[Arrrlex]

Could someone help me understand the all ages excess mortality graph? Naively it looks like

a) there were way more deaths in 2018 and 2019 than were expected, and

b) despite Covid-19, the excess mortality still hasn't caught up with 2018 levels

What am I missing here?

[henrikschroder]

It's a cumulated graph. Earlier flu seasons were responsible for excess deaths in 2018 and 2019, and since flu season starts earlier in the year than pandemic deaths, by April, more people had time to die than has died in total from covid-19 in 2020 so far.

You can click the "weekly" button and you'll see that covid-19 is deadlier than earlier flus, but has been killing people for a shorter amount of time.

[mrow84]

IANAE but ... all three years show the pattern of what I presume is the flu season (broadly speaking), causing deaths (mostly in the elderly), and differing in severity/rate between the three years.

2020 is different because there is a clear step-change in death rate after week 11, and the implication/fact is that that is COVID-19.

The conclusion is that those deaths are excess deaths on top of those that we would have had as a consequence of the 2020 flu season.

[mkagenius]

> clear step-change in death rate after week 11, and the implication/fact is that that is COVID-19

How does that compare with 2018 excess death graph, clearly some other novel virus would have caused that, it broke all previous highs quickly.

[mrow84]

In 2018 excess deaths increased at a roughly consistent rate of around 7,500/week to 113,302 excess deaths in week 15 and then stopped.

In 2019 excess deaths increased at a roughly consistent rate of around 5,100/week to 56,806 excess deaths in week 11 and then stopped.

In 2020 excess deaths increased at a roughly consistent rate of around 2,500/week to 27,552 excess deaths in week 11 and then started increasing at a rate of around 15,600/week to 105,564 in week 16.

Again IANAE, but different flu strains have different mortality rates, but broadly similar behaviour in that they mostly go away after winter. The step change (caused by COVID-19) is the difference, and the new excess death rate is much higher than the 2018 rate.

[mrow84]

Posting separately so as not to distract from the point about the data: I'm sorry (for myself) to be driven to being snarky/whatever, but I don't get why this isn't completely plain from that excess deaths chart. I don't get why people are trying so hard to insist that there isn't an issue. Really, wtf?

[cma]

I don't know if it is the case, but sometimes there is something like a flu A and flu B that peak at different times.

[rjtavares]

The excess deaths in the winter are flu related. A flu season can be worse than usual for several reasons: e.g. a particular bad strain or a strain outside of the coverage of the seasonal influenza vaccine.

[soco]

So much interesting data and no idea what it means... why the visible bumps for SOME countries in December 2016 or January 2018? Why Spain has that higher mortality in winter? Why why why...

[rjtavares]

Every European country has more deaths in the winter, mainly due to the flu season (which can happen from December to February).

[ekianjo]

> a) there were way more deaths in 2018 and 2019 than were expected

Flu season every winter brings excess mortality. Also very hot summers, but there's less of a pattern.

> b) b) despite Covid-19, the excess mortality still hasn't caught up with 2018 levels

You are missing a big factor: COVID is not behind us, and this excess mortality is happening DESPITE a lockdown, which never happens for regular flu seasons.

[egman_ekki]

The numbers have been revised, apparently, now it's higher.