Coronavirus - General Chat - No statistics

[redsturgeon]

https://www.theguardian.com/commentisfr ... g-covid-19

Interesting opinion piece here on how we affect the virus by our actions.

Suggests quarantine makes for a milder mutation than letting it spread naturally.

John

[Lootman]

The father of a female friend of one of my sons has tested positive and he is showing mild symptoms.

His daughter made the interesting decision to spend a lot of time with her father because, in her words, she "wants to catch it sooner rather than later"

The idea, apparently, is that if the entire family catch it, then they can all self-isolate together, rather than have 14 day quarantine periods that overlap.

I wonder how widespread this desire to "get it over with" is amongst young people? I have heard reports of "Covid parties" where young people try and infect each other.

Irresponsibility or an acceleration of herd immunity? Either way, my son is avoiding her.

[XFool]

Operation Moonshot ignored screening experts. No wonder it's failing

The Guardian

Putting mass Covid testing in place before financial support for those required to self-isolate is putting the cart before the horse


https://www.bmj.com/content/371/bmj.m4436

[kiloran]

The father of a female friend of one of my sons has tested positive and he is showing mild symptoms.

His daughter made the interesting decision to spend a lot of time with her father because, in her words, she "wants to catch it sooner rather than later"

The idea, apparently, is that if the entire family catch it, then they can all self-isolate together, rather than have 14 day quarantine periods that overlap.

I wonder how widespread this desire to "get it over with" is amongst young people? I have heard reports of "Covid parties" where young people try and infect each other.

Irresponsibility or an acceleration of herd immunity? Either way, my son is avoiding her.

This example was quite widely reported: https://www.theguardian.com/world/2020/ ... arty-texas

--kiloran

[Nimrod103]

https://www.telegraph.co.uk/news/2020/1 ... ted-study/

Quarter of people may already be immune to coronavirus – even though many have never been infected PHE researchers believe people with high levels of T-cells likely to have picked up immunity from coronaviruses like common cold. A quarter of people may already be immune to coronavirus even though many of them have never been infected, a new study by Public Health England (PHE) suggests. Over the past few months, researchers have followed nearly 2,850 key workers from the police, fire and health services to gauge levels of immunity to the virus. They discovered that, by June, one in four had high levels of T-cells which recognised Covid, suggesting they had some level of protection against the virus – but nearly half had never been infected. Researchers believe they probably picked up immunity from similar coronaviruses such as those that cause the common cold. In the four months of follow-up, nobody with a high T-cell count became infected with Covid, suggesting they were protected against it.


No wonder the Ferguson predictions have been so wrong.

[Mike4]

https://www.telegraph.co.uk/news/2020/11/19/many-could-immune-covid-despite-never-having-infected-study/

Quarter of people may already be immune to coronavirus – even though many have never been infected PHE researchers believe people with high levels of T-cells likely to have picked up immunity from coronaviruses like common cold. A quarter of people may already be immune to coronavirus even though many of them have never been infected, a new study by Public Health England (PHE) suggests. Over the past few months, researchers have followed nearly 2,850 key workers from the police, fire and health services to gauge levels of immunity to the virus. They discovered that, by June, one in four had high levels of T-cells which recognised Covid, suggesting they had some level of protection against the virus – but nearly half had never been infected. Researchers believe they probably picked up immunity from similar coronaviruses such as those that cause the common cold. In the four months of follow-up, nobody with a high T-cell count became infected with Covid, suggesting they were protected against it.


No wonder the Ferguson predictions have been so wrong.

I thought they were "projections" rather than predictions. Is that wrong then? He made firm predictions?

[Nimrod103]

https://www.telegraph.co.uk/news/2020/11/19/many-could-immune-covid-despite-never-having-infected-study/

Quarter of people may already be immune to coronavirus – even though many have never been infected PHE researchers believe people with high levels of T-cells likely to have picked up immunity from coronaviruses like common cold. A quarter of people may already be immune to coronavirus even though many of them have never been infected, a new study by Public Health England (PHE) suggests. Over the past few months, researchers have followed nearly 2,850 key workers from the police, fire and health services to gauge levels of immunity to the virus. They discovered that, by June, one in four had high levels of T-cells which recognised Covid, suggesting they had some level of protection against the virus – but nearly half had never been infected. Researchers believe they probably picked up immunity from similar coronaviruses such as those that cause the common cold. In the four months of follow-up, nobody with a high T-cell count became infected with Covid, suggesting they were protected against it.


