Hacker News new | ask | show | jobs
by ArkanExplorer 1912 days ago
These weekly death stats are not useful because if no one dies in week 1, but everyone dies instead in week 2 (plus those who would have died anyway in week 2), it will show week 2 as being catastrophic, and everyone will just ignore week 1.

We need the raw annual data which basically removes this noise.

1 comments

If you look at the chart the death in 2020/2021 are well outside the extremal values for the previous two decades, and the "mass" contained there is much larger than the cumulative discrepancies for earlier in 2020. At that level of signal I don't think the weekly data can be dismissed as not useful.
I looked up Slovakia data specifically.

Deaths: 2020: 59,000 [1] 2019: 52,234 [2] 2018: 53,914 [2]

So a 9.4% increase from 2018 to 2020.

By contrast, Sweden had a 5.7% increase. [3]

Slovakia has lockdown and curfew.

Isn't the conclusion here that lockdown in Europe is failure, when with only voluntary measures Sweden was able to achieve a lower death rate?

[1]https://spectator.sme.sk/c/22594536/in-2020-slovakia-suffere... [2]https://countryeconomy.com/demography/mortality/slovakia [3]https://www.statista.com/statistics/525353/sweden-number-of-...

It seems implausible that lockdowns in Slovakia have achieved anything - by contrast lockdown may be contributing to deaths since people are not seeking regular medical care:

https://news.sky.com/story/coronavirus-lockdown-may-have-ind...

Well this is a different topic - the issue I was addressing was the utility of weekly data presented by ciceryadam, which you were questioning.

As to you new point - you are drawing your conclusions from disparate data sources, whilst ignoring confounding factors (as pointed out by ljf). The data on wikipedia give a less dramatic picture, and after adjusting for population things look even less dramatic (10.0 -> 10.8 per 1000 for Slovakia and 9.0 -> 9.5 for Sweden). I don't think you can draw serious conclusions from such limited data.

What else do you know about Sweden that could have skewed their figures? Great healthcare? Strong public support of the measures that were put in place? Trust in govt? Healthy community to start with?

To say that Sweden and Swedes took no steps to protect themselves and their communities is baseless - I assume you do not live there? I have friends in Sweden, and the lives they are living are very very different to the friends in Australia and NZ.

Many deaths in care homes early in the pandemic, that's what skewed their figures. If you took number for the second wave only they would score way better. Lithuania already surpassed Sweden by number of cases per capita, and is really close to surpass it by number of deaths per capita. Sweden never had lockdowns. Lithuania has been living in strict lockdowns for 6 or 7 months, I am not even sure.
If you check the graph in my first comment, you can see that the 9.4% increase happened onward from week 42, so it was the second wave that hit the country hard. By that time, the country gave up on contact tracing. Plus in the second wave , the lockdown was enforced only for people not tested by mass antigen testing.

You can compare Denmark and Slovakia (similar population size) for 2 different approaches to the 2nd wave, with 2 completely different outcomes. Denmark as a whole has as many people in ICUs as Slovakia has in one hospital.

Exactly, if you have a surplus of dead people for 15 weeks straight, that means something.