Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

This is an antibody test which will only be positive days after the onset of symptoms.

This isn't really what your Twitter link says. That links says: "Both IgM & IgG (those are two different kinds of antibodies, with IgA being a third) were low or undetectable at day 0, but increased by day 5 in nearly all patients (N=16)"

-- Reports strongly indicate there are many infected people who are asymptomatic after five days.

Which is to say, this test could be extremely useful if applied widely and systematically to many people; food service workers, health care workers and so-forth.

Currently the virus test has a week turn around time. So both tests effectively find people with a week's exposure.



Turn around time is mostly due to shipping samples to outside labs, in addition to outside labs lacking automation.


But that doesn't change the current situation.

Even you could change that, just cost would make this approach very useful.


Does "N=16" mean they've approved this test after such limited testing?

I think my high school stats class told me not to trust studies where N < 30.


I think the acceptable value for N depends on the size of the effect being studied.

For example if it’s a massive effect, maybe 10 people is sufficient, but if it’s tiny enough to get swallowed up in statistical noise until you have 1000, then you need N >= 1000


I would be told off if I coded with magic global variables and is why I find math and bio frustrating. Care to explain the purpose of N?


Its the number of people they used to test.


Is this test able to detect that someone is infected before they start to infect others?


No. This test tells us that you had covid-19 in the past and recovered from it.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: