[abarro81]

For Norway vs Sweden total death rate changes you'd probably need to dig into the other causes of death and see where increases/decreases had occurred, and whether they were likely due to pandemic-related issues or something else entirely. I have zero inclination to do this so will leave it to someone very bored or who's being paid to do it.

[petejh]

I’m not claiming papal infallibility barrows. I know I’m a twat. But it remains that I made an effort to admit the flaws of my posts.

Pete’s theory I get in a U.K. vs Sweden argument but the comparison is between Norway and Sweden. Tom’s analysis seems fair based on ‘covid’ statistics but not total mortality which is what I was wondering about.

And I do think the guardian is a rag

You've missed my point.

As you rightly pointed out, Sweden has no overall increase in its excess deaths in 2020. I thought you were correct to point this out and ask why. Other posters on here may try to claim Sweden 'fucked up'. But their protestations fly in the face of the actual evidence - which says that relative to any other 'Swedish' year in the past 10, Sweden clearly did not fuck up because they haven't suffered any excess deaths. (if the evidence changes, I'll change my opinion).

They only fared badly relative to other Scandavian countries, who locked down harder. But the question that then needs to be asked (and which you are asking) is was locking down hard worth it in Scandanavia, if the evidence shows that Sweden's excess death rate is unaltered despite not locking down hard?

So my point / hypothesis was that maybe there are some underlying geographical and cultural factors at play for the low death rates in Scandanavian (when compared to the rest of Europe). And these factors negate the need for harder restrictions such as those applied by Norway. I.e. whose lives are the harder restrictions in Norway actually saving if Sweden isn't suffering any excess death from covid? Could there be less unrelated non-covid deaths in Norway? If so then that wouldn't be the purpose of a lockdown but just a happy side-effect of a socially and economically damaging policy.

None of the above is meant to imply I think the UK's response should be like Sweden (except for maybe being less of a nation of fat unhealthy walking fat-bergs and more a nation of healthy active people).

[A.Greenhorne]

.

[A.Greenhorne]

I don’t have one, that’s why I asked on here

I’m about to head out etc etc.

Isn’t this all a little Uriah Heep?

Collar (assertion of lowly status) and Cuffs (vocabulary and grammar) don’t match.

Edit: Obviously I’m wandering into Ad Hominem, too busy to flesh it out properly.

God knows I’m a knob. You take being a condescending wanker to new heights. Eat shit

[A.Greenhorne]

Thanks Pete, that makes sense. I did misinterpret your original point but understand it now and cheers for not being a twat about it. Hope that’s not too ‘Uriah fucking Heap’

[tomtom]

So Alan, during the first wave and lockdown South Africa had a massive drop in total deaths.  A significantly negative excess mortality.

So what do you think excess mortality is going to tell you?

What is the question you are hoping to answer?

As said earlier - each government response is in effect a mini experiment in how to handle covid. The only thing that has been shown to work is to restrict social contact - via lockdowns, persuasion, contact tracing and/or travel restrictions. Nothing else has worked (if it has - show me..).

There is a point at which a countries health service becomes over-run by covid cased. At that point mortality rises very rapidly, as you cannot treat the covid patients as well as all the other people needing help of ongoing treatment. At that point the excess mortality will rocket. Sweden may well have not reached that point. Italy did in the first wave, we did during the first wave abs are probably close to that now. Other countries (like NZ) knew that their health service could not cope with even a moderate amount of cases - so were forced into a very comprehensive lockdown (also helped by geography).

[A.Greenhorne]

Pete answered the question I was asking, cheers. The rest seems too much of a mess for me to comment on.

[AJM]

As you rightly pointed out, Sweden has no overall increase in its excess deaths in 2020. I thought you were correct to point this out and ask why. Other posters on here may try to claim Sweden 'fucked up'. But their protestations fly in the face of the actual evidence - which says that relative to any other 'Swedish' year in the past 10, Sweden clearly did not fuck up because they haven't suffered any excess deaths. (if the evidence changes, I'll change my opinion).

I had a very brief look at the paper and one thing which struck me was that the winter mortality for 19/20 versus the 5 year average looked notably lighter for Sweden than Norway.

I've no idea why, but it potentially points to a cross-subsidy within the overall excess deaths figure (depending exactly what you are defining your deaths as "excess" compared to) - the excess deaths versus the average in March and April are being offset against lighter winter mortality (and winter mortality is usually the highest but also, in the UK at least, is the most volatile year on year, and the full paper does note a light flu season pre pandemic) to get to the overall "no excess deaths" position.

