SARS-CoV-2 testing

Covid-19 discussion, bring your own statistics
Post Reply
User avatar
jimbob
Dorkwood
Posts: 1204
Joined: Mon Nov 11, 2019 4:04 pm
Location: High Peak/Manchester

Re: SARS-CoV-2 testing

Post by jimbob » Thu Jul 16, 2020 8:47 pm

https://www.theguardian.com/world/2020/ ... are_btn_tw

Randox was awarded a £133m contract in March to produce the testing kits for England, Wales and Northern Ireland without any other firms being given the opportunity to bid for the work.
On Thursday the health and social care secretary, Matt Hancock, told MPs: “We’ve identified some swabs that are not up to the usual high standard that we expect, and we’ll be carrying out further testing of this batch as a precautionary measure.

Have you considered stupidity as an explanation

User avatar
Bird on a Fire
Stummy Beige
Posts: 2987
Joined: Fri Oct 11, 2019 5:05 pm
Location: nadir of brie

Re: SARS-CoV-2 testing

Post by Bird on a Fire » Thu Jul 16, 2020 10:56 pm

Oh look! Corruption.
Born at 356.32 ppm CO2

User avatar
JQH
Catbabel
Posts: 785
Joined: Mon Nov 11, 2019 3:30 pm
Location: Sar Flandan

Re: SARS-CoV-2 testing

Post by JQH » Sat Jul 18, 2020 8:17 am

Australian researchers come up with a rapid test.

https://www.aol.co.uk/news/2020/07/17/a ... lood-test/
And remember that if you botch the exit, the carnival of reaction may be coming to a town near you.

Fintan O'Toole

User avatar
headshot
Fuzzable
Posts: 243
Joined: Tue Nov 12, 2019 9:40 am

Re: SARS-CoV-2 testing

Post by headshot » Sat Jul 18, 2020 11:13 am

And the U.K. has a new 20 min antibody fingerprint test too:

https://www.theguardian.com/uk-news/202 ... SApp_Other

User avatar
Woodchopper
After Pie
Posts: 1620
Joined: Sat Oct 12, 2019 9:05 am

Re: SARS-CoV-2 testing

Post by Woodchopper » Sat Jul 18, 2020 11:32 am

Woodchopper wrote:
Fri Jul 10, 2020 4:38 am
Woodchopper wrote:
Tue Jul 07, 2020 7:14 am
Spain has 28 388 confirmed Covid deaths and according to the FT about 48 000 excess deaths. Making the IFR between about 1.2-2%.

ETA The latter number is pretty high. Could be a sign of what happens when the healthcare system gets overwhelmed which causes higher mortality in milder Covid cases and non-Covid cases who can’t get treatment.

Alternatively, another explanation is that large numbers of people who were infected didn’t have detectable antibodies when they did the tests.

EETA as mentioned here people with a mild infection may not have have detectable antibodies for long: https://www.nature.com/articles/d41586-020-01989-z
Another comment on the likelihood that antibodies aren’t detectable after a period.

https://blogs.sciencemag.org/pipeline/a ... -ignorance
On the other hand:

SARS-CoV-2 infection induces robust, neutralizing antibody responses that are stable for at least three months
Here we report that the vast majority of infected individuals with mild-to-moderate COVID-19 experience robust IgG antibody responses against the viral spike protein, based on a dataset of 19,860 individuals screened at Mount Sinai Health System in New York City. We also show that titers are stable for at least a period approximating three months, and that anti-spike binding titers significantly correlate with neutralization of authentic SARS-CoV-2. Our data suggests that more than 90% of seroconverters make detectible neutralizing antibody responses and that these titers are stable for at least the near-term future.
https://www.medrxiv.org/content/10.1101 ... 20151126v1

That’s good in terms of people not being reinfected (at least within three months). But if so we’re looking at something close to a 2% IFR in Spain (with total excess deaths). Which is really bad.

User avatar
Martin Y
Dorkwood
Posts: 1129
Joined: Mon Nov 11, 2019 1:08 pm

Re: SARS-CoV-2 testing

Post by Martin Y » Sat Jul 18, 2020 11:36 am

headshot wrote:
Sat Jul 18, 2020 11:13 am
And the U.K. has a new 20 min antibody fingerprint test too:

https://www.theguardian.com/uk-news/202 ... SApp_Other
Prick.

