Data : Death Rates

[TheFrendo]

Here is a graph of weekly death rate for the past four years:

assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/934792/Weekly_report_mortality_W46.pdf

It demonstrates the clear seasonality in death rates. We get a low of ~8000 a week in summer. In winter the rate increases to over ~11,000 a week in a good winter to 14,000+ if we get a bad one.

The same link show that the current death rate is a whisker under 11,000 a week.

At the same time, we have about 3000 deaths a week classed as covid deaths "Deaths within 28 days of positive test by date reported".

The implication is that covid is responsible for the 3000 deaths, that these deaths are somehow extra, over and above the normal.

I do not see how that can be true, given our current low death rate.

How can we be in the middle of deadly second wave of covid when the weekly death rate is so low?

What is going on?

Because lots of accidental deaths aren’t happening, less surgeries, less illness of different varieties being spread.

We’re in an artificial situation right now and winter hasn’t even officially begun yet. It’s far to early to call shenanigans when we’re only in autumn and there are lots of preventative measures in place .

It seems that deaths from respiratory illnesses in 2020 are at their usual level. Instead of flu and pneumonia, the killer is covid. Overall deaths this year seem set to match every other year.

Lockdown is becoming harder to justify.

@Cornettoninja

Because lots of accidental deaths aren’t happening, less surgeries, less illness of different varieties being spread.

We’re in an artificial situation right now and winter hasn’t even officially begun yet. It’s far to early to call shenanigans when we’re only in autumn and there are lots of preventative measures in place .

Those causes of death can't possibly be 3000 in a week.

I don't know how many patients die as a result of surgery but the increase in cancer deaths due to stopped treatment would need to be taken against those.

I'm not suggesting any kind of shenanigans but if we're going to comment on data it's only valid if you have actual numbers

Are you reading the same graph as me OP? The impact of CV looks pretty clear to me

Big spike in deaths earlier this year, tracking above baseline for most weeks since the spike, and we're a couple of months away from the seasonal spike - the graph shows pretty clearly the trajectory is higher than normal.

It's low because we're in (sort of) lock down. If we go back to normal, deaths from covid will shoot up.

I see what titchy sees. Plus we are taking lots of measures to keep things low. Although i cant help wondering if there are less elderely and vulnerable to pass away this winter.

Of course I see the April/May spike in deaths.

The question was:

How can we be in the middle of deadly second wave of covid when the weekly death rate is so low?

The deaths attributed to covid in this second wave cannot be extra deaths. If you reduce current death rates by the ~3000 attributed to covid the death rates would drop way down below baseline to numbers not even seen in summer.

The intention of lockdown is to reduce deaths, the graph clearly shows the deaths started rising above the normal level and hopefully are peaking now again. We have a raft of restrictions to keep cases low precisely so that deaths do not soar. It doesn't matter if even 2000 of those 3000 people would have died soon of another illness (flu / etc) if they hadn't died of covid first, because exponential growth means deaths would start to reach 5000, then 10,000, 20,000... As that excess grows, so do the number of very early deaths who would not have died for many more years. That's before any consideration of overwhelming the health service because people take so much time and resource to recover, the impact of ongoing long covid health problems, higher viral loads in society making people more unwell, the risk of virus mutation making it even worse (more hosts = more mutation) etc.

The deaths attributed to covid in this second wave cannot be extra deaths. If you reduce current death rates by the ~3000 attributed to covid the death rates would drop way down below baseline to numbers not even seen in summer.

Eh? The numbers are above the baseline - by around 2000 I'd estimate from that graph - they're covid deaths. That's with lockdown, local restrictions etc. If you want to imagine what the graph would have done without restrictions look to the left of the graph a little...

One thing I noticed when looking at the chart was a line showing corrected deaths, I assume the information gets updated regularly and some of the figures may increase when adjusted. I can see clearly the surge in deaths earlier this year and the start of increase in November.

titchy,

As I understand it, that baseline curve is not an average but rather a sine curve used for modelling. It does not represent what deaths should be.

Death rates for this winter will have to rise considerably over the next month or so to match the levels seen in the winters of 2016/17 and 2017/18.