No wonder the Ferguson predictions have been so wrong.

I thought they were "projections" rather than predictions. Is that wrong then? He made firm predictions?

If he said projections, what did 99% of the public hear?

[Mike4]

https://www.telegraph.co.uk/news/2020/11/19/many-could-immune-covid-despite-never-having-infected-study/

Quarter of people may already be immune to coronavirus – even though many have never been infected PHE researchers believe people with high levels of T-cells likely to have picked up immunity from coronaviruses like common cold. A quarter of people may already be immune to coronavirus even though many of them have never been infected, a new study by Public Health England (PHE) suggests. Over the past few months, researchers have followed nearly 2,850 key workers from the police, fire and health services to gauge levels of immunity to the virus. They discovered that, by June, one in four had high levels of T-cells which recognised Covid, suggesting they had some level of protection against the virus – but nearly half had never been infected. Researchers believe they probably picked up immunity from similar coronaviruses such as those that cause the common cold. In the four months of follow-up, nobody with a high T-cell count became infected with Covid, suggesting they were protected against it.


No wonder the Ferguson predictions have been so wrong.

I thought they were "projections" rather than predictions. Is that wrong then? He made firm predictions?

If he said projections, what did 99% of the public hear?

I don't think he 'said' anything. The Imperial College projections were carefully explained probability projections based on various scenarios, IIRC. Then widely mis-interpreted by journalists who can't read properly.

[XFool]

https://www.telegraph.co.uk/news/2020/11/19/many-could-immune-covid-despite-never-having-infected-study/

No wonder the Ferguson predictions have been so wrong.

OK. I'll bite.

Any particular "prediction" of Neil Ferguson you had in mind?

[XFool]

I thought they were "projections" rather than predictions. Is that wrong then? He made firm predictions?

If he said projections, what did 99% of the public hear?

I don't think he 'said' anything. The Imperial College projections were carefully explained probability projections based on various scenarios, IIRC. Then widely mis-interpreted by journalists who can't read properly.

'Alternative' explanation: Then widely mis-represented by journalists, who may have had an agenda all there own.

[Nimrod103]

If he said projections, what did 99% of the public hear?

I don't think he 'said' anything. The Imperial College projections were carefully explained probability projections based on various scenarios, IIRC. Then widely mis-interpreted by journalists who can't read properly.

'Alternative' explanation: Then widely mis-represented by journalists, who may have had an agenda all there own.

I am aware of the subtle but important difference between prediction and projection. But AIUI a crucial assumption in all his modelling was that 100% of the population was vulnerable to catching the virus. PHE says that is not the case.

[XFool]

I don't think he 'said' anything. The Imperial College projections were carefully explained probability projections based on various scenarios, IIRC. Then widely mis-interpreted by journalists who can't read properly.

'Alternative' explanation: Then widely mis-represented by journalists, who may have had an agenda all there own.

I am aware of the subtle but important difference between prediction and projection. But AIUI a crucial assumption in all his modelling was that 100% of the population was vulnerable to catching the virus. PHE says that is not the case.

Well here is a reference to some of his "predictions":

[url]https://en.wikipedia.org/wiki/Neil_Ferguson_(epidemiologist)#COVID-19_–_2020[/url] {Usual problems there}

His team's publication in mid-March of the projections that the UK could face hundreds of thousands of deaths from COVID-19 without strict social distancing measures, gained widespread media attention. In late March, he calculated that with "strict social distancing, testing and isolation of infected cases", deaths in the UK could fall to less than 20,000.

The computer models which Ferguson authored were initially criticised as "unreliable" and "a buggy mess," but subsequent efforts to reproduce the results were successful.

[XFool]

Covid lockdown shows signs of working in England, expert says

The Guardian

Prof Neil Ferguson points to slow decline but warns it is too early for December forecasts

[servodude]

Covid lockdown shows signs of working in England, expert says

The Guardian

Prof Neil Ferguson points to slow decline but warns it is too early for December forecasts

Sheesh - someone should tell him it would have happened anyway!

