Coronavirus outbreak

[roubaixtuesday]

Let's try a different approach.

Explain herd immunity to me.

Sure, but it's irrelevant to the point at hand. Let's do that first.

The ONS survey randomly samples people and tests them with PCR (polymerase chain reaction).

PCR detects viral RNA - you only have this during or very shortly after (days at most) an infection. It's very sensitive, to the point that a fair proportion of people who test positive will not actually be infectious - their viral load will be too low. So it gives, if anything, an upper bound on the proportion of infectious people.

They do the survey weekly. The week you quoted was the peak of the second wave. Since then, all the people tested positive then would now be negative by PCR. So it is *not* cumulative.

The latest figures, in the link I gave, show 0.3% prevalence in London. So in a crowd of 300, just one person would be positive, on average.

Once you've recovered, and if you test negative by PCR, you are no longer infectious.

I how that makes sense. Please come back if not.

 

[Pale Rider]

You've managed to 'explain' herd immunity without mentioning herd immunity.

I'm guessing you now agree - with the ONS, not me - that about 1 in 25 or 30 Londoners have had Covid.

How many of those remain infectious is simply not known, although the ability of a Covid carrier to infect others does appear to drop over time.

A better argument for allowing mass gatherings is the success of the vaccination programme.

In other words, it largely doesn't matter if a vaccinated person catches Covid because the vaccine will usually suppress the impact of Covid to a trivial level.

Although even that is not much comfort to those who have been vaccinated but who still suffer serious symptoms or death.

 

[roubaixtuesday]

Next, the testing for antibodies.

Antibodies are produced by the immune system in response to infection.

This *is* cumulative (roughly); antibody response is long lived, though there is some decline over months, and we don't know how much yet. It's testing whether someone has *ever* been infected, not whether they currently are.

This testing is sometimes called serology, because its a test of serum, from blood, not a swab test of the nose and throat which is what's used for PCR.

The ONS also does a serology survey, but much less frequently. I think this is the most recent, showing 15% are antibody positive.

https://www.ons.gov.uk/peoplepopula...ctionsurveyantibodydatafortheuk/3february2021

 

[roubaixtuesday]

that about 1 in 25 or 30 Londoners have had Covid

No, please see the next post. Its about one in seven nationally. Though that's an underestimate as antibody response declines over time, plus it was done a few weeks ago so will have increased, plus London is probably much higher than the uk average.

 

[midlife]

I thought the problem with PCR was that it was very sensitive and could pick up fragments of dead virus a long time after infection?

 

[Craig the cyclist]

I am really enjoying this battle of the improbable stats

 

[roubaixtuesday]

How many of those remain infectious is simply not known

This is not the case.

If you test negative by PCR you're pretty much unequivocally not infectious.

Only 0.3% of the current population of London are PCR positive, according to that ONS survey.

I'll try and find something authoritative on this for you.

 

[roubaixtuesday]

I thought the problem with PCR was that it was very sensitive and could pick up fragments of dead virus a long time after infection?

Nah, not really. A few days.

 

[roubaixtuesday]

@Pale Rider

I wanted to explain the different tests before going to herd immunity.

The R number defines, on average how many people each infected person will infect.

But if some are immune, that number reduces.

If R=4 for current COVID strain, then on average every infected person will infect for others.

So let's imagine 5 people.

One catches COVID.

R=4, so they infect all four others, and the virus expands exponentially.

But now imagine three of those are immune. Whether through vaccination or previous infection.

Now, the one person only infects that one other person. And R=1.

We don't have a definitive test for immunity, but the serology test is the closest we have.

Once that 15% becomes 75% we have herd immunity.

[There are many caveats and details to this, but that's the gist of it]

 

[Craig the cyclist]

I think it's time for you both to reveal your credentials in the use of all these stats! I am expecting big things from roubaixtuesday!

 

[Ajax Bay]

You've managed to 'explain' herd immunity without mentioning herd immunity.
I'm guessing you now agree - with the ONS, not me - that about 1 in 25 or 30 Londoners have had Covid.