I mean, you meant "fingerprick" until autocorrupt decided otherwise.

User avatar
headshot
Fuzzable
Posts: 243
Joined: Tue Nov 12, 2019 9:40 am

Re: SARS-CoV-2 testing

Post by headshot » Sat Jul 18, 2020 11:39 am

I did indeed. Could have been autocorrect, could have been phat phingers.

User avatar
shpalman
Dorkwood
Posts: 1585
Joined: Mon Nov 11, 2019 12:53 pm
Location: One step beyond

Re: SARS-CoV-2 testing

Post by shpalman » Fri Jul 24, 2020 11:42 am

Just mentioned on the news that 7.7% of Milan public transport staff were found to have antibodies.
molto tricky

User avatar
shpalman
Dorkwood
Posts: 1585
Joined: Mon Nov 11, 2019 12:53 pm
Location: One step beyond

Re: SARS-CoV-2 testing

Post by shpalman » Mon Aug 03, 2020 9:17 pm

Results from mass antibody testing in Italy (a sample of nearly 65,000 subjects from all over Italy) would indicate that 2.5% of Italians had antibodies (that would be 1.4 million people, six times higher than the official number). In Lombardy, 7,5% which would be 750,000 people vs. about 96,000 official cases so far.
molto tricky

bob sterman
Stargoon
Posts: 136
Joined: Mon Nov 11, 2019 10:25 pm

Re: SARS-CoV-2 testing

Post by bob sterman » Mon Aug 03, 2020 9:25 pm

shpalman wrote:
Mon Aug 03, 2020 9:17 pm
Results from mass antibody testing in Italy (a sample of nearly 65,000 subjects from all over Italy) would indicate that 2.5% of Italians had antibodies (that would be 1.4 million people, six times higher than the official number). In Lombardy, 7,5% which would be 750,000 people vs. about 96,000 official cases so far.
If there have been just 1,400,000 infections and if the reported total of 35,000 deaths is accurate - that puts the IFR around 2.5%. Close to estimates coming out of Wuhan early in the pandemic - but perhaps influenced by the the age structure of the populations affected in Italy.

User avatar
Woodchopper
After Pie
Posts: 1620
Joined: Sat Oct 12, 2019 9:05 am

Re: SARS-CoV-2 testing

Post by Woodchopper » Mon Aug 03, 2020 10:14 pm

bob sterman wrote:
Mon Aug 03, 2020 9:25 pm
shpalman wrote:
Mon Aug 03, 2020 9:17 pm
Results from mass antibody testing in Italy (a sample of nearly 65,000 subjects from all over Italy) would indicate that 2.5% of Italians had antibodies (that would be 1.4 million people, six times higher than the official number). In Lombardy, 7,5% which would be 750,000 people vs. about 96,000 official cases so far.
If there have been just 1,400,000 infections and if the reported total of 35,000 deaths is accurate - that puts the IFR around 2.5%. Close to estimates coming out of Wuhan early in the pandemic - but perhaps influenced by the the age structure of the populations affected in Italy.
Excess deaths were 48 600, which gives us an IFR of 3.47%, which seems too high.

Options seem to be:

a) many people who were infected don't have detectable antibodies any more.
b) something else killed thousands of Italians (maybe possible if the healthcare system collapsed and everyone else couldn't get treatment).
c) the sample is biased - for example if people who were sick are more likely to agree to be tested.
d) There was an IFR of 3.47%, which is what happens when the healthcare system collapses.

User avatar
Woodchopper
After Pie
Posts: 1620
Joined: Sat Oct 12, 2019 9:05 am

Re: SARS-CoV-2 testing

Post by Woodchopper » Tue Aug 04, 2020 3:32 pm

More on declining antibody prevalence. See page 30 of the latest Public Health England Covid surveillance report: https://assets.publishing.service.gov.u ... NAL_V2.pdf

Overall SARS-CoV-2 antibody seroprevalence in blood donors in London has declined from circa 16% in mid-May to circa 9% in the end of July. Yes, blood donors aren't a random sample so there are caveats. But still, unless something odd is happening to blood donors it looks like antibody presence is declining over a few months.

It may be that people with mild or asymptomatic infections experience rapid decline in antibodies. If so the seroprevalence tests will be an under count (and will over estimate the IFR).

