How do we know that the R0 is below 1?

[Kitchendoctor]

They haven’t been testing everyone who has symptoms, so how do they know how many other people everyone is passing the virus to?

Is this statement based an assumption of the real figures based on those diagnosed in hospital?

I just don’t understand how they can be so sure of it when they don’t seem to have all the information they would need.

Even during lockdown, I’m thinking patterns of infection would be something like:

Person at work infects 2 colleagues
1 colleague lives alone and doesn’t infect anyone else.
Other colleague lives with wife and 2 kids, they all become infected.

So I’m imagining most infections still lead to more than one other?

AIUI it’s been calculated by the number of new cases these have in recent days been lower than the day before. This will give them figures to generate the average

But they don’t know numbers of all new cases.

The number of cases has increased in the last couple of days, as more people have been given access to testing. But prior to that numbers of new cases were mostly those tested in hospital and healