Page 12 of 15
Re: SARS-CoV-2 testing
Posted: Thu Oct 08, 2020 8:45 pm
by badger
badger wrote: ↑Thu Oct 08, 2020 7:50 pm
bob sterman wrote: ↑Thu Oct 08, 2020 5:41 pm
So the number of new cases reported today was about 17,500 - but the Kings College Zoe symptom study is only estimating 21,903 new cases per day.
Can we really be detecting 80% of infections????
Interesting, because if you spread out today's 17,500 over date of specimen, and then extrapolate (assuming 9 day doubling) you get 21/22k today.
See this thread:
https://twitter.com/RP131/status/1314222917580980224
Oh wait. I see. *Cases*. f.cking hell. I have lost the ability to read today. I'm retiring.
Re: SARS-CoV-2 testing
Posted: Fri Oct 09, 2020 1:26 pm
by jimbob
lpm wrote: ↑Thu Oct 08, 2020 8:21 pm
bob sterman wrote: ↑Thu Oct 08, 2020 5:41 pm
So the number of new cases reported today was about 17,500 - but the Kings College Zoe symptom study is only estimating 21,903 new cases per day.
Can we really be detecting 80% of infections????
No of course we can't.
The King College reports have been disasters, week after week.
Have they?
In July and August, they seemed to be picking up noise in most areas, although their sampling did point out what seemed like a cluster in South Wales and the North East even in August.
But now, they've been looking plausible.
Or have I missed something?
Re: SARS-CoV-2 testing
Posted: Fri Oct 09, 2020 1:58 pm
by lpm
For August they claimed their cases halved when official confirmed cases doubled.
Worse than that, they made no attempt to reconcile the conflicting storyline. It's like your accountant telling you your sales are halved while the official figures show they've doubled - you immediately demand an explanation because it's such a critical difference in views.
Re: SARS-CoV-2 testing
Posted: Mon Oct 12, 2020 3:07 pm
by raven
sTeamTraen wrote: ↑Tue Oct 06, 2020 9:11 pm
raven wrote: ↑Mon Oct 05, 2020 2:27 pm
Johnson used the word 'truncated' which does, if he wasn't just using a big word to sound clever yet again, have a specific meaning wrt databases. Or so MrRaven tells me. Something something limits on the size of a datadump something something.
Truncate (or TRUNCATE) is a command in the SQL database language, but it is also a perfectly good computer-related word to describe what happens to any file containing 80,000 records of data when you open them in a system that can't handle that many, and it can be used for a plain text file as well. The use of this word in SQL is derived from its ordinary English meaning. So I think MrRaven is being a little too keen to bash Boris here. (Which is otherwise entirely understandable, of course.)
MrRaven was just consulted on tech stuff; any Boris bashing was me
Re: SARS-CoV-2 testing
Posted: Fri Oct 30, 2020 7:12 pm
by shpalman
Slovakia to test all citizens over age of 10 for coronavirus
That would be 4 million people, to be tested over three weeks, using one of those tests which gives a result in half an hour.
Re: SARS-CoV-2 testing
Posted: Sun Nov 01, 2020 5:43 pm
by shpalman
Re: SARS-CoV-2 testing
Posted: Mon Nov 02, 2020 10:41 pm
by Brightonian
Re: SARS-CoV-2 testing
Posted: Mon Nov 02, 2020 11:57 pm
by sTeamTraen
It's a trap! When the swab goes up your nose it implants a chip in the base of your brane!! Then you're a zombie under control of the 5G signal!!! AND THEIR DOING IT TO A HOLE CITTY!!!!!!!
Re: SARS-CoV-2 testing
Posted: Tue Nov 03, 2020 7:51 am
by shpalman
... or Slovakia, as the two posts immediately above yours explain.
But then this is "military personnel being there to help with manpower" not "dragging people out of their houses at gunpoint" which the use of the form of words "armed forces" might imply.
Re: SARS-CoV-2 testing
Posted: Tue Nov 03, 2020 8:01 am
by shpalman
Re: SARS-CoV-2 testing
Posted: Tue Nov 03, 2020 9:32 am
by JQH
I'm vaguely surprised Johnson has involved the public sector. Or do none of his mates run military contract companies?
Re: SARS-CoV-2 testing
Posted: Sun Nov 15, 2020 11:46 pm
by shpalman
Austria has now decided to try the same thing.
Re: SARS-CoV-2 testing
Posted: Sun Nov 15, 2020 11:49 pm
by shpalman
Another one of those things saying that
covid was already in Italy in September 2019 based on testing of blood samples given for cancer screening.
Re: SARS-CoV-2 testing
Posted: Mon Nov 16, 2020 4:15 am
by Millennie Al
Data from earlier this year suggested that when unchecked, Covid-19 cases would double twice in a week. If we take some time in September to be the 30th, then there are 13 weeks to the end of 2019, or 26 doublings. That means we would expect 67,108,864 cases by the end of the year, which exceeds the total population of Italy. If there really were cases as early as September we also need an explanation of why it was speading so slowly.
Re: SARS-CoV-2 testing
Posted: Mon Nov 16, 2020 8:57 am
by shpalman
Millennie Al wrote: ↑Mon Nov 16, 2020 4:15 am
Data from earlier this year suggested that when unchecked, Covid-19 cases would double twice in a week. If we take some time in September to be the 30th, then there are 13 weeks to the end of 2019, or 26 doublings. That means we would expect 67,108,864 cases by the end of the year, which exceeds the total population of Italy. If there really were cases as early as September we also need an explanation of why it was speading so slowly.
