UKC

Another thread in response to the latest Covid guidelines...sorry to the Jonny Ball fans.

If we do not know who has had the virus and we are far from sure who currently has it then how is it possible to calculate a meaningful value for R?

To say that it is somewhere between 0.5 and 0.9 is surely just plucking a number from the air? The range says it all. If my employer were to tell me that they would pay me something between £500 and £900 I’d tell them where to go.

What is most frightening of course is that assuming (is it anything more than an assumption?) it to be below 1 is potentially playing with people’s lives.

Jason

you calculate it from the daily admissions into the hospitals.

Ok. Thank you.

So is the assumption then that the trends in the curve for hospital admissions agree with those of the curve for infection in the population as a whole?

yes.

Presumably that means its essentially an estimate based on non-random sampling of the population. 

I think I would be more worried if they gave a precise figure. That would mean they were making it up. Giving a range reflects the fact that there are accepted  unknowns.

You don't need to assume it's below 1 in the general population + hospitals, since the covid admission criteria doesn't seem to have radically changed you can see which side of 1 R is in gently falling hospital admissions.

I doubt it's low enough to get us our lives back for a while yet. 

Jk

Yes, there is an assumption behind that the vulnarable/not vulnarable ratio stays constant.

But in some sense the hospital numbers are more useful than an accurate estimate of the amount of cases and R0 in determining policies: the first goal is not to overwhelm the NHS, so tracking the number of hospital cases over time compared to hospital capacity gives a good handle on the situation.

Yes, that makes sense, ta.

Thanks, I am getting a better understanding of it now.

J

In other words you create a three of four day lag, before you get any useful information.

This is not what they are doing in countries which have a good grip on the virus and it's control.

You don't, you also lose any chance to keeping the spread localised, because of the lag in cases presenting to hospital and those catching it.

I the most recent Boris-a-thon, the chief medical examiner was replying to a question about how we open things up again, he said, "there should be something in place to help us trace cases".

Yet no one asked WHAT will be in place for us to do any tracing, or WHY haven't we got this in place before we start lifting restrictions. If we're going to have something, we should have it NOW, so we know it works and if there are issues, we sort them out before we test it with live people!!

It's all a f*cking shambles that's getting worse and worse.

It sometimes feels like Boris is trying to out Trump, Trump.

Well buried in today's press conference were the estimates that 4% of the population had been infected "a couple of weeks" ago (antibodies) and that 136,000 currently have infection (swabs). This is going to be based on data from our population sampling studies rather than hospital admissions and I'd expect that this is also involved in estimating R

Also, based on the proportion of the population infected 2 weeks ago you can actually calculate a fatality rate of about 1% - as was initially predicted.

All this means is that there's no chance of doing any meaningful contact tracing with so many people infected currently and there's a high risk of a surge in cases when relaxing lockdown measures.

The fear that we are too late is one I have shared for some time. We were in the fortunate position to be able to learn from other countries’ experiences and we blew it. The virus is likely now so deep set in the population that learning to live with it feels like our only option. Sometimes I think that the government has gone for the herd immunity route by stealth; cloaking their plans with the least plausible lockdown.

Jase

There was a reasonable discussion of R in tonight's Newsnight (Mon 11th May), including an interview with David King (a previous Chief Scientist) who gave his frank and full opinion about the recent announcement and the likelihood of any scientific input affecting it; he was then followed by other interviewees attempting to defend the government.

This very clear 20 minute explanation using a spreadsheet shows you how to do simple calculations for yourself: youtube.com/watch?v=oRCn5xPLCKs&

The sensible epidemiological course is to continue the lockdown; the sensible economic course is to get the economy going again. At the moment, there's no testing or contact tracing capability, so we can only guess what's going to happen, and have few ways to control it. Let's hope we get lucky in the next few weeks.

I watched that and found it good.

I was also very impressed by Lawrence Freedman's later contribution on the same programme. Clever chap.