Whilst they also argue mortality displacement, from deaths dipping below the line in the final few weeks, I don't personally find that enormously convincing in that it returns the 19/20 to below the 5 year average, exactly as it was at the start of their graphs, and furthermore it's seen across all age groups where the excess mortality was not.

I think my view from a brief read of the paper is that Sweden looked like it would have recorded deaths significantly less than its 5 year average due in part to a light winter if it weren't for a spike of covid deaths in the spring.

[petejh]

That's an interesting detail. And it raises an interesting moral question.
What level of sacrifice to prevent deaths does a society deem acceptable? And at what point does the sacrifice become unacceptable?

Perhaps Sweden's government could be shown to be using its 'societal sacrifice credit' wisely, given it probably knew the bigger picture of their lower than average deaths - would you want to severely restrict people's liberties if you knew covid would likely have zero upward impact on the 10-year average number of deaths statistic? That's a cynical view obvs, because the people would still die whether or not they moved the needle on a chart.
Or perhaps they could have prevented *even more* deaths than in an average year, by restricting people more.
(I don't have an answer).

If what you interpet is true AJM then it also raises the obvious argument that covid (in the specific context of Sweden's experience) has been no more deadly *for them* than 'a bad flu year', so why take extraordinary measures for one but not the other.
 
Sweden's total deaths, last 10 years including 2020:

[tomtom]

That's an interesting detail.
If what you interpet is true AJM then it also raises the obvious argument that covid (in the specific context of Sweden's experience) has been no more deadly *for them* than 'a bad flu year', so why take extraordinary measures for one but not the other.

And that’s clearly what Sweden and to an extent our government has tried to do - until it reaches....

There is a point at which a countries health service becomes over-run by covid cased. At that point mortality rises very rapidly, as you cannot treat the covid patients as well as all the other people needing help of ongoing treatment. At that point the excess mortality will rocket. Sweden may well have not reached that point. Italy did in the first wave, we did during the first wave abs are probably close to that now. Other countries (like NZ) knew that their health service could not cope with even a moderate amount of cases - so were forced into a very comprehensive lockdown (also helped by geography).

[AJM]

Perhaps Sweden's government could be shown to be using its 'societal sacrifice credit' wisely, given it probably knew the bigger picture of their lower than average deaths - would you want to severely restrict people's liberties if you knew it would have zero impact on the 10-year average number of deaths statistic? That's a cynical view obvs, because the people would still die whether or not they moved the needle on a chart.
Or perhaps they could have prevented *even more* deaths than in an average year, by restricting people more.
(I don't have an answer).

If what you interpet is true AJM then it also raises the obvious argument that covid (in the specific context of Sweden's experience) has been no more deadly *for them* than a bad flu year, so why take extraordinary measures for one but not the other.
 
Sweden's total deaths, last 10 years including 2020:

I'm an actuary and not a moral philosopher, so I'll duck that part. But a view that they "knew" that their light actions would not lead to something significantly worse than within the 10-year range, in late March, feels like we're crediting them with a lot of foresight - having watched Wuhan go into full lockdown, South Korea blow up, and having watched Lombardy and the rest of North Italy buckle, to be confident that light restrictions would do the job to that degree feels like something they can't possibly have known.

I've seen various arguments as to how much of their approach was framed by the freedoms enshrined in their constitution and the difficulty of removing those, but since I'm not a lawyer either I don't have much of an informed opinion beyond the fact I've seen it quoted.

In terms of your charts, without knowing how the underlying demographics are changing and what that represents as a rate of improvement and so on I wouldn't want to draw too much out of them - I'd prefer something like a standardised mortality rate comparison to help remove some of those potential confounding factors.

On the point about the flu season, it's worth noting that the areas under and over the trend line in their paper didn't obviously look equal - by eye Id have expected higher than average mortality overall, with more above the trend line than under - I'm only assuming they might offset quantitatively because of what others have said (what's your source for the cumulative 2020 excess deaths being zero - is it just the graph you linked? I tried to look to see if I could tie things back together but didn't find anything enormously helpful - everything I could see pointed to obvious excess deaths in the pandemic window with no obvious move beyond the statistical lower bound to compensate for it) . Which loops a bit back to the point about the underlying demographics and how they link from the rates per 100k to the overall death figures. But it doesn't necessarily invalidate your point in concept, in that *if* it looked like it sat within the bounds of winter variation for Sweden, then someone could probably have an interesting discussion about the right balance of how to react to it!

[petejh]

That's an interesting detail.
If what you interpet is true AJM then it also raises the obvious argument that covid (in the specific context of Sweden's experience) has been no more deadly *for them* than 'a bad flu year', so why take extraordinary measures for one but not the other.

