lpm wrote: Sat Aug 22, 2020 8:26 am
It's the numbers you are providing that are duds. Where did you get them from? Your numbers cannot possibly right - unless the UK track and trace service miraculously doubled its success rate over the last 3 or 4 weeks.
I have to say that I have quite a bit of doubt about some aspects of their results
Looking at the map for the NW of England, there are LAs with estimates of >100 cases per 100K directly adjacent to LAs with zero.
Which seems unlikely
covid2.PNG
One thing they are doing wrong (I do contribute to the app) is they send out test kits on a sample basis to those reporting symptoms (which is good) but not to those not reporting symptoms.
What they should do is compare both in order to see not just the probability of the various symptoms being COVID-19, but also the proportion not displaying symptoms, or those who later develop symptoms.
In terms of age structure of cases, here's what the distribution of confirmed cases looks like in Portugal:
Screenshot_2020-08-22_15-57-46.png (24.9 KiB) Viewed 7719 times
A small but perceptible increase in under 20s getting it, and a corresponding decline in the proportion of cases in the 70+ groups.
For context, proper lockdown from mid-March, gradual unlocking from mid-May and (AFAICT) high compliance with things like masks and distancing (at least in public).
I haven't yet found a similar plot for Italy (and I haven't found the raw data to make my own plot) but from this pdf we can see how the average age of covid-positive cases has decreased
av-age.png (110.4 KiB) Viewed 7689 times
and how more than half of the positive covid results are in asymptomatic cases (green)
symptoms.png (105.25 KiB) Viewed 7689 times
The dark blue refers to cases with barely any symptoms, yellow is light symptoms, and orange is severe symptoms.
having that swing is a necessary but not sufficient condition for it meaning a thing
@shpalman@mastodon.me.uk
@shpalman.bsky.social / bsky.app/profile/chrastina.net
threads.net/@dannychrastina
lpm wrote: Sat Aug 22, 2020 8:26 am
It's the numbers you are providing that are duds. Where did you get them from? Your numbers cannot possibly right - unless the UK track and trace service miraculously doubled its success rate over the last 3 or 4 weeks.
I have to say that I have quite a bit of doubt about some aspects of their results
Looking at the map for the NW of England, there are LAs with estimates of >100 cases per 100K directly adjacent to LAs with zero.
Which seems unlikely
covid2.PNG
One thing they are doing wrong (I do contribute to the app) is they send out test kits on a sample basis to those reporting symptoms (which is good) but not to those not reporting symptoms.
What they should do is compare both in order to see not just the probability of the various symptoms being COVID-19, but also the proportion not displaying symptoms, or those who later develop symptoms.
To be fair, it IS the Covid-19 SYMPTOM tracker.
The original idea was for them to track the various symptoms that Covid presents. It was partly through their work that the loss of smell/taste was added to the official symptoms.
They don't purport to track asymptomatic cases, or cases in total. They're only tracking symptomatic cases and doing research around the range of symptoms and how symptomatic patients are affected by other factors in their lifestyle.
lpm wrote: Sat Aug 22, 2020 8:26 am
It's the numbers you are providing that are duds. Where did you get them from? Your numbers cannot possibly right - unless the UK track and trace service miraculously doubled its success rate over the last 3 or 4 weeks.
lpm wrote: Sat Aug 22, 2020 1:15 pm
They, and you, are claiming cases have nearly halved when the official UK statistics claim cases have nearly doubled.
Provide the link to where they get stuck into this massive discrepancy. If true, it's a major finding with huge implications for public policy.
If they haven't thought to get deep into the causes of this, they are a bunch of f.cking idiots .
I have to say that I have quite a bit of doubt about some aspects of their results
Looking at the map for the NW of England, there are LAs with estimates of >100 cases per 100K directly adjacent to LAs with zero.
Which seems unlikely
covid2.PNG
One thing they are doing wrong (I do contribute to the app) is they send out test kits on a sample basis to those reporting symptoms (which is good) but not to those not reporting symptoms.
What they should do is compare both in order to see not just the probability of the various symptoms being COVID-19, but also the proportion not displaying symptoms, or those who later develop symptoms.
To be fair, it IS the Covid-19 SYMPTOM tracker.
The original idea was for them to track the various symptoms that Covid presents. It was partly through their work that the loss of smell/taste was added to the official symptoms.
They don't purport to track asymptomatic cases, or cases in total. They're only tracking symptomatic cases and doing research around the range of symptoms and how symptomatic patients are affected by other factors in their lifestyle.
True, but they do have 4-million contributors and are now sending out tests. So as well as seeing how many positive results there are from those reporting symptoms, they should also see the converse.
The level is also low at the moment.
In most constituencies, we're talking about say under three thousand contributors so with 1000 cases per million, that's only 3 amongst the contributing population.
And even overall 1265 cases in 67-million means that there are only about 80 symptomatic people reporting.
That number will be subject to quite a bit of noise.
It would be useful to know too how well the Kings study samples the demographics of the infection hotspots.
I could well imagine they have a sample bias towards IT specialists currently working from home etc, and poorer coverage of the zero-hours contract workers at the local abbatoir.
Long twitter thread about how the UK is organising the quarantines. The author is someone who has just arrived back from Croatia and the thread details the various measures in place and how his quarantine will be enforced.
Marcus Fysh battles hard to stay in the tight race for stupidest MP of 2020. I have faith in you Marcus.
The Tory backbencher Marcus Fysh said: “Masks should be banned in schools. The country should be getting back to normal not pandering to this scientifically illiterate guff,” he said. “ It is time to end the fear. And keep it away from our kids thank you very much.”
This place is not a place of honor, no highly esteemed deed is commemorated here, nothing valued is here.
What is here was dangerous and repulsive to us.
This place is best shunned and left uninhabited.
Little waster wrote: Tue Aug 25, 2020 7:49 pm
Marcus Fysh battles hard to stay in the tight race for stupidest MP of 2020. I have faith in you Marcus.
The Tory backbencher Marcus Fysh said: “Masks should be banned in schools. The country should be getting back to normal not pandering to this scientifically illiterate guff,” he said. “ It is time to end the fear. And keep it away from our kids thank you very much.”
There was a U turn on the U turn as well. Journalists were briefed it was to be masks for all secondary school pupils but then once the announcement was made it changed to only in lockdown areas and up to individual heads elsewhere.
Poor heads! They aren't public health experts and whatever they decide will piss some people off.
This place is not a place of honor, no highly esteemed deed is commemorated here, nothing valued is here.
What is here was dangerous and repulsive to us.
This place is best shunned and left uninhabited.
Key points being (1) the employees were wearing masks (2) it appears patient 1 of the cluster was sitting directly under the aircon unit.
Even more key points beng that the employees were standing up doing work, while customers were sitting down eating and drinking. The airflow was probably somewhat downward (due to it being chilled by aircon). And that the employees were probably further away from the customer most of the time as they had work to do in areas where customers were not permitted. And, of course, one example is poor evidence. There are lots of people running shops wearing masks. Even by chance when an infectious person is present, there will be cases where no employee caught the disease and cases where every employee caught the disease.
And, of course, it is highly relevant when the 27 people were in the shop and for how long compared to the infectious person.
Awarded gold star 4 November 2021
