Hacker News Reader

Comment by ghj | Hacker News Reader

NYC is a good case study since it's one of the first place hit and is now more or less "done" with covid in the same sense used in this article (it had a peak and is now on the right tail).

NYC is at 2716 cases per 100k and 281 deaths per 100k: https://www.cdc.gov/covid-data-tracker/#cases

States with aggressive reopening policies like Florida/Louisana/Arizona etc had spikes recently but are also coming back down. They are sitting at around 2-3k cases per 100k (deaths will lag).

Even if the reason why those states "peaked" is because of herd immunity, it seems like it still requires at least 3k cases per 100k. That would translate to 300k cases for a population of 10 million but Sweden is only at 80k right now.

I highly doubt they have herd immunity.

View on HN

Your numbers are way off.

25% percent of NYC was seropositive in May. So that works out to about 25k per 100k people.

The reported cases have always been just the tip of the iceberg. And we never knew how big the underwater part was. We still down't.

Take the current situation, is the disease penetration larger than that in March/April? Most likely not, not even close, yet the reported new cases are about 5x larger.

The most recent, randomized studies of the NYC metro area finds SARS-CoV-2 antibodies in about 14% of participants when they randomly solicit people directly. Prior studies put out ads looking for participants, naturally drawing in people who "had that unexplained thing a few months back".

Even when directly appealing to people the participation rate is likely higher among people who suspect they may have had it.

We know that testing was very limited early in the outbreak, when NYC was hit, and that it caught only a fraction of the people infected. But all metrics are that there is still a huge victim base in the area.

The drop is more likely entirely behavioral.

Good to know. My 25% numbers were from an announcement by Cuomo where the report claimed that the sample collection and data processing (corrections) were statistically sound.

Even at 14% seropositivity the true number of people that had the virus is likely larger. There are reports that some (many?) people that recover do so without needed antibodies.

Working backwards, taking the fairly accepted value of 0.6% fatality rate, the 30K deaths in NYC would indicate around 5 million infections. Now the the NYC fatality rate is likely higher due to Cuomo's mind boggling order regarding nursing homes.

Regardless I think it is fair to estimate that the true number of cases in NYC is likely to fall between the reported 400k and the upper limit of 5 million. At 25% seropositivity would be around 2 million, to me sounds believable.

I don't care how big the underwater part is though. I am just using the case rate as a proxy for when states seem to hit their peak (yes, extremely handwavy).

As long at the confirmed cases is within some constant fraction of true number, it will be a decent proxy. This makes sense to me since some x% of the people who get infect will need hospitalization so they will get counted towards the statistics.

> I don't care how big the underwater part is though

You should care. The disease propagation is all about the "unseen" part.

Otherwise the numbers are not comparable at all. Back then we caught 10% of cases, today perhaps 50%

Thus you can't extrapolate from old numbers into the new.

if anything you'd need to triple the numbers or more

Case rate is unreliable due to testing issues. It's highly unlikely NYC only had that many cases with an implied fatality rate of over 10%.

Louisiana was not aggressive about reopening; we have Democratic governor

A depressing sentence

I think we really are losing sight of the before versus now comparison. NYC is an absolutely ghost town compared to how it was. Everyone stays remarkably limited in their social contact. Professional sports are a ghost of what they were, running in an incredibly limited fashion. Travel is rare.

The same is true of Sweden as well, of course. Social interaction has collapsed to very close to nil, and any article that claims that things are "normal" is lying to push an agenda.

How are sports teams a ghost of what they were? NHL playoffs are incredible right now. European soccer is wrapping up and play is great. NBA is just getting started. Most of them just canceled games or took a break, which, for example, the NHL used to do every 4 years for the Olympics. The only sport that seems to be having real trouble is the MLB, who just had a terrible plan. I admit, I have no belief the NFL has any chance due to team and personal size but that is just life. Not to mention esports which has barely slowed down after the first few months where they had to cancel in person events.

"NHL playoffs are incredible right now"

No fans. Playing in two bubble cities in Canada.

"NBA is just getting started"

No fans. Playing in one bubble city.

And I was talking about NYC. Zero NBA or NHL games are taking place in NYC.

In an average month in NYC, pre-COVID, there are dozens of enormous sports events. Hundreds of concerts. Tens of thousands of significant parties. If social interaction were on a scale from 0 to 100, and 100 was the before, right now NYC is at a less than 1. So when someone looks at NYCs current infection rate and presumes that means there is some herd immunity or the like, I don't think they're fully taking into account the reality on the ground.

I'm not sure how playing in a bubble without fans means they are a ghost. Honestly I wish fans didn't come back to live sports. Hearing the game and player noise makes everything much more enjoyable. I think most pro-sports are currently doing great besides missing ticket revenue, especially compared to a few months ago when they weren't going at all.

Anyways, I don't think measuring recovery in the number of events and parties is a very good metric. Though I agree without that there probably isn't going to be true herd immunity for this and I wish people would stop throwing around the term.

a) The context is specifically in relation to communicability

b) Further, it was about New York City!

Extremely few sports are taking place in NYC, and even if we discount that, from the overarching context those sports having a tiny, tiny fraction of the communicability potential.

Here’s a graph which speaks to exactly this topic - the R value charted versus two axis; social connectivity, and herd immunity.

Both are equally important pieces of the puzzle. They work together to get R below 1.

https://twitter.com/trvrb/status/1291860668342079490?s=21

Looks useless to me; dropping R to below (but not much below l) 1 means a super-slow decline; you might need years with such a regime to eliminate the virus. And it doesn’t actually provide herd immunity, just a slow die-off.

Which leaves most of the population vulnerable, and re-introduction a possibility.

You have to get R significantly below 1 and the whole ecosystem (for the US: the whole world) to do it.

Or, you need herd immunity compatible with the R that you can maintain indefinitely - which la likely >2 so immunity needs to be >60%.

Any other way leaves you with the possibility of reintroduction as far as I understand.

Testing has been off in FL due to the hurricane closing sites.

Reply on news.ycombinator.com