Greenwood2
Member of DD Central
Posts: 4,251
Likes: 2,694
|
Post by Greenwood2 on Jul 18, 2020 10:24:48 GMT
From the graph above it looks like that may be starting to happen, the last couple of points are slightly below last year for the first time since this all began (may all change again though). Due to general life improvements, even if Covid did no happen you would expect the 2020 line to be lower than 2019 (ignoring statistical noise). If your excess deaths later in the year was so negative that it offset the earlier spike then yes you could say Covid just accelerated deaths rather than killed a significant extra amount of people. That is a theory that some hold that might prove to be the case. I think the yearly variations are more to do with how bad the flu season was, although you might expect an overall downward trend over several years. On the other hand I seem to remember life expectancy in the UK is not going up much if at all currently.
|
|
IFISAcava
Member of DD Central
Posts: 3,665
Likes: 2,989
|
Post by IFISAcava on Jul 18, 2020 10:46:21 GMT
From the graph above it looks like that may be starting to happen, the last couple of points are slightly below last year for the first time since this all began (may all change again though). Due to general life improvements, even if Covid did no happen you would expect the 2020 line to be lower than 2019 (ignoring statistical noise). If your excess deaths later in the year was so negative that it offset the earlier spike then yes you could say Covid just accelerated deaths rather than killed a significant extra amount of people. That is a theory that some hold that might prove to be the case. But it isn't either/or. It's likely to be both. The below trend deaths is supportive (but not definitive proof) of some element of accelerated death. In other words, people who would have died now are not available for dying now because they died earlier in the year with COVID-19. A longer time series will help, and there is delay and noise in the system as we are seeing.
|
|
slippery
Member of DD Central
Posts: 83
Likes: 61
|
Post by slippery on Jul 18, 2020 10:51:58 GMT
"What was the cause of all of those "non"-Covid 19 deaths in March, April, May?"
Assuming the stats are right, one contribution may have been the necessity to suddenly cut off all the extra care normally provided by partners/close relatives visiting nursing homes. Dealing with PPE etc will presumably have meant that the most vulnerable residents had even less of the carers' time. Some nurses and carers are wonderful but others are not. Especially behind closed doors.
|
|
|
Post by bracknellboy on Jul 18, 2020 11:23:35 GMT
From the graph above it looks like that may be starting to happen, the last couple of points are slightly below last year for the first time since this all began (may all change again though). Due to general life improvements, even if Covid did no happen you would expect the 2020 line to be lower than 2019 (ignoring statistical noise).uhh, no. Improvements in life expectancy between 2020 and 2019 cohort would be so small as TO BE statistical noise. There is way way more variation between years in even a 5 or 10 year comparison driven by e.g. severity of seasonal flu, probably even weather (think hot summers) than would ever be seen in extended average life expectancy.
|
|
|
Post by bracknellboy on Jul 18, 2020 11:31:18 GMT
From the graph above it looks like that may be starting to happen, the last couple of points are slightly below last year for the first time since this all began (may all change again though). And interestingly excess deaths are below average while 100 people per day are still dying from/with COVID-19. but as a I recall, if you look at the 2019/20 winter excess deaths prior to covid becoming a thing, it was trending below the 5/10 year average already. Which could equally well be an indicator that a mild "deaths from seasonal effects" was otherwise on the cards, and that therefore 'excess deaths' against the average will itself have some undercounting effect in the mix as well. Clearly C19 will have accelerated other deaths, with a bias to those who had shorter live expectancy: we have after all managed to allow it to do an aegeon stable job on the UKs care homes.
|
|
starfished
Member of DD Central
Posts: 296
Likes: 216
|
Post by starfished on Jul 18, 2020 12:06:02 GMT
Due to general life improvements, even if Covid did no happen you would expect the 2020 line to be lower than 2019 (ignoring statistical noise).uhh, no. Improvements in life expectancy between 2020 and 2019 cohort would be so small as TO BE statistical noise. There is way way more variation between years in even a 5 or 10 year comparison driven by e.g. severity of seasonal flu, probably even weather (think hot summers) than would ever be seen in extended average life expectancy. I think it depends on how you think about it. If something is a key part of the data then I take it out so I tend to think of it like this year's mortality being a function of, last year mortality, longevity improvements (A), seasonal flu effects (B) and then statistical noise (C). What I was trying to say is that I expect "A" to be negative if "C" is removed. However, I do agree "B" can sometimes dominate from year to year. My understanding is that "A" has continued to improve in the UK, what had slowed significantly recently was the rate of that improvement.