Yes, even if there were ten times more cases than the official numbers (reasonable considering the death statistics) that would have only have meant that it started a week earlier than we thought.
Re: SARS-CoV-2 testing
Posted: Mon Nov 16, 2020 9:09 am
by Woodchopper
Millennie Al wrote: ↑Mon Nov 16, 2020 4:15 am
Data from earlier this year suggested that when unchecked, Covid-19 cases would double twice in a week. If we take some time in September to be the 30th, then there are 13 weeks to the end of 2019, or 26 doublings. That means we would expect 67,108,864 cases by the end of the year, which exceeds the total population of Italy. If there really were cases as early as September we also need an explanation of why it was speading so slowly.
Yes, as with the French reports, a false positive perhaps caused by another coronavirus is far more likely than a non-epidemic a year ago. (The latter might be possible, for example, if somehow there was one person who was infected, that one person somehow got their blood tested, and they then somehow didn't pass it on to anyone else. But that is hardly likely.)
Re: SARS-CoV-2 testing
Posted: Mon Nov 16, 2020 10:59 am
by shpalman
Out of their 959 subjects, there was generally more IgM+ than IgG+, but from September to February there isn't an obvious trend. If anything there was a slightly higher prevalence in September to November, decreasing a bit in December, almost gone in January, and then back quite a lot in February.
It was scattered all over Italy in a way which may or may not correlate with official covid prevalence in the period up until 10th March.
Could there have been a different coronavirus around in the second half of 2019 which induces similar antibodies?
I'm not reading the full text, just looking at the pictures.
Re: SARS-CoV-2 testing
Posted: Mon Nov 16, 2020 11:59 am
by shpalman
they wrote:The serologic assay used in this study is an in-house designed RBD-based ELISA, namely, VM-IgG-RBD and VM-IgM-RBD, and is a proprietary assay developed by using spike glycoprotein (S-protein), which mediates binding to target cells through the interaction between the RBD and the human angiotensin-converting enzyme 2 (ACE2) receptor. The S-protein has been found to be highly immunogenic, and the RBD is considered the main SARS-CoV-2–specific target in the effort to elicit potent NAbs.
ref In our preliminary study, an excellent correlation between the neutralization titer and the IgG, IgM, and immunoglobulin A ELISA response against the RBD of the S-protein was observed,
ref confirming that the RBD-based ELISA can be used as a valid surrogate for neutralization. Therefore, the specificity of the assays used in the present study strongly supports our seroprevalence findings in a relevant number of asymptomatic individuals well before the overt pandemic period, with positive patients in September–October 2019.
Re: SARS-CoV-2 testing
Posted: Mon Nov 16, 2020 10:35 pm
by AMS
Hmm. As you say, crossreactivity to another coronavirus is a big possibility. I haven't looked at the paper, but I'd want to see data on whether their assay picks up other CoV spike proteins too. If they haven't tested for this, I think Occam's razor might apply here.
Re: SARS-CoV-2 testing
Posted: Thu Nov 19, 2020 11:02 am
by raven
That's interesting. My money would be on confusion with antibodies for something else, rather than Covid circulating last autumn.
My dad did antibody testing as part of the, I think, REACT study last week. It's the one like a pregnancy test, with three lines - one to indicate you did the test properly, one for recent antibodies, and one for longer-term ones that would indicate you'd had it in the past.
He swears that last one had a faint line (Mum wasn't convinced though), but he hasn't been ill with anything Covid-like this year. So either it's picking up antibodies to something similar or possibly he had a really mild case. Which seems unlikely as he's been super careful.
OTOH, MrRaven's terrible fluey-cold with a cough that he had in January that he was half-wishing might have been it, probably wasn't as his boss had that and is currently ill with Covid.
Re: SARS-CoV-2 testing
Posted: Sun Nov 22, 2020 10:47 pm
by shpalman
Re: SARS-CoV-2 testing
Posted: Mon Nov 23, 2020 7:20 am
by Woodchopper
A national test is completed every five years or so. But results are only released at a population level.
Re: SARS-CoV-2 testing
Posted: Mon Nov 23, 2020 10:31 am
by badger
Looking at the figures for the recent mass testing pilot in Liverpool
https://www.theguardian.com/uk-news/202 ... o-symptoms
153,000 tests
1,897 positive tests (lateral flow positives "have been asked" to confirm by PCR)
597 of positives asymptomatic (on day of test) which comes in at 24%, which I think is about expected? (All of these from Lateral Flow - anyone with symptoms goes straight to PCR test)
Does anyone know if there's any follow up on whether these cases end up developing symptoms?
The end of the article has a lot of stuff on false positives and concerns around the lateral flow test (as mentioned here too), but if positives are being confirmed by PCR and if there is follow up re symptoms then there will be some useful data for future roll outs, right? Does anyone have any idea how robust the follow up on PCR confirmation (and symptom checking, if any) might be?
Also, it seems take up was above 50% in the 'burbs, but "relatively lower" in poorer areas. There's still another two weeks on the project.
Re: SARS-CoV-2 testing
Posted: Mon Nov 23, 2020 5:31 pm
by Sciolus
Who gives a toss about false positives? At the moment we've got 99% of the population with a false positive result (assuming 1% have a true positive). If that goes down to less than 1%, which is what the article seems to be saying, that's a huge win and a sound basis for reopening normal society.
The important question is the false negative rate. Any indication what that is?
Re: SARS-CoV-2 testing
Posted: Tue Nov 24, 2020 9:15 am
by RoMo
Any tests of the lateral flow devices compare against PCR as the gold standard, which we already know has a false negative rate of 20-30%. So it's at least that.