to others in worst case. You think you don't have corona so you go to work anyway even though you feel a bit sick. A false negative is worse than a false positive.
- zenexer 7y agoA single false negative is worse than a single false positive, but false positives affect far more people with equivalent percentages. They’re both problematic, but in different ways. Our medical infrastructure is already strained. If you tell 100x the number of actually infected people that they are infected, everything will grind time a halt. It’s important to find a balance. A single false negative could be disastrous—but so could 1000 false positives. If the false negative and false positive rate are both 10%, and 0.1% of the population is infected, there will be 1000 false positives for every false negative. Both of those are going to cause huge (but very different) problems.
- ImaCake 7y agoNope, that is exactly what they are saying. It really is 10% false positive. They tested 128 negative blood samples and got a positive result for 12 of them. There is a frequency table at the end of their paper (0). So I guess you are right that if you sample a population of 1000 people (and only 1 actually sick) you would get 100 positive results, and 88% chance of detecting the one person who is actually sick. [0] https://rapidward.com/assets/downloads/Li_et_al-2020-Journal_of_Medical_Virology.pdf https://rapidward.com/assets/downloads/Li_et_al-2020-Journal...