And that’s clearly what Sweden and to an extent our government has tried to do - until it reaches....

There is a point at which a countries health service becomes over-run by covid cased. At that point mortality rises very rapidly, as you cannot treat the covid patients as well as all the other people needing help of ongoing treatment. At that point the excess mortality will rocket. Sweden may well have not reached that point. Italy did in the first wave, we did during the first wave abs are probably close to that now. Other countries (like NZ) knew that their health service could not cope with even a moderate amount of cases - so were forced into a very comprehensive lockdown (also helped by geography).

Of course. But you're still not really acknowledging the wider context and the moral question it raises. You're just going on about covid deaths and overwhelmed health services as if that's the be all and end all, and any country that fails to lower deaths is 'failing'. It clearly isn't that simple.
You can go on about covid deaths, or deaths from an overwhelmed health service all you like. If, as in Sweden's case those deaths don't actually raise the annual total deaths above the countries long-term average then you can at least ask a valid question: why *should* they take extraordinary preventative measures to prevent those deaths when we know those measures have negative side effects of their own? (I don't have an answer either way)
Also, if Sweden's health service hasn't been/isn't overwhelmed, then why even raise it?

And who are you or anyone else to claim they've 'failed'.

[petejh]

Perhaps Sweden's government could be shown to be using its 'societal sacrifice credit' wisely, given it probably knew the bigger picture of their lower than average deaths - would you want to severely restrict people's liberties if you knew it would have zero impact on the 10-year average number of deaths statistic? That's a cynical view obvs, because the people would still die whether or not they moved the needle on a chart.
Or perhaps they could have prevented *even more* deaths than in an average year, by restricting people more.
(I don't have an answer).

If what you interpet is true AJM then it also raises the obvious argument that covid (in the specific context of Sweden's experience) has been no more deadly *for them* than a bad flu year, so why take extraordinary measures for one but not the other.
 
Sweden's total deaths, last 10 years including 2020:

I'm an actuary and not a moral philosopher, so I'll duck that part. But a view that they "knew" that their light actions would not lead to something significantly worse than within the 10-year range, in late March, feels like we're crediting them with a lot of foresight - having watched Wuhan go into full lockdown, South Korea blow up, and having watched Lombardy and the rest of North Italy buckle, to be confident that light restrictions would do the job to that degree feels like something they can't possibly have known.

I've seen various arguments as to how much of their approach was framed by the freedoms enshrined in their constitution and the difficulty of removing those, but since I'm not a lawyer either I don't have much of an informed opinion beyond the fact I've seen it quoted.

In terms of your charts, without knowing how the underlying demographics are changing and what that represents as a rate of improvement and so on I wouldn't want to draw too much out of them - I'd prefer something like a standardised mortality rate comparison to help remove some of those potential confounding factors.

On the point about the flu season, it's worth noting that the areas under and over the trend line in their paper didn't obviously look equal - by eye Id have expected higher than average mortality overall, with more above the trend line than under - I'm only assuming they might offset quantitatively because of what others have said (what's your source for the cumulative 2020 excess deaths being zero - is it just the graph you linked? I tried to look to see if I could tie things back together but didn't find anything enormously helpful - everything I could see pointed to obvious excess deaths in the pandemic window with no obvious move beyond the statistical lower bound to compensate for it) . Which loops a bit back to the point about the underlying demographics and how they link from the rates per 100k to the overall death figures. But it doesn't necessarily invalidate your point in concept, in that *if* it looked like it sat within the bounds of winter variation for Sweden, then someone could probably have an interesting discussion about the right balance of how to react to it!

I don't think Dan's question was about what intent or assumptions the people in Sweden's government held.

It's about the numbers. They are what they are, whether or not they got lucky.

The data is from Statisca here: https://www.statista.com/statistics/525353/sweden-number-of-deaths/

If you want to look at death rates per 100,000 population as you say, the study that Dan (Greenhorne) linked to gives you that. The answer is the same - Sweden had a big spike compared to Norway, but no actual increase in annual deaths compared to their 10 year average. https://www.news-medical.net/news/20201116/Study-compares-deaths-in-Sweden-and-Norway-before-and-after-COVID-pandemic.aspx

Interesting that Sweden's 5-year mean death rate is higher than Norway's. Life is cheaper in Sweden than Norway..

Text from the study:

In this study, the researchers calculated weekly mortality rates with 95% confidence intervals per 100,000 individuals and mortality rate ratios to compare the epidemic year (July 29, 2019, to July 26, 2020) with the four preceding years (July 2015 to July 2019).

