Hacker News Reader

>> ...the CDC models project that 70 MILLION Americans have already gotten sick with symptomatic Covid, but decided to just write it off as a bad cold and never got tested...

> Why is this hard to believe?

I don't agree with the article more broadly, but it does make the point that 70 million prior infections isn't consistent with hospitalization numbers over the summer vs now. The only way I can reconcile those numbers is by adding two assumptions: first, that severity depends on initial viral dose; second, various social and weather patterns associated with summer drive lower initial viral doses. I don't think these are implausible assumptions--in fact, I'd guess they're true--but I would need significantly more evidence to make me think they account for many tens of millions of missed COVID cases.

There is really no reliable evidence that symptom severity is correlated with initial viral dose. It's a plausible hypothesis but remains unproven.

There's another assumption you can add - that some people are more susceptible than others.

A good portion of hospitalizations and deaths have been the elderly via nursing homes. If those have become safer, it'd make sense that we can have more cases without heavily increasing hospitalizations

They have become safer by adapting ... and by having the most vulnerable filtered out by earlier waves.

In November my daughter came down with a slight fever and runny nose. We didn't think it was COVID, but of course wanted to make sure. It was much more difficult than I thought it would be to get tested. Many places wouldn't test a child under 10. Once we finally found a location to test the soonest they had an appointment was 24 hours later. It took several days to get the negative result as well and I had to call them twice to get the results. I'm sure many people wouldn't go through the effort.

Until we have more precise IFR numbers, my simplistic "model" is just to divide the current death count by 0.005, which suggests around 78 million have had an infectious exposure in the US so far. The CDC number doesn't feel surprising.

Unfortunately a large number of those infections (far more than recorded cases) happened last spring, so millions of people could be losing their immunity if they've managed to avoid substantial re-exposures since then.

The CDC has published their best estimates of IFR by age range.

https://www.cdc.gov/coronavirus/2019-ncov/hcp/planning-scena...

SARS-CoV-2 is very similar to other endemic human coronaviruses like HCoV-OC43. Immunity isn't a binary yes or no condition and there are other components to the human immune system beyond just antibodies. The vast majority of recovered patients will likely retain a high level of immunity for many years. As with most viral diseases a small fraction of patients will suffer reinfections but even those typically experience less severe symptoms.

https://www.technologyreview.com/2021/01/06/1015822/covid-19...

Agreed. Some people just don't care about testing.

I am friends with a doctor who tested positive. They went into quarantine with their whole family, assuming they would all get it one way or another, and didn't bother with further testing. I suspect this is the case for a lot of people.

> Covid test just satisfies your curiosity, because you're going to be isolating anyway

This might be your personal bubble speaking. I have two non-tech friends (EMT and a labourer) who contracted Covid and both of their workplaces required a positive Covid test if they wanted to take two weeks off for quarantine.

> getting a Covid test just satisfies your curiosity, because you're going to be isolating anyway.

Ehhh, people might be avoiding the test ... to avoid having to isolate. I am making this shit up, but I wouldn't be surprised.

That's a good point. Covid testing can be irrelevant to your decision-making in the wrong direction as well.

The whole thing comes off as a rant supported by gut feelings about his own questionable figures and quickly dismissing contradictory sources as political.

Author does not provide sources or analysis as to why he trusts his figures and remains skeptical of others.

> Why is this hard to believe?

That's the core of what the article is about; it's hard to believe based on comparing testing and hospitalization numbers.

> "Social behaviors" are not what slows down Rt ... the number of targets gradually diminishes

Those aren't mutually exclusive, and it is possible to change social behaviors fast enough that it happens before you reach a threshold where you get significant slowing due to reduction in available "targets".

>"Social behaviors" are not what slows down Rt when you get to 60% of the entire human population infected; the number of targets gradually diminishes, slowing the spread.

Or the virus mutates, and off you go on another round. We have a couple of corona viruses that are already endemic in the human population, what makes you think this one will just die out naturally?

I'm not talking about dying out entirely, I'm talking about an end to exponential growth.

There's a finite population on the earth, clearly it has to end at some point for that particular strain, but again, it can evolve and the new strain could then grow exponentially again. It's likely already happened.

>It's likely already happened.

Please explain? I'm not aware of any variants of SARs-CoV-2 which are not covered by existing immunity.

Is there evidence that surviving infection results in immunity forever? I don’t think we know since the disease is so new. For all we know, you can get it year after year like the flu or a cold.

AFAIK we don’t (can’t, really) know either way, but viruses it’s more closely related to (SARS 1, MERS) don’t seem to have that kind of flu behavior.

Correct, transmission Re decreases linearly with the exposed population (e.g. 50% exposed, 50% growth reduction).

The author makes the point that the assumption behind that number is that people test for Covid-19 at the same rate as they would test for the flu. I know of very few people who have tested for flu but many who have tested for Covid-19.