Updated Covid datasets for the UK and antibodies

New Topic
This topic has been archived, and won't accept reply postings.
 wintertree 19 Jun 2020

The ONS have updated their Coronavirus Infection Survey [1].  They have roughly doubled the number of blood samples in their antibody test to ~1,800 and report a rate of (5.4 ± 1.1)% consistent with their previous report.

The weekly surveillance report [2] has updated figures on antibody testing in blood donations - these are presumably less well statistically controlled than the ONS data but have similar numbers - with the notable exception that it's much higher in London but still a factor of 5 or so less than would be needed for effective herd immunity.

A lot of posters, friends and colleagues have wondered if they had the virus back in January; by some quick estimates it's less than 5% of people who I know or who post here so it's not incompatible with the ONS report - let the wondering carry on!  I'm assuming anyone who was hit hard enough in January to consider this would have developed antibodies.

Considering these latest reports, it feels to me like the virus is now largely confined to care settings (ref 2, figs 11 and 26, ref 3 figure 4 suggests 4× the prevalence in care workers to the general population although the confidence on the data isn't great) although it is starting to rear its head in schools (ref 2 fig 26) which is worrying given that they've barely reopened yet in terms of numbers, and with some regions only reopening this week.  It seems to me that we would be best from an infection control perspective to balkanise the schools into those for children of care workers and those for other children. I don't for a minute suggest this is the best from a social or child welfare perspective and now is a dangerous time politically for any "othering".  It does however hi light the importance of widespread testing with rapid turn around times and for equally fast and effective contact tracing.

I hope to hear some other people's thoughts...

[1] https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/con...

[2] https://assets.publishing.service.gov.uk/government/uploads/system/uploads/... - figure also attached to post as an image 

[3] https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/con...

ps - it would be nice if we could upload PNG images to preserve some quality from plots etc.  

Post edited at 11:19

 Flinticus 19 Jun 2020
In reply to wintertree:

My brother-in-law and his wife think they may have had it in January. She's a doctor and was going to get herself tested recently- not sure if that has happened as we're not frequently in touch with them. 

 plyometrics 19 Jun 2020
In reply to Flinticus:

My sister in law was adamant she had it in January, as she had the dry cough that lasted for ages. She paid for a test recently and it transpires she’s not had it. 

My wife and I also had the horrible cough in January, as it felt most of Kendal did too. Must admit, it crossed my mind whether we’d had the virus, but suspect strongly now it wasn’t; particularly in light of my sister in law’s test results. 

 girlymonkey 19 Jun 2020
In reply to wintertree:

Do we know yet how long any antibodies last? If people had it 6 months ago they may no longer have antibodies? I presume we would need long term virus and antibody testing following people for a number of months. Has this happened at all?

OP wintertree 19 Jun 2020
In reply to girlymonkey:

I’ve not seen a discussion of repeat measurements.  The seroprevalence testing in blood donors would seem ideal for this.  I don’t know if the data is collected of collated at a sufficient level.

My non expert number-from-my-arse leaning is that they should persist for at least a year but I’d take that with a pinch of salt.

 galpinos 19 Jun 2020
In reply to plyometrics:

There's a 20% false negative on the Roche test, if that's what she had.

 MG 19 Jun 2020
In reply to wintertree:

Just heard out neighbour tested positive for antibodies, so assume my wife who was similarly ill too will when results back. Will be interesting to see if I do, having been in close contact with her but not noticeably ill.

OP wintertree 19 Jun 2020
In reply to plyometrics:

Interesting; thanks.  This is the alternate possibility batted about on here at the time - covid came at the tale end of some other ungodly awful respiratory infection meaning that a lot of people where in poor respiratory healthy when exposed to covid.  I’ve seen some suggestion that another nasty coronavirus was circulating around December / January.  In a twisted way this could be quite promising as it might suggest the virus doesn’t have the same level of opportunity in did early in the year.  All highly speculative still.

OP wintertree 19 Jun 2020
In reply to MG:

Do report back to class after your tests.  It’s been similar in our household so we should really take an antibody test.  

Looking back over the last few years I’ve decided that exposure to large numbers of nationally and globally mixed students, often in small and poorly ventilated rooms, combined with my advancing years is probably behind my increasingly awful respiratory infections each winter.   Enough might be enough.

 neilh 19 Jun 2020
In reply to wintertree:

What about it being just flu?Are you certain it was respiratory. Only a thought.

 jonny taylor 19 Jun 2020
In reply to wintertree:

> combined with my advancing years

You're not that old...

 SDM 19 Jun 2020
In reply to girlymonkey:

> Do we know yet how long any antibodies last? If people had it 6 months ago they may no longer have antibodies?