These scientists have also compared COVID-19 associated deaths and mortality rates for the weeks of the epidemic in Norway and Sweden (between March 16 and July 26, 2020). The data were obtained from the main data registries in these two countries that are close to 100% complete due to mandatory reporting.

All COVID-19 associated mortality (defined as deaths among individuals with a positive COVID-19 test up to thirty days before death) stratified by age has been retrieved from the Institute of Public Health in Norway and the Public Health Agency of Sweden.

[tomtom]

Easy Pete - I was deliberately not raising the moral questions.

I think the moral Q you’re asking is one that’s easier with hindsight - is the pain of lockdown worse than a softer approach. Yes?

There are also the moral questions of which people you choose not to treat (and ultimately to die) because your health service is over-run - if you misjudge how well your population will behave and follow the guidance.

Now fresh challenges as the balance between hard and soft options - will now all need to be recalibrated in light of a version that’s 70% more readily transmitted...

[A.Greenhorne]

I wonder if they’re vaccinating those with moderate to severe Alzheimer’s / Dementia? Certainly an interesting moral debate to be had there.
33 meditations on death is worth a read from the perspective of a retired consultant in elderly medicine. TT’s rabid paternalism is a bit nauseating

[petejh]

Easy Pete - I was deliberately not raising the moral questions.

I think the moral Q you’re asking is one that’s easier with hindsight - is the pain of lockdown worse than a softer approach. Yes?

There are also the moral questions of which people you choose not to treat (and ultimately to die) because your health service is over-run - if you misjudge how well your population will behave and follow the guidance.

Now fresh challenges as the balance between hard and soft options - will now all need to be recalibrated in light of a version that’s 70% more readily transmitted...

I don't think you can avoid the moral question TT if people are going to make assertions that Sweden 'failed'. And I was specifically referring to your and other's assertions to Dan about Sweden - whose health service has not been overwhelmed.
He wasn't talking about anywhere else.

[AJM]

Hmm, the statistica page says it's under maintenance - I did find that link but couldn't get into it. I'll try again in the morning.

I wasn't really responding to Dan, but to your "given it probably knew the bigger picture of their lower than average deaths" comment which implies to me at least the idea of taking decisions with that "headroom" in mind (although maybe I misinterpret you?) - even if you know how much "room you have to play with" in a very utilitarian sense I just don't believe anyone could have had the confidence to know they could stick within it at that point in time.

I wasn't looking for deaths per 100k - SMR is the mortality rate for a reference population (i.e. you take your own rates per 100k by age or age bucket and apply them to a standardised population mix. Removes the potential for distortion by age as well as by population size so more helpful for comparing populations and time periods). I did have a look at the paper but didn't find much of what I had hoped to see in it.

[petejh]

Can you provide a link to where the SMR is used for other studies, in this context?

[tomtom]

Easy Pete - I was deliberately not raising the moral questions.

I think the moral Q you’re asking is one that’s easier with hindsight - is the pain of lockdown worse than a softer approach. Yes?

There are also the moral questions of which people you choose not to treat (and ultimately to die) because your health service is over-run - if you misjudge how well your population will behave and follow the guidance.

Now fresh challenges as the balance between hard and soft options - will now all need to be recalibrated in light of a version that’s 70% more readily transmitted...

I don't think you can avoid the moral question TT if people are going to make assertions that Sweden 'failed'. And I was specifically referring to your and other's assertions to Dan about Sweden - whose health service has not been overwhelmed.
He wasn't talking about anywhere else.

Fair enough Pete - I wasn’t really sure what Dan was arguing about to be honest - more general points about strategies for managing the pandemic.

[Oldmanmatt]

Hmm, the statistica page says it's under maintenance - I did find that link but couldn't get into it. I'll try again in the morning.

I wasn't really responding to Dan, but to your "given it probably knew the bigger picture of their lower than average deaths" comment which implies to me at least the idea of taking decisions with that "headroom" in mind (although maybe I misinterpret you?) - even if you know how much "room you have to play with" in a very utilitarian sense I just don't believe anyone could have had the confidence to know they could stick within it at that point in time.

I wasn't looking for deaths per 100k - SMR is the mortality rate for a reference population (i.e. you take your own rates per 100k by age or age bucket and apply them to a standardised population mix. Removes the potential for distortion by age as well as by population size so more helpful for comparing populations and time periods). I did have a look at the paper but didn't find much of what I had hoped to see in it.

Aren’t we a little early into this discussion?