Last things first. About 1 in 5 Londoners have had C19, but not many (1 in 300) have it now (positive test or would test positive).
Herd immunity. Here's my text from 29 Jan, with a bit of comment:
"Easing of lock down in March may (will) have an effect of increasing the R number (R was 0.98 (range of 0.92 to 1.04) on 22 Jan, and dropping, so by March, well below 1.0) but that will be countered (in terms of number of cases) by the reduced number of the population 'in play': ~18% (12M) will have been vaccinated, rising weekly, to which add those who've had COVID-19 and who still have a sufficient level of antibodies - maybe 7M. So the 'new case' rate will be low and probably fall. Comment: This happened.
These factors and effects mean that reducing the restrictions (aka ending lock down) then will be a reasonable government decision, and one which avoids accusations of delay and procrastination (though no doubt there will be other views and criticisms, and rightly so, we live in an open society).
"Provided vaccine supplies remain sufficient (to allow 2.5M doses delivered per week), we should have vaccinated all the Groups (1-9 - #30M (NB 33M less 10%)) laid out by JCVI by mid Apr (all over 50s). NB takes into account that second doses will take about half the delivery from 14 Mar onwards. Comment: JCVI 1-9 will now be achieved by Easter, thanks to increased supply to 4M doses a week"

 

[Ajax Bay]

From my post from yesterday in the vaccine thread ('liked' by @roubaixtuesday so I'll assume it has their imprimatur)

Key factors which will affect the population percentage where we might expect national herd immunity are:
1) transmissibility of whatever is the by then dominant variant (looking like B.1.1.7) - reflected in Rt, NB the summer months will have a beneficial effect,
2) actual reduction in transmission by vaccinated but infected people of the virus which vaccines afford - hopefully 70+% will not 'transmit',
3) the limited depressive effect on 'R' of whatever restrictions remain,
4) number of unvaccinated million or two who've had COVID-19 (tested and asymptomatic) and retained sufficient antibodies (eg in the under 20 cohort and in those who choose not to be vaccinated).
The 90+% level of uptake recently reported also means that UK herd immunity may be achievable without having to vaccinate the under 16s (UK and USA trials under way to check/allow the latter). Bear in mind primary school age children seem to be less susceptible and contagious than adults, in which case they may be partially omitted from the computation of herd immunity. https://www.medrxiv.org/content/10.1101/2020.07.19.20157362v2 I have not made this exclusion below.
The population of UK is 67M - ONS (incl 13M under 16). At the forecast vaccination rate from yesterday (13 Mar - 4M per week, increasing numbers second dose from now till mid June) we might hope to reach 54M first doses by mid July: which would be 80%.
Opinions vary on the percentage required and to an extent on the definition. It may be we never reach herd immunity for C19 VoC. Haven't managed it against flu.

 

[roubaixtuesday]

I think it's time for you both to reveal your credentials in the use of all these stats! I am expecting big things from roubaixtuesday!

Beware of credentials.

Look for reputable citations. And check for views from other, reputable sources.

Many Internet cranks have impeccable credentials. At least one nobel prize winner has crank status on COVID https://en.m.wikipedia.org/wiki/Michael_Levitt

 

[Craig the cyclist]

Beware of credentials.

Look for reputable citations. And check for views from other, reputable sources.

Many Internet cranks have impeccable credentials. At least one nobel prize winner has crank status on COVID https://en.m.wikipedia.org/wiki/Michael_Levitt

Thanks for the warning, I will beware of your credentials when you tell me what they are.

So you just cut and paste then?

 

[roubaixtuesday]

Thanks for the warning, I will beware of your credentials when you tell me what they are.

So you just cut and paste then?

I would very much encourage you to beware of my credentials. Or those of anyone else, even Nobel Prize winners.

I have some experience and skills relevant to help me, which I would rather not go into, but there is nothing I've posted that cannot be readily researched by anyone with good numeracy, basic science knowledge, and an ability to judge the reliability of sources (libertarian websites bad, public health and published science good).

I am absolutely not an original researcher into COVID, not by a very long distance.