Latest EU Excess Mortality Stats: Covid-19 Visible in Some Countries, But No Worse Than a Flu, Sweden Fine

Euromomo has updated deaths stats out. They give the impression that they go up to the end of week 13 (March 27), although in fact for a lot of countries the stats only go for half the week, and the rest of the week is a prediction (more on this below). It looks like we can now see some Covid-19 deaths there, but there’s still no sign of a horror movie.

Here’s your headlines: Italy, Spain, Switzerland and Belgium all have definite recent upturns. But none of them massive. All within normal range. Italy is the worst, but it’s on about the same par as the winter of 2016-17. You know, the winter when we had that other World War 3 and we had to shut down the whole world… no, hang on, nobody even noticed that.

As the graphs may have changed by the time you read this post here’s a snip of some of the countries. It’s the last bit of the graph on the right-hand side that has the most recent figures:

(For those who didn’t read my last post on the euromomo figures, these graphs don’t present the ‘absolute’ numbers, but the ‘z-scores’, which are the deviations from what is normal for this country.)

France is up a bit, but everywhere else (apart from England, which we’ll talk about in a minute) is flat, or even down. That includes Wales, Scotland, NI, Germany, and most importantly … Sweden. Yes, Sweden is stubbornly showing no signs of impending doom despite their refusal to have a lockdown.

Here’s a snip of some more of them:

We can also see that the deaths are once again very predominantly occurring in the elderly. If you click on the age range figures — see the top of the country z-scores page — this is very clear. We can also see that in the pooled European data page, which is pided up into age groups — again it’s the last bit of the graph on the right that has the recent weeks’ figures:

This again tells us that this is a nasty-but-fairly-normal winter virus, and not a general once-in-a-century mass killer.

There’s one more thing to talk about, and that is the ‘delay-adjusted’ blue lines on the z-score by country page. The green lines represent the actual death figures that have come in so far (in z-score format), whereas the blue lines represent the ‘delay-adjusted figures’. The latest figures won’t be complete because some death certificates take time to do, and so the certificates for some recent deaths won’t have come in yet. Euromomo therefore makes predictions about what the ultimate figures will be for recent weeks based on the data they have from previous years that tell them how these things normally go. At least, that seems to be the general idea from what they say on their methods page (which isn’t the clearest).

However, I don’t get what is happening with England’s delay-adjusted data. The green line goes to the middle of week 13, and is well below normal. Then there’s a blue delay-adjusted line spiking upwards for the rest of the week. I don’t see how they get that from past data. Which makes me wonder whether they are actually basing that spike on what they’ve been told is in the pipeline by the British stats people. So we need to take some care in declaring that this graph shows that the British death rates have gone up recently. It doesn’t show that the official figures have gone up, although they may well do so in a few days (we’ll have a better idea of that anyway on Tuesday when the latest batch of ONS stats comes out). Italy’s upturn, however, is real, although as I said, it’s not that big.

Source: Hector Drummond