Hacker News Reader

Comment by RamRodification | Hacker News Reader

While this is clearly a second wave, "New cases" are kind of meaningless as it is very much tied to the amount of testing performed.

Here are the stats for number of deaths and also patients in ICU which gives a different picture. Still alarming but not as crazy as this "New cases"-graph might make you believe.

(two links to the same image on separate image hosting services)

https://postimg.cc/rzyCv57W https://imgbox.com/zI14ri2b

Sources: https://experience.arcgis.com/experience/09f821667ce64bf7be6... Public Health Agency of Sweden

View on HN

New cases are not meaningless. They might just not be 1:1 comparable in numeric values to the numbers a couple of months before. But if you are an an exponential raise, even pretty large factors are caught up with literally within a couple of days.

You are correct. That's why I went with "kind of meaningless". Maybe a bit too harsh by me. What I really meant was "not as meaningful as people make it out to be." (or "seem to want it to be").

The post kind of goes from straight talk about deaths to "and here are stats" which are not at all showing deaths. Comes off a little dishonest to me. And the stat on deaths being "40% above the U.S. equivalent" just seem plain wrong. See the comments on the post for that.

New cases do not say anything about how many dead you can expect, case in point Czech Republic: https://www.worldometers.info/coronavirus/country/czech-repu...

How is this a case in point? The curve of the deaths follows the curve of infections closely.

Both Sweden and New York have seen an upward trend of infections, but without a big increase in deaths as you have in the Czech Republic. Here is the data for New York: https://www.worldometers.info/coronavirus/usa/new-york/

No, since in this situation the error is growing exponentially too. There can never be any "catching up" in this case.

Deaths are a lagging indicator, in part because it takes a while to die - Herman Cain didn't die for a month - and in part because it takes a while to report them after death. (https://www.cdc.gov/nchs/nvss/vsrr/covid19/index.htm: "Death counts are delayed and may differ from other published sources...")

You also need to look at the positivity rate, which answers the question are cases going up because of increased testing, or despite of?

https://ourworldindata.org/coronavirus/country/sweden?countr...

Indeed, and large-scale testing was just getting started during the first peak. IIRC the bump around June in the "new cases" data corresponds to a scale-up of testing capacity.

Reply on news.ycombinator.com