Hacker News Reader

Comment by glangdale | Hacker News Reader

> Luckily, Bay Area and LA residents travel to China a lot, so we already had substantial herd immunity by February 1.

Ummn, what? This sounds like an entirely made up factoid, and certainly isn't supported by the dashboard. The estimates of how many people in a population have to have had COVID-19 are very high. Even if you buy the Stanford study, which is itself controversial, the estimated numbers for how many have been exposed to the virus are still an order of magnitude below what you'd need for herd immunity.

View on HN

It’s not just the Stanford study: antibody studies everywhere, from New York to Sweden to Italy are showing similar results.

The “criticism” on those studies are more hypothetical and not universal. It’s like “this study could be wrong because X”, and “that study could be wrong because Y”, but there are more than enough now to draw statistical conclusions.

The difference between 20% (NYC antibody positives) and herd immunity estimates (75%) is not an order of magnitude.

No, we don’t have herd immunity, but enough that the rate of infection is slowing.

And actually, at this rate we’ll get to herd immunity before a vaccine is found. Maybe the lockdown breakers are ironically the ones that (1) saves the economy (2) ends the corona pandemic.

> but enough that the rate of infection is slowing.

You don’t think that has something to do with the fact we drove our economy off a cliff by asking (or forcing) people to stay home?

I'm not sure the criticism is just "hypothetical." My roommate just received an antibody test and was told by the provider that the error bars are so wide on positive results that it's basically only useful to confirm that you haven't had it. Maybe they're just hedging, and sure that's anecdotal, but I think we need more evidence before drawing a strong conclusion. I hope you're right.

From what I've read, the false positive rate on most of the antibody tests available in the US is high enough that you'd be just as well served by flipping a coin.

OK right. Let's look at that NY data. If you figure you need 75% to get herd immunity, and we have 20% in NY, we need 3.35x the current number of infections to get herd immunity. Of course there is also overshoot but let's forget about that.

24,708 people have died in NY so far, but we'd need 82775 to die to get herd immunity. That's about 1% of NYC's population. Of course, NYC's population is a lot less since so many people left the city as this was getting started, and it also takes a while for many hospitalized people to lose their fight with this disease, so the 24,708 is an undercount. It also undercounts excess mortality.

Still, if we neglect all of that, you're looking at 3 million Americans dead in order for your strategy to work.

After we sacrifice the 3 million, we don't even know what the immunity will look like. Our experiences with other coronaviruses are not all that promising. People may be reinfected within months.

I don't know about results beyond the Stanford study, but they were talking about 4-5% as a highest estimate of those who had had it, which is indeed an order of magnitude off.

The NYC study is interesting; of course, it's a random sample of people at grocery stores and community centers i.e. "those who are out and about in a pandemic", not a truly random sample. But it's intriguing nonetheless.

Still, I'm not sure what the course of action that it supposedly suggests is: make everywhere in the USA 3x as infected as NYC at its peak? That's 40K more dead there and I doubt that the extrapolation to the rest of the country would be fun.

> No, we don’t have herd immunity, but enough that the rate of infection is slowing.

The rate of infection is slowing because of shelter-in-place orders. Suggesting we're to the point of developing herd immunity is completely baseless and improbable.

Reply on news.ycombinator.com