Coronavirus - Data Analysis Thread
Discussion
1. This shows tests and cases since 1st March. I am beginning to wonder if the recent peak was a rather artificial one, driven by unlocking whilst the Euro's were underway with students still at University. I have added a trendline to show what it would look like if we are returning to the more gentle growth rate observed in May and early June:
2. This effect looks even more obvious if we look at cases per 100k tests:
3. This shows Cases per 100k tests, admissions and deaths with 7 and 14 day lags. Admissions are on a secondary Y axis and deaths are rebased by a factor of 10 to align the curves. Its clear that they are all moving in tandem, albeit the proportion of cases resulting in admission and deaths reduced significantly in May. I believe this is due to the vaccines:
4. Finally, this shows the rolling 7 day total of Cases per 100k Tests and the %age change across those 7 days. When the orange bars are above zero cases are growing. The height of the orange bars shows the rate of change. You can see that the recent slowdown kicked in as we reopened on the 19th July and that it started to bounce back just one week later on the 26th July:
2. This effect looks even more obvious if we look at cases per 100k tests:
3. This shows Cases per 100k tests, admissions and deaths with 7 and 14 day lags. Admissions are on a secondary Y axis and deaths are rebased by a factor of 10 to align the curves. Its clear that they are all moving in tandem, albeit the proportion of cases resulting in admission and deaths reduced significantly in May. I believe this is due to the vaccines:
4. Finally, this shows the rolling 7 day total of Cases per 100k Tests and the %age change across those 7 days. When the orange bars are above zero cases are growing. The height of the orange bars shows the rate of change. You can see that the recent slowdown kicked in as we reopened on the 19th July and that it started to bounce back just one week later on the 26th July:
Edited by Elysium on Wednesday 11th August 18:11
Update on weekly ONS excess deaths
Following 17 straight weeks of below average deaths we began to see deaths above the 5 year average through July 21.
COVID has steadily risen in significance shifting from under 2% of all deaths in week 27 to 4% in week 30.
Excess deaths in the last 4 weeks are about double the total deaths 'due to' COVID in that period. So some other factor is contributing to this. It may be a natural bounce back following months of restrictions or perhaps a deeper signal about long term healthcare disruption. Either way, it is statistically significant as deaths are usually very stable at this time of year.
This graph shows deaths 'due to' COVID vs excess deaths for the entire pandemic. Its interesting to see how the relationship between these metrics has varied over time.
Following 17 straight weeks of below average deaths we began to see deaths above the 5 year average through July 21.
COVID has steadily risen in significance shifting from under 2% of all deaths in week 27 to 4% in week 30.
Excess deaths in the last 4 weeks are about double the total deaths 'due to' COVID in that period. So some other factor is contributing to this. It may be a natural bounce back following months of restrictions or perhaps a deeper signal about long term healthcare disruption. Either way, it is statistically significant as deaths are usually very stable at this time of year.
This graph shows deaths 'due to' COVID vs excess deaths for the entire pandemic. Its interesting to see how the relationship between these metrics has varied over time.
Muncher said:
Another ONS antibody prevalence update released today.
They have revised downwards their projections for previous months, (down 0.6% on their last modelled figure for 12 July), but again the we are up to 94.2% of adults in England having antibodies as at 26 July.
Do you know if their model takes into account waning immunity from the vaccine?They have revised downwards their projections for previous months, (down 0.6% on their last modelled figure for 12 July), but again the we are up to 94.2% of adults in England having antibodies as at 26 July.
Ashfordian said:
Muncher said:
Do you know if their model takes into account waning immunity from the vaccine?Smollet said:
Ashfordian said:
Muncher said:
Do you know if their model takes into account waning immunity from the vaccine?Israel seems to be the 'canary in the mine' regarding all this and they are reporting efficacy on early vaccinations, hence the boosters, is way down. The data behind the actions still seems to be speculative rather than officially published.
Ashfordian said:
Smollet said:
Ashfordian said:
Muncher said:
Do you know if their model takes into account waning immunity from the vaccine?Israel seems to be the 'canary in the mine' regarding all this and they are reporting efficacy on early vaccinations, hence the boosters, is way down. The data behind the actions still seems to be speculative rather than officially published.
https://twitter.com/davidwdowdy/status/14280997758...
The older people in their sample, who were jabbed first, are not seeing any reduction in vaccine efficacy over time. However this does seem apparent in younger age groups, which in turn shows up in the overall picture
This suggests that this may be a social, rather than biological effect. Increases in infection because precautions are being abandoned amongst the young as society returns to normal.
Ashfordian said:
Muncher said:
Do you know if their model takes into account waning immunity from the vaccine?The pool of adults for whom Covid 19 would be a new thing to their immune system is rapidly diminishing by the day, I think we are only about 5 weeks away from the point where you would really struggle to find any adult in England who didn’t have antibodies.
Muncher said:
It will do as they effectively resample the population every 2 weeks, any drop off would show up here.
The pool of adults for whom Covid 19 would be a new thing to their immune system is rapidly diminishing by the day, I think we are only about 5 weeks away from the point where you would really struggle to find any adult in England who didn’t have antibodies.
Fair enough, but why are cases creeping up over 30k right now? Will there be a sudden downward trend in the next couple of weeks?The pool of adults for whom Covid 19 would be a new thing to their immune system is rapidly diminishing by the day, I think we are only about 5 weeks away from the point where you would really struggle to find any adult in England who didn’t have antibodies.
Muncher said:
Ashfordian said:
Muncher said:
Do you know if their model takes into account waning immunity from the vaccine?The pool of adults for whom Covid 19 would be a new thing to their immune system is rapidly diminishing by the day, I think we are only about 5 weeks away from the point where you would really struggle to find any adult in England who didn’t have antibodies.
According to the data an estimated 92.8% of 75 to 79-year-olds in England had Covid-19 antibodies in the latest week, down from 94.4% a month earlier.
For people aged 70 to 75, the figure has dropped from 94.9% to 92.9% over the same period. For those aged 80 and over it has fallen from 92.9% to 92.4%.
This report in the BMJ is relevant: https://www.bmj.com/content/374/bmj.n2074
“There is now quite a lot of evidence that all vaccines are much better at reducing the risk of severe disease than they are at reducing the risk from infection. We now know that vaccination will not stop infection and transmission, [but it does] reduce the risk. The main value of immunisation is in reducing the risk of severe disease and death.”
It isn't a long article, and is worth reading.
“There is now quite a lot of evidence that all vaccines are much better at reducing the risk of severe disease than they are at reducing the risk from infection. We now know that vaccination will not stop infection and transmission, [but it does] reduce the risk. The main value of immunisation is in reducing the risk of severe disease and death.”
It isn't a long article, and is worth reading.
never have so many people been tested when not even unwell but yeah #casesGassing Station | News, Politics & Economics | Top of Page | What's New | My Stuff