[Lootman]

Covid lockdown shows signs of working in England, expert says

The Guardian

Prof Neil Ferguson points to slow decline but warns it is too early for December forecasts

Sheesh - someone should tell him it would have happened anyway!

Good point, various reports indicate the UK infection rate peaked on October 19th, just a few days before the announcement of the new lockdown.

Panic, perhaps?

[zico]

Interesting article in the Spectator about whether face-masks work or not. It's interesting because it's a good example how you start from the result that you want to tell people, and then select your statistics and narrative to fit the purpose. Maybe this is best suited for the Science board, but I thought I'd put it here, in case people believed what was stated in this article.

It quotes a study in Denmark in the spring (when masks weren't being widely used). In the study 3,000 people were randomly selected, 50% asked to wear masks, and 50% asked not to wear masks.
It says 1.8% of mask-wearers caught Covid compared to 2.1% of non-masked wearers, and that this difference is not statistically significant (3 fewer people). But it's not significant because the sample is too small, not because there isn't a difference. Not proving a difference is not the same as proving there isn't a difference (Another way of expressing the results is that mask-wearers were 15% less likely to catch covid - but the article doesn't state this).
(To use a football analogy, based on this season's results so far, you can't say that Liverpool are a significantly stronger team than Southampton, or that Manchester City are a significantly stronger team than Newcastle United).

But there are other things wrong with how this study is quoted.
1. It's now widely known that masks primarily protect others from an infected person, rather than protect you from being infected. In this study, mask-wearers would always be a tiny minority of the people they came into contact with.
2. The study has no way of allowing for the fact that people may have felt safer wearing a mask, so more likely to get closer to other people, or allow other people to get close to them.

The last sentence of the article says

"And now that we have properly rigorous scientific research we can rely on, the evidence shows that wearing masks in the community does not significantly reduce the rates of infection."

It absolutely does not show that. The best you can say is that "wearing masks in a community that doesn't wear masks may not significantly reduce the rates of infection, but further research with larger samples would be necessary to determine this". Article was written by Prof Carl Heneghan & Tom Jefferson from the University of Oxford, but it doesn't say it's been checked by a qualified statistician - almost certainly because it hasn't.

https://spectator.us/landmark-danish-st ... nt-effect/

[XFool]

...Furthermore:

https://www.rigshospitalet.dk/presse-og-nyt/nyheder/nyheder/sider/2020/november/dansk-studie-er-nu-offentliggjort.aspx

"Det understreges, at studiet ikke undersøgte maskernes funktion som kildekontrol, dvs. til begrænsning af smitte fra en inficeret person, der bærer maske, til andre. Studiet belyste heller ikke effekt af mundbind i situationer, hvor det ikke er muligt at opretholde den sociale afstand.
Det skal derfor understreges, at disse resultater ikke kan anvendes til at rejse tvivl om, at bred anvendelse af masker uden for sundhedsvæsenet kan være et effektivt middel til at reducere SARS-CoV-2-infektioner."

"It is emphasized that the study did not examine the function of the masks as a source control, ie. to limit infection from an infected person wearing a mask to others. The study also did not elucidate the effect of bandages in situations where it is not possible to maintain the social distance.
It should therefore be emphasized that these results can not be used to cast doubt on the widespread use of masks outside healthcare to be an effective means of reducing SARS-CoV-2 infections."

The original says 6000 people were involved. Half of which is 3000.

[servodude]

The original says 6000 people were involved. Half of which is 3000.

well that is true in the cardinality sense

[johnhemming]

The Danish study only studied whether mask wearing protects the user and concluded that it might, but it also might not, but if it does it does not have lot of an effect.

It did not study whether wearing masks protect other people. I personally am not a fan of masks, but I would think that masks do something to reduce people being infected from someone wearing a mask.

The US Military study showed that even with high levels of compliance with social distancing etc then people still got infected. It demonstrates that we cannot really control the virus although we can influence the rate of spread.

[77ss]

Covid lockdown shows signs of working in England, expert says

The Guardian

Prof Neil Ferguson points to slow decline but warns it is too early for December forecasts

It baffles me why anybody takes Ferguson seriously. A long track record of what seems to me to be gross alarmism.

He and his team nay be great at modelling, but they don't seem to be too good at the underlying assumptions.

Garbage in, garbage out.