Hacker News Reader

Comment by rossdavidh | Hacker News Reader

If you're looking at human cost, then yes, deaths is what matter. If you're evaluating protocols to prevent infection, then cases per capita is exactly what you care about. The likelihood of death when you get infected could be a better measure of how well the hospital can treat it, for example. But if you're evaluating how well mask-wearing impacts the spread of the virus, then looking at new cases per capita is the metric of interest.

View on HN

Cases per capita is still very misleading without accounting for number of tests though, no?

Looking at some countries that offer more detailed data, you can sometimes see the cases double, while hospitalizations and deaths grow at a much lower rate (if at all).

Compare for example the UK:

- This chart for the daily new cases shows a dramatic second wave that even exceeds the first one: https://www.worldometers.info/coronavirus/country/uk/

- This chart for the positive rate on the other hand shows a modest increase to ~3% over the past few weeks, far below the 30% in April: https://ourworldindata.org/coronavirus-testing#the-positive-...

- And if you look at the number of tests you'll see it has more than tripled since April, hence the increase in "cases": https://ourworldindata.org/coronavirus-testing#how-many-test...

> If you're looking at human cost, then yes, deaths is what matter. If you're evaluating protocols to prevent infection, then cases per capita is exactly what you care about.

But there is no reason to care whether infection is prevented or not, unless infection carries some sort of cost. That metric has no value on its own.

I'm genuinely confused at what you're trying to get at here. I mean, in the abstract, sure, if an infection had no "cost" we wouldn't care about how widespread it was. But it's fairly well-established by now that this specific infection carries not only a small but obviously real risk of death, but a risk of lingering, debilitating effects that go far beyond the respiratory system.

It's a metric of interest, but it's uninteresting if we can't measure it properly.

To get a good estimate we'd have to randomly test the population, instead of measuring people with symptoms. Since a lot of people are asymptomatic or don't get a test since they don't feel too sick. Picking people at random would give us a good overview of active cases.

In March in Germany they were reporting a doubling of cases on TV every day, but they failed to mention that they had administered almost twice the tests as well. So basically the number we were measuring was the number of tests we could administer, which was naturally rising as testing capacities expanded, not the number of Corona cases.

Deaths just seem way more reliable a measure.

Well you are correct about the reliability factor. In theory, we could have randomly tested, but for reasons that escape me this was done almost nowhere. Deaths is more reliable.

However, the metric that is actually at least attempting to measure the underlying reality of interest, here, is not deaths, but cases, that's my only point. But I totally concede that in some cases deaths may be the less bad metric for cases, than confirmed cases.

Sweden actually does it. It's PCR tests to track the number and change in ongoing infections in the whole population.

0.9% end of April

0.3% end of May

Zero of the tested were infected end of August

End of September is being published next week

https://www.folkhalsomyndigheten.se/nyheter-och-press/nyhets...

On Tuesday, at the press conference, they said that it was 0 in September as well. The more detailed report is published next week.

If you are looking at human cost, it is not just death that matters. The people who have long term consequences matter too. It is that if you dont die you walk away scott free.

It is that there is range of consequences from absolutely nothing to death through "sick for months" and through "maybe lifelong consequences but we dont know yet".

Do you have evidence that permanent damage occurs in significant numbers?

It's studied, of course, but, yes.

Even if 10 % of patients end up with permanently reduced lung function it's a lot because the disease is so infectious, given the right conditions, and it isn't going anywhere soon, at least not this winter.

Here's an overview: https://www.nature.com/articles/d41586-020-02598-6

I think you have misread the article. It says that less than 10% will even have intermediate-term damage. There is no evidence that any significant numbers of people will have long-term, let alone permanent, damage.

Reply on news.ycombinator.com