Death rates for this winter will have to rise considerably over the next month or so to match the levels seen in the winters of 2016/17 and 2017/18.

Which they undoubtedly would do if we didn't have restrictions in place.

The baseline does represent what deaths would be btw the way without seasonal epidemics/pandemics. Your argument seems to be that in 17/18 we had a much larger number of expected deaths because of H3N2 therefore we should just let covid do its thing. Given that the death rate would be significantly higher than H3N2 I'm not sure that's an argument I support.

mobile.twitter.com/ActuaryByDay/status/1329353960843567104

@Flyonawalk

It seems that deaths from respiratory illnesses in 2020 are at their usual level. Instead of flu and pneumonia, the killer is covid. Overall deaths this year seem set to match every other year.

Lockdown is becoming harder to justify.

This.

No statistically significant increased mortality in the devolved nations. With our sensible firebreaks and efficient and competent track and trace system.

@Ritasueandbobtoo9

No statistically significant increased mortality in the devolved nations. With our sensible firebreaks and efficient and competent track and trace system.

Agree that's the significant take away in the link.

There are plots of excess deaths that are helpful to look at. I have attached one of them.

What should be clear from the plot is that excess deaths are now increasing rapidly but are still far below the peak in the first wave. The bad news is that death data is always weeks delayed from case data. So we won't see much effect of the lockdown until December.

The data is at fingertips.phe.org.uk/static-reports/mortality-surveillance/excess-mortality-in-england-latest.html

Maybe this makes it clearer. It's the next week on from the graph you have.

In week 45, there were 11,812 deaths registered, against a 5 year average of 10,331, an excess of 1,481. There were 1,937 deaths involving Covid.

This is for the week ending 8th Nov, which of course was 2 weeks ago; in the week 7-14th there were 2,351 deaths within 28 days of a Covid test.

Overall deaths this year are some 58,000 above average, or 13%, so far.

It could be that one would expect that deaths should be lower than average this winter - given that earlier in the year many of the sickest and most vulnerable people died.

Why are deaths higher than modelled?/blue line? And this is with restrictions in place. What will happen if you remove restrictions? Maybe the same thing as happened earlier in the year, when people were not wearing masks, few tests being done?

You wouldn’t be getting the peak of winter deaths yet. It’s only November. Looking at the ONS stats for week 44 & 45

Week 44
2016 - 10152
2017 - 9984
2018 - 9529
2019 - 10,164
2020 - 10,887

Week 45
2016 - 10470
2017 - 10346
2018 - 10151
2019 - 10182
2020 - 11812

I’m fairly certain the number of deaths over the 5 year average is higher in week 45 than week 44. Which is going in the wrong direction and definitely justifies the lockdown.

@Flyonawalk

It seems that deaths from respiratory illnesses in 2020 are at their usual level. Instead of flu and pneumonia, the killer is covid. Overall deaths this year seem set to match every other year.

Lockdown is becoming harder to justify.

I’ve stolen this from elsewhere. You should be able to spot 2020 on it. It ends in week 43, so I suspect we might need to start resurrecting people in order to make the total deaths this year look anything like normal.

Although if absolutely nobody dies of anything between now and new year it might not look too bad.

RafaIsTheKingOfClay,

Interesting graphs, thanks.

Looking at the one on the right we can see the 2020 line split in April/May/June when the virus made its impact.

Since the middle of the year the 2020 line can be seen moving up and right in parallel to the group lines of earlier years. This shows that current death rates are similar to those in the previous years, i.e. death are are now in the normal range.

This is my point.

Deaths are now in the normal range

WITH: mask wearing, hand sanitizer everywhere, lots of movement restricted (even if lots of people claim it doesn't feel like a lockdown, this time last year I was in cosy rammed pubs, in hurdles of people watching light switch ons, parades, fireworks); lots working from home; no hugs or handshakes on meeting people; very very different attitudes and procedure when meeting older relatives; schools operating so differently with frequent hand washing; people staying off work and school if they have any cold/ flu/ covid like symptoms.

Isn't that the whole point of the restrictions and lockdown: to try to keep deaths in their normal range and not overwhelm the NHS?