User avatar
jimbob
Dorkwood
Posts: 1204
Joined: Mon Nov 11, 2019 4:04 pm
Location: High Peak/Manchester

Re: SARS-CoV-2 testing

Post by jimbob » Tue Aug 04, 2020 3:55 pm

https://www.gov.uk/government/publicati ... -july-2020
Table 1: people newly tested for COVID-19 under pillars 1 and 2, England
9 July to 15 July: number of people 16 July to 22 July: number of people Since Test and Trace launched. 28 May to 22 July: number of people
Pillar 1 132,005 126,414 964,448
Tested positive 885 640 11,646
Pillar 2 231,330 239,983 1,730,749
Tested positive 3,122 3,488 31,318
Total 363,335 366,397 2,695,197
Tested positive 4,007 4,128 42,964
EelZc_UX0AA5sx8.png
EelZc_UX0AA5sx8.png (87.95 KiB) Viewed 292 times
Consistent with a slight upturn in cases.
Have you considered stupidity as an explanation

User avatar
shpalman
Dorkwood
Posts: 1585
Joined: Mon Nov 11, 2019 12:53 pm
Location: One step beyond

Re: SARS-CoV-2 testing

Post by shpalman » Tue Aug 04, 2020 4:39 pm

7-day average cases per day has increased from about 600 to about 800 in the UK during July.

The git I used for the numbers isn't updating anymore.
molto tricky

User avatar
shpalman
Dorkwood
Posts: 1585
Joined: Mon Nov 11, 2019 12:53 pm
Location: One step beyond

Re: SARS-CoV-2 testing

Post by shpalman » Wed Aug 05, 2020 8:19 am

Woodchopper wrote:
Mon Aug 03, 2020 10:14 pm
bob sterman wrote:
Mon Aug 03, 2020 9:25 pm
shpalman wrote:
Mon Aug 03, 2020 9:17 pm
Results from mass antibody testing in Italy (a sample of nearly 65,000 subjects from all over Italy) would indicate that 2.5% of Italians had antibodies (that would be 1.4 million people, six times higher than the official number). In Lombardy, 7,5% which would be 750,000 people vs. about 96,000 official cases so far.
If there have been just 1,400,000 infections and if the reported total of 35,000 deaths is accurate - that puts the IFR around 2.5%. Close to estimates coming out of Wuhan early in the pandemic - but perhaps influenced by the the age structure of the populations affected in Italy.
Excess deaths were 48 600, which gives us an IFR of 3.47%, which seems too high.

Options seem to be:

a) many people who were infected don't have detectable antibodies any more.
b) something else killed thousands of Italians (maybe possible if the healthcare system collapsed and everyone else couldn't get treatment).
c) the sample is biased - for example if people who were sick are more likely to agree to be tested.
d) There was an IFR of 3.47%, which is what happens when the healthcare system collapses.
The excess deaths peaked here before the official covid deaths, in a way which suggests that in the early phase of the pandemic the testing wasn't keeping up with the actual spread of the disease. It's certainly possible that the health service was quickly overwhelmed in that period - there were 4000 covid patients in the ICU at the end of March, when there had only been 100,000 official positives of which 80,000 were still active (and two weeks later there were about 20,000 deaths). Right now there are something like 12,000 active positives, but only 40 cases in intensive care.

So it makes sense if there were actually about 500,000 cases of covid back when the official number was only 80,000.

The official protocol that a positive antibody test meant that you needed to go into self-isolation until you could get a negative swab test (at your own expense) was stupid. I don't think a person is shedding viable virus even if they do have a positive swab for viral RNA while also having antibodies and no symptoms.

The health service here also didn't really know what to do at the beginning. Protocol based on SARS was apparently to get people face-down intubated as soon as possible, and it was known that older patients wouldn't necessarily deal with that very well, so there was a sense of leaving someone at home as long as they weren't too bad, and when they suddenly got worse, oh well it's too late to do anything. We now know better how that works, with the decreasing blood oxygen saturation not necessarily causing any respiratory distress.

Italy also made the mistake of letting the covid get into care homes and hospitals (as well as the thing of extended families living together); 363 medical charts just have been seized by the police here in the province of Como because care homes are accused of manslaughter (and there is also an offense here of epidemia colposa). (I'm not sure how that would map to UK law; "colposa" would mean someone was at fault but through negligence; "dolosa" would mean it was deliberate, but not premeditated.)