We don't know. The lack of high profile deaths following a second infection of people who had previously been hospitalised with covid-19 is confirmation that there is some level of immunity for some period of time.

If you didn't get a reasonable level of immunity for at least 3 months, we would be starting to hear about deaths from reinfections by now from Italy etc.

But how long that immunity will last, we don't know and we also don't know how well it will hold up to future mutations.

I think I am also correct in saying that we still don't have an accurate, reliable antibody test so the numbers from any antibody test need to be taken with a fistful of salt.

> I presume we would need long term virus and antibody testing following people for a number of months. Has this happened at all?

The ONS study is weekly for the first 5 weeks, then switches to monthly for 1 year. So we might get some useful data from that, assuming the test is/will become accurate and reliable.

 SDM 19 Jun 2020
In reply to plyometrics:

We are seeing more and more reports that the virus was present in Europe since about December.

What is always absent from these reports is that the timing and pace of the outbreaks in Europe are not compatible with this timeline unless either:

a) our understanding of the rate of spread and incubation is incorrect or

b) the virus mutated so that the virus that was present in December was not the same virus that killed so many people in February-May.

If the virus that caused so many fatalities in Europe was present in Europe as early as December, the spike in excess deaths would have begun much earlier.

OP wintertree 19 Jun 2020
In reply to neilh:

> What about it being just flu?Are you certain it was respiratory. Only a thought.

I don't know; but the symptoms experienced by many seemed well beyond typical flu.  One change I am thinking of making is taking and storing personal blood samples during future illnesses so that if something like this happens again I can get an unambiguous answer when the dust settles.  If we do antibody tests in the household now it won't actually confirm when ​​​​​​​we had it.

 mik82 19 Jun 2020
In reply to wintertree:

I think it means we have to be extremely careful not to be complacent. 57% tested positive in the centre of Italy's outbreak, Bergamo, so despite the downplaying of the severity in parts of the media there was definitely the potential for 10 times the current number of deaths, even without the healthcare system being overwhelmed.

You can't really allow an infection with a fatality rate of over 1% to pass through the community so we really need to ensure that our public health measures, particularly contact tracing are good enough to control any further outbreaks. Otherwise we're left with the blunt tools of travel restrictions, closing non-essential things, and further lockdowns. I really do not want everyone to be stuck, unable to see anyone again over the miserable winter months.

OP wintertree 19 Jun 2020
In reply to jonny taylor:

> > combined with my advancing years

> You're not that old...

Still not old enough to know better at any rate..

 mik82 19 Jun 2020
In reply to wintertree:

> I don't know; but the symptoms experienced by many seemed well beyond typical flu.  One change I am thinking of making is taking and storing personal blood samples during future illnesses so that if something like this happens again I can get an unambiguous answer when the dust settles.  If we do antibody tests in the household now it won't actually confirm when we had it

Maybe H3N2 flu? It was the dominant strain this winter, and it did cause quite a nasty outbreak in Australia earlier in the year. In fact we were quite worried that the NHS would be overwhelmed by a similar outbreak here but it never materialised. Proper flu wipes people out for at least a week, and can take weeks to get over.

 Luke90 19 Jun 2020
In reply to girlymonkey:

> Do we know yet how long any antibodies last? If people had it 6 months ago they may no longer have antibodies?

Some preliminary but not very encouraging studies are starting to come out...

https://arstechnica.com/science/2020/06/immunity-to-covid-19-may-wane-just-...

 DaveHK 19 Jun 2020
In reply to wintertree:

> it is starting to rear its head in schools (ref 2 fig 26) which is worrying given that they've barely reopened yet in terms of numbers, and with some regions only reopening this week.  

It says in that doc that the school outbreaks predate the wider reopening of schools so maybe relates to schools being open as key worker child care hubs and suchlike?

Post edited at 16:01
OP wintertree 19 Jun 2020
In reply to DaveHK:

> It says in that doc that the school outbreaks predate the wider reopening of schools so maybe relates to schools being open as key worker child care hubs and suchlike?

Thank you - I missed that and was out by a week.  So we must wait another 6 days to see how the first week of school re-opening is going...

 DaveHK 19 Jun 2020
In reply to wintertree:

As a teacher in Scotland where we don't reopen until August I'll be watching that bit with interest!

 cathsullivan 20 Jun 2020
In reply to SDM:> We are seeing more and more reports that the virus was present in Europe since about December.

....

> If the virus that caused so many fatalities in Europe was present in Europe as early as December, the spike in excess deaths would have begun much earlier.