|
|
michaelc
Member of DD Central
Posts: 4,896
Likes: 2,768
|
Post by michaelc on Jul 18, 2020 12:13:30 GMT
As dan1 has said excess deaths gives the best indication. IMO, no it doesn't. We are all going to die, the only question is when. Excess deaths includes a lot, possibly a majority, of people who would have died anyway in the near future. If Covid recedes we will see excess deaths go negative. Will you then say that Covid has saved lives? I think not. The only way of assessing the severity of a threat using morbidity is to measure lifetime lost. So a new born infant would lose, say, 80 years. A seriously ill 80 year old would lose <1. I agree with your first paragraph but don't feel comfortable suggesting a new born is worth 80 times that of an 80 year old. Comes down to core belief but I've always thought a life is a life and they are all equal.
|
|
michaelc
Member of DD Central
Posts: 4,896
Likes: 2,768
|
Post by michaelc on Jul 18, 2020 12:46:23 GMT
Wow! One simple, self-explanatory chart seems to have hit a very raw nerve. What was the cause of all of those "non"-Covid 19 deaths in March, April, May? I wouldn't say a raw nerve. I found the chart extremely interesting and thought provoking thus a great catalyst to debate.
|
|
IFISAcava
Member of DD Central
Posts: 3,665
Likes: 2,989
|
Post by IFISAcava on Jul 18, 2020 14:03:29 GMT
And interestingly excess deaths are below average while 100 people per day are still dying from/with COVID-19. but as a I recall, if you look at the 2019/20 winter excess deaths prior to covid becoming a thing, it was trending below the 5/10 year average already. Which could equally well be an indicator that a mild "deaths from seasonal effects" was otherwise on the cards, and that therefore 'excess deaths' against the average will itself have some undercounting effect in the mix as well. Clearly C19 will have accelerated other deaths, with a bias to those who had shorter live expectancy: we have after all managed to allow it to do an aegeon stable job on the UKs care homes. Or that people's deaths had already been postponed and there was a catch up effect (regression to the mean) etc etc. Care homes has been in my view the biggest failure of all, given that protecting the elderly has been a key aim all along.
|
|
michaelc
Member of DD Central
Posts: 4,896
Likes: 2,768
|
Post by michaelc on Jul 18, 2020 15:59:47 GMT
Last Saturday also seemed to show a jump. I'm hoping its a Saturday blip.
|
|
|
Post by bracknellboy on Jul 18, 2020 16:59:55 GMT
Last Saturday also seemed to show a jump. I'm hoping its a Saturday blip. none of them actually died of C19, they were in fact all run over by a bus 29 days after testing +ve (attribution: Matt Hancock)
|
|
|
Post by dan1 on Jul 18, 2020 19:00:22 GMT
Last Saturday also seemed to show a jump. I'm hoping its a Saturday blip. none of them actually died of C19, they were in fact all run over by a bus 29 days after testing +ve (attribution: Matt Hancock) Typical, just typical. You wait all that time for a bloody bus and 70,000 turn up at once
|
|
duck
Member of DD Central
Posts: 2,589
Likes: 5,769
|
Post by duck on Jul 19, 2020 5:17:58 GMT
One point on dan1 's posted graph that I don't think has been mentioned, occupancy. Assuming care homes are taking in few/no new residents since Covid first struck wouldn't you expect the number of deaths be below 'normal' (something that the graph is starting to show) due to the lower occupancy? Whilst I generally hate %'s IMHO this is possibly a better measure here?
|
|
benaj
Member of DD Central
Posts: 4,884
Likes: 1,603
|
Post by benaj on Jul 19, 2020 7:41:28 GMT
Sir Ian Diamond said on Sky News this morning, “stick with ONS data.”
|
|
Greenwood2
Member of DD Central
Posts: 4,251
Likes: 2,694
|
Post by Greenwood2 on Jul 19, 2020 11:28:22 GMT
One point on dan1 's posted graph that I don't think has been mentioned, occupancy. Assuming care homes are taking in few/no new residents since Covid first struck wouldn't you expect the number of deaths be below 'normal' (something that the graph is starting to show) due to the lower occupancy? Whilst I generally hate %'s IMHO this is possibly a better measure here? You could also argue that care homes had more (and possibly more unwell) residents during the pandemic as hospitals released beds for covid patients, also unfortunately one route that covid got into care homes.
|
|