But how to we know what the IFR should be?
molto tricky

User avatar
shpalman
Dorkwood
Posts: 1585
Joined: Mon Nov 11, 2019 12:53 pm
Location: One step beyond

Re: SARS-CoV-2 testing

Post by shpalman » Wed Aug 05, 2020 2:47 pm

shpalman wrote:
Mon Aug 03, 2020 9:17 pm
Results from mass antibody testing in Italy (a sample of nearly 65,000 subjects from all over Italy) would indicate that 2.5% of Italians had antibodies (that would be 1.4 million people, six times higher than the official number). In Lombardy, 7,5% which would be 750,000 people vs. about 96,000 official cases so far.
An graph:
FB_IMG_1596638794622.jpg
FB_IMG_1596638794622.jpg (35.51 KiB) Viewed 188 times
molto tricky

User avatar
shpalman
Dorkwood
Posts: 1585
Joined: Mon Nov 11, 2019 12:53 pm
Location: One step beyond

Re: SARS-CoV-2 testing

Post by shpalman » Wed Aug 05, 2020 3:19 pm

shpalman wrote:
Wed Aug 05, 2020 2:47 pm
shpalman wrote:
Mon Aug 03, 2020 9:17 pm
Results from mass antibody testing in Italy (a sample of nearly 65,000 subjects from all over Italy) would indicate that 2.5% of Italians had antibodies (that would be 1.4 million people, six times higher than the official number). In Lombardy, 7,5% which would be 750,000 people vs. about 96,000 official cases so far.
An graph:

FB_IMG_1596638794622.jpg
https://www.laprovinciadicomo.it/storie ... 366947_11/

3.5% in my province, which would be about 21,000 cases, while the official number is 4169 cases. (ETA today's numbers just came out)
molto tricky

PeteB
Sindis Poop
Posts: 78
Joined: Wed Nov 13, 2019 1:02 pm

Re: SARS-CoV-2 testing

Post by PeteB » Thu Aug 06, 2020 7:39 am

Interesting thread on whether cases are really going up

James Annan tried reconciling this idea with his model

User avatar
Woodchopper
After Pie
Posts: 1620
Joined: Sat Oct 12, 2019 9:05 am

Re: SARS-CoV-2 testing

Post by Woodchopper » Thu Aug 06, 2020 7:48 am

PeteB wrote:
Thu Aug 06, 2020 7:39 am
Interesting thread on whether cases are really going up

James Annan tried reconciling this idea with his model
The key indicator is hospital admissions. They don’t yet show an increase. https://coronavirus.data.gov.uk/healthcare

But there will be a lag of up to a couple of weeks between infection and needing hospital treatment.

PeteB
Sindis Poop
Posts: 78
Joined: Wed Nov 13, 2019 1:02 pm

Re: SARS-CoV-2 testing

Post by PeteB » Thu Aug 06, 2020 8:01 am

Woodchopper wrote:
Thu Aug 06, 2020 7:48 am
PeteB wrote:
Thu Aug 06, 2020 7:39 am
Interesting thread on whether cases are really going up

James Annan tried reconciling this idea with his model
The key indicator is hospital admissions. They don’t yet show an increase. https://coronavirus.data.gov.uk/healthcare

But there will be a lag of up to a couple of weeks between infection and needing hospital treatment.
I guess it depends a bit on the age group being infected, I could well see initially a surge in young people being infected which wouldn't result in too many new hospital admissions before they pass it onto parents / grandparents

User avatar
Woodchopper
After Pie
Posts: 1620
Joined: Sat Oct 12, 2019 9:05 am

Re: SARS-CoV-2 testing

Post by Woodchopper » Thu Aug 06, 2020 8:15 am

PeteB wrote:
Thu Aug 06, 2020 8:01 am
Woodchopper wrote:
Thu Aug 06, 2020 7:48 am
PeteB wrote:
Thu Aug 06, 2020 7:39 am
Interesting thread on whether cases are really going up

James Annan tried reconciling this idea with his model
The key indicator is hospital admissions. They don’t yet show an increase. https://coronavirus.data.gov.uk/healthcare

But there will be a lag of up to a couple of weeks between infection and needing hospital treatment.
I guess it depends a bit on the age group being infected, I could well see initially a surge in young people being infected which wouldn't result in too many new hospital admissions before they pass it onto parents / grandparents
True, and that seems to be what happened in the US. If so we can add another few weeks to the lag.

User avatar
lpm
Dorkwood
Posts: 1451
Joined: Mon Nov 11, 2019 1:05 pm
Location: IMPEACH AND EXTERMINATE

Re: SARS-CoV-2 testing

Post by lpm » Thu Aug 06, 2020 8:28 am

It would be rational for young people to go to pubs - very low risk - but fewer 50 year olds. Social get-togethers are group decisions, so six 20 year old friends might agree to meet up in a pub, but in a group of six 50 year olds two might veto the decision so the entire group goes to a garden BBQ instead.

John Burn-Murdoch has the charts of activity:
https://twitter.com/jburnmurdoch/status ... 1873455110

Image

It would be very useful to see something like this split by age grouping. Anecdotally, it's young people out and about.

Economic factors are also relevant. With the end of the furlough scheme and job losses, millions won't be able to afford shopping and restaurants and pubs. This might shift the mix.
I'll miss him after he's died in the pandemic

PeteB
Sindis Poop
Posts: 78
Joined: Wed Nov 13, 2019 1:02 pm

Re: SARS-CoV-2 testing

Post by PeteB » Thu Aug 06, 2020 8:47 am

lpm wrote:
Thu Aug 06, 2020 8:28 am
It would be rational for young people to go to pubs - very low risk - but fewer 50 year olds.
Yes, but it's not the risk to the young people themselves, it's everyone else they come into close-ish contact with on public transport, offices, family (and then all their contacts) - I don't think there is any practical way of limiting your close-ish contacts to your own age group

User avatar
shpalman
Dorkwood
Posts: 1585
Joined: Mon Nov 11, 2019 12:53 pm
Location: One step beyond

Re: SARS-CoV-2 testing

Post by shpalman » Thu Aug 06, 2020 9:12 am

Fine, the number of cases in the UK isn't increasing. It was always that high but testing was sh.t. Well, it's still sh.t, but getting a bit less sh.t than before. There's no clear idea now of how many swabs have been performed, and the number of people tested was lost track of ages ago. And there's no record of how many people have recovered.

And the UK is still on over 800 cases per day and currently about 400 deaths per week (compared to an ONS death rate of about 8000-9000 per week at this time of year, so maybe 4-5% of all deaths). If you take that death rate and assume an IFR of 2.5-3.5% as the Italian antibody study suggests, you're looking at the order of 2000 cases per day (yeah I know that number applies to two weeks ago or whatever but unless deaths go down significantly in the next two weeks, it's valid now).

Because comparisons with other countries are meaningless,* Italy has a covid-positive death rate of about 50 per week at the moment. (Italy is testing 27,000 cases per day, performing 50,000 swabs, and the fraction of people tested who turn out to be positive is about 1.1% at the moment. Cases per day is going up in Italy because of a few localized outbreaks, for example in Mantova (Lombardy was actually stable at about 60 cases per day until that happened; the nearby regions of Veneto and Emilia-Romagna are increasing but from much lower baselines).

* - Weird how the UK decided comparisons with other nations were meaningless at the point at which it started to obviously do worse than other nations.
molto tricky

User avatar
lpm
Dorkwood
Posts: 1451
Joined: Mon Nov 11, 2019 1:05 pm
Location: IMPEACH AND EXTERMINATE

Re: SARS-CoV-2 testing

Post by lpm » Thu Aug 06, 2020 9:22 am

shpalman wrote:
Thu Aug 06, 2020 9:12 am
And the UK is still on over 800 cases per day and currently about 400 deaths per week (compared to an ONS death rate of about 8000-9000 per week at this time of year, so maybe 4-5% of all deaths). If you take that death rate and assume an IFR of 2.5-3.5% as the Italian antibody study suggests, you're looking at the order of 2000 cases per day (yeah I know that number applies to two weeks ago or whatever but unless deaths go down significantly in the next two weeks, it's valid now).
The ONS survey for week of 20 and 26 July estimated 4,200 new infections per day.

A reasonable rule of thumb seems to be to use 3x the official confirmed cases per day.
I'll miss him after he's died in the pandemic

Post Reply