I keep thinking about this. I had something weird and lingering in December (actually started right at the end of November) with major fatigue, gastrointestinal symptoms, sore chest and throat, headaches. Thought maybe I had glandular fever. Anyway ... will probably remain a mystery. But I am tempted by the idea it was COVID, as we all love a neat explanation. Except I keep thinking that, as you say, if COVID was around that early surely we'd have seen an earlier spike in excess deaths.

 girlymonkey 20 Jun 2020
In reply to Luke90:

Thanks for that. That is indeed disappointing. 

 oldie 20 Jun 2020
In reply to cathsullivan:

> But I am tempted by the idea it was COVID, as we all love a neat explanation. Except I keep thinking that, as you say, if COVID was around that early surely we'd have seen an earlier spike in excess deaths. <

There have been other UKC threads where people have thought they may have had the virus earlier. I was sceptical (my wife and I were bedbound in late Dec/early Jan with bad cough, wife's breathing affected, and several neighbours were similar). I'm still sceptical, but less so now. I spoke to another couple recently who had similar in late Nov/Dec, they'd just come back from a trip to China with several days in Wuhan. Does anyone know of a reputable source for a DIY antibody test kit? I've saved money staying at home so would be prepared to fork out £60ish, out of interest rather than any practical reason.  Though I suppose a positive test would mean there was at least a possibility that we would have a degree of immunity or be less likely to be carriers. I've read that three positive results from different sources would be needed for it to be reliable

 SouthernSteve 20 Jun 2020
In reply to wintertree:

So much more to be known. What proportion of mildly infected people produce antibodies and then if they do how long does that antibody response last? We know after hospitalisation that 92+% produce antibodies and they last at least 2 months, but there is a trend in the data which suggested that younger people had less antibody response which highlights the question over  the mild- or non-clinically infected people and their Ig levels and subsequent immunity.  (https://doi.org/10.1101/2020.06.07.20124636).

From coronavirus more generally immunity might possibly last 3 years, but that was not necessarily mediated by the antibody response.

Recent outbreaks in meat plants in the UK and Germany would suggest that there are other factors at play in transmission, be that temperature or something else, but might given the modellers something tangible to work with. 

 StefanB 20 Jun 2020
In reply to wintertree:

I was convinced I had it in March (at the peak of the outbreak here in Spain) and have now had a negative antibody test (at the hospital, not a DIY test kit).

Not very useful in isolation, but it has served to show me that it is easy to convince oneself of something.

Post edited at 15:03
 jkarran 22 Jun 2020
In reply to wintertree:

This is interesting, not that R is rising so much* but buried in there is strong evidence against the argument some make that most of us are somehow naturally immune without exposure: 657 of 1000 abattoir workers were positive with active covid cases. We can safely assume more than that were exposed and recovered (the seeds of that outbreak) and that there will be a few false negatives there too.

https://news.sky.com/story/coronavirus-germanys-r-number-rockets-again-from...

It'll be interesting to see if they follow this up with antibody testing what fraction show detectable markers.

*I suspect this has more to do with the type of outbreaks, it is no longer representative of normal community transmission but localised outbreaks in confined pressurised workplaces, likely with culture or staffing issues.

jk

 Dave Garnett 22 Jun 2020
In reply to jkarran:

> This is interesting, not that R is rising so much* but buried in there is strong evidence against the argument some make that most of us are somehow naturally immune without exposure: 657 of 1000 abattoir workers were positive with active covid cases. We can safely assume more than that were exposed and recovered (the seeds of that outbreak) and that there will be a few false negatives there too.

I'm starting to think that a plausible model is that a less virulent version of the current virus may have been in widespread circulation before the current pandemic, which would account for some partial immunity being present in some populations.  

At some point, this virus acquired one or more mutations to its S-protein, imparting an extra enzymatic cleavage site (for furin, for example, but also other enzymes like TMPR-SS2).  Enzymatic cleavage at this site seems to be associated with at least increased viral infectivity for cells in the lung (and for cell to cell transmission) and some reports say it may be required for infection of these cells.  When this new, more infectious and more pathogenic, virus took off, it would seem logical that it would hit those with no previous exposure (as well as those with co-morbidities) hard.

All speculation, and possibly bollocks, but something weird does seem to be happening with the pattern of infection and the paradox of the apparently low level of immunity as measured by antibody titre.  Maybe this is because the circulating IgG and IgM levels aren't really correlated with protective immunity (because, presumably, that's more strongly related to an effective CD8+ T-cell response). 

Equally, other, non-immunological, mechanisms might be involved; levels and pattern of tissue expression of ACE-2 expression, ditto TMPR-SS2 or other cellular protease expression, genetic factors controlling intensity and control of inflammatory response... who knows?  


New Topic
This topic has been archived, and won't accept reply postings.
Loading Notifications...