We know the 19/20 Flu season was relatively light, but that occurred pre Covid arrival in Europe (at least significant infection rates).
These numbers, excess deaths in particular, will need to be compared at the end of the 20/21 Flu season  (say March to March?) before we’ll have a meaningful comparison?
And then, there’s the correction to be expected in the Covid mitigation strategy, also mitigating Flu infections etc.

[Oldmanmatt]

I don’t have one, that’s why I asked on here

I’m about to head out etc etc.

Isn’t this all a little Uriah Heep?

Collar (assertion of lowly status) and Cuffs (vocabulary and grammar) don’t match.

Edit: Obviously I’m wandering into Ad Hominem, too busy to flesh it out properly.

God knows I’m a knob. You take being a condescending wanker to new heights. Eat shit

Ah shucks!

That’s sweet of you. I was worried you weren’t bothered. I really couldn’t give it my full attention, just five minutes here and there. I guess I just have a talent for it (being a wanker, I mean. Too many bloody years at sea, you have to develop certain “skills”).

Thing is (I guess you didn’t notice), I really only hand you back your own shit with a little garnish.
Even when I do have time, I don’t “do it” when you make a good point, or ask a valid question.

I just read you insults and see a juicy target. Glad to know I scored a point.
Duck next time.

[AJM]

Do you mean specifically in terms of Covid? (The "in this context" bit)

https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/articles/comparisonsofallcausemortalitybetweeneuropeancountriesandregions/januarytojune2020

I don't know if they'll do an update at some point to bring in the full year. But in terms of what I was looking for it's pretty good, from initial inspection, in that it shows standardised mortality rates, then compared to the five year average weekly and cumulatively, amongst other things, for cities and countries in western Europe (including, for the purpose of this discussion, Stockholm and Sweden).

[A.Greenhorne]

I don’t have one, that’s why I asked on here

I’m about to head out etc etc.

Isn’t this all a little Uriah Heep?

Collar (assertion of lowly status) and Cuffs (vocabulary and grammar) don’t match.

Edit: Obviously I’m wandering into Ad Hominem, too busy to flesh it out properly.

God knows I’m a knob. You take being a condescending wanker to new heights. Eat shit

Ah shucks!

That’s sweet of you. I was worried you weren’t bothered. I really couldn’t give it my full attention, just five minutes here and there. I guess I just have a talent for it (being a wanker, I mean. Too many bloody years at sea, you have to develop certain “skills”).

Thing is (I guess you didn’t notice), I really only hand you back your own shit with a little garnish.
Even when I do have time, I don’t “do it” when you make a good point, or ask a valid question.

I just read you insults and see a juicy target. Glad to know I scored a point.
Duck next time.

What do you want? I’d like you to get off my case. Please

[petejh]

Do you mean specifically in terms of Covid? (The "in this context" bit)

https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/articles/comparisonsofallcausemortalitybetweeneuropeancountriesandregions/januarytojune2020

I don't know if they'll do an update at some point to bring in the full year. But in terms of what I was looking for it's pretty good, from initial inspection, in that it shows standardised mortality rates, then compared to the five year average weekly and cumulatively, amongst other things, for cities and countries in western Europe (including, for the purpose of this discussion, Stockholm and Sweden).

Cheers.


From the text below Figure 3. -
''Throughout 2020, no significant deviations from the five-year average mortality rate were observed in Iceland, Norway, Finland, Denmark, Lithuania, Czechia, Slovakia, Bulgaria, Austria, Estonia, and Hungary.''

No mention of Sweden...

(Dubious - perhaps unfairly - of the data from a couple of those countries!)

[AJM]

Cheers.


From the text below Figure 3. -
''Throughout 2020, no significant deviations from the five-year average mortality rate were observed in Iceland, Norway, Finland, Denmark, Lithuania, Czechia, Slovakia, Bulgaria, Austria, Estonia, and Hungary.''

No mention of Sweden...

(Dubious - perhaps unfairly - of the data from a couple of those countries!)

I thought figure 5, the cumulative change in mortality, was quite useful too, in that it helps to prove/disprove some of my speculations above about whether the light winter offsets the covid deaths (answer seems to be - not over age 65), and also shows that "light winter" pattern of lower than baseline cumulative mortality from Jan to March across the Nordics and indeed much of europe and then a sharp divergence in outcomes after that for over 65s in particular. Also maybe helps explain why Sweden's aggregate excess deaths aren't that high, in that the under 65 year looks to have been fairly light for mortality overall, driven mostly by the light winter with a small offset for covid spike in April.

You can see some of the same breakdowns or presentations for the UK in the link below if it's of any interest.

https://www.actuaries.org.uk/learn-and-develop/continuous-mortality-investigation/other-cmi-outputs/mortality-monitor