Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

In 4 US state prisons, 3,300 inmates test positive, 96% without symptoms

reuters.com

Post reply on HN
Print view View original

41–50 of 649 posts

Re: In 4 US state prisons, 3,300 inmates test positive, 96% without symptoms

#41

Post by marcell »

DanBC wrote: post #22
marcell wrote: post #3

I've been tracking the antibody study results in a spreadsheet, and they are suggesting a 10-20x undercount of cases in the official "confirmed" numbers. You can see the data I've collected here: https://docs.google.com/spreadsheets/d/16onEUBWIV5IqN1RCvTla...

What numbers are you using for sensitivity and specificity?

It depends on the study, I've been using their reported confidence intervals. The two California studies (Santa Clara and Los Angeles) used the same kit, which has 2 false positives of 371 tests, and 10-20% false negative rate.

Re: In 4 US state prisons, 3,300 inmates test positive, 96% without symptoms

#43

Post by MrPatan »

Does anybody know the false positive rate of the test?

Could this be an artifact of that plus low actual incidence like:

- False positive rate of 10%

- 100 tests

- 1 true positives. All with symptoms

- 9 false positives. All without symptoms (duh!)

Headline: "90% asymptomatic!"

Truth: "Shitty test procedure!"

Yes, yes, I know you know this. Do you think people doing tests and writing headlines know this?

What is the false positive rate?

Re: In 4 US state prisons, 3,300 inmates test positive, 96% without symptoms

#44

Post by ouid »

alexbanks wrote: post #2

Before the deluge of "But wait two weeks" comments, I just want to ask at what point we accept that the potentially of totally asymptomatic cases is insanely high, far higher than anyone thought?

You have to account for events like the Diamond Princess, if you want to run with that story.

Re: In 4 US state prisons, 3,300 inmates test positive, 96% without symptoms

#45

Post by taeric »

vmception wrote: post #32
taeric wrote: post #4

Earlier quoted context omitted.

I also want to add that this doesn't make the virus less deadly. It just gives it a very population varied IFR. It appears far more deadly for some populations. Near harmless for others. But no clear way to confine it to the harmless crowd.

> But no clear way to confine it to the harmless crowd. That won't be what the public policy decisions will be based around if the asymptomatic and already exposed rate is so high. We don't know enough to make that decision yet, but testing more broadly was the first step. Remaining steps: - Do they get sick in two weeks? - Skip testing for current sickness, and test for antibodies instead. - Get a better antibody te…

Not sure I follow. I confess I'm biased in thinking we could have locked down nursing homes and shaved a large chunk of deaths. Catch is, we should have done that in January.

Re: In 4 US state prisons, 3,300 inmates test positive, 96% without symptoms

#46

Post by SpicyLemonZest »

DanBC wrote: post #27
nodesocket wrote: post #15

Isn't this a bit of promising news, means herd immunity is probably much deeper and larger than we are currently aware.

WHO says we can't assume previous infection means future immunity. https://www.who.int/news-room/commentaries/detail/immunity-p...

The WHO says that we can't assume someone is 100% guaranteed to be immune when they have antibodies. That's true, and it's an important flaw in the idea of immunity passports that they're tackling - you couldn't send people with immunity passports into quarantined nursing homes or allow them to attend potential superspreading events.

Herd immunity doesn't require perfect immunity or a 100% guarantee of it. No reasonable expert doubts that herd immunity is possible, although some argue it's too costly.

Re: In 4 US state prisons, 3,300 inmates test positive, 96% without symptoms

#47

Post by FabHK »

alexbanks wrote: post #2

Before the deluge of "But wait two weeks" comments, I just want to ask at what point we accept that the potentially of totally asymptomatic cases is insanely high, far higher than anyone thought?

It’s not like this is the first population that has been nearly exhaustively tested. There was the cruise ship, there was the navy ship, etc. Those have shown between This certainly adds another data point, but I wouldn’t throw conventional wisdom out the window yet.

Re: In 4 US state prisons, 3,300 inmates test positive, 96% without symptoms

#48

Post by IAmEveryone »

Plugging the 96% into New York numbers results in 155,000 / 4 * 100 = 3,875,000 infected people in NYC, or about half the population.

This is assuming that tests in NYC currently include every infected and symptomatic person. Considering the official advice for people with mild symptoms is to stay home (and not seek a test), that assumption is ... optimistic.

It also wouldn't fit with the anti-body tests that have been done in NYC that showed figures closer to 15-20%.

So I'd expect about 1/2 to 2/3 of these 96% to develop symptoms within the next week.

The other possibility is the prison population not being representative of the general population. That's probably true in terms of fatality rates, because they are younger. I'm not entirely sure if that age imbalance is just as strong for any symptoms as it is for risk of hospitalisation and death.

Re: In 4 US state prisons, 3,300 inmates test positive, 96% without symptoms

#49

Post by nodesocket »

DanBC wrote: post #27
nodesocket wrote: post #15

Isn't this a bit of promising news, means herd immunity is probably much deeper and larger than we are currently aware.

WHO says we can't assume previous infection means future immunity. https://www.who.int/news-room/commentaries/detail/immunity-p...

Not sure why I am getting downvoted, I guess covid-19 shaming is a thing. From what I've read people who appear to get it again is actually just a relapse of the original infection[1].

[1] https://thehill.com/changing-america/well-being/prevention-c...

Re: In 4 US state prisons, 3,300 inmates test positive, 96% without symptoms

#50

Post by marcell »

cies wrote: post #30
marcell wrote: post #3

I've been tracking the antibody study results in a spreadsheet, and they are suggesting a 10-20x undercount of cases in the official "confirmed" numbers. You can see the data I've collected here: https://docs.google.com/spreadsheets/d/16onEUBWIV5IqN1RCvTla...

I did similar calculations, and found the institutions in charge give us very unreliable data. The term "corona case" is very, very, ambiguous and cannot the understood as such without a detailed explanation on how the counting was done. Thanks for sharing. I found the peek in all-case mortality also very interesting, because that way counting is much more unambiguous: dead is dead. They showed a clear diversion from…

Yea, I agree that the case counts are very different between regions. I think the more interesting column is the IFR estimate based on the antibody study results, since dead is dead as you say.

It is interesting though that the median undercount is converging to ~10-20x. Perhaps the protocols across regions are similar enough that the confirmed case counts are somewhat comparable.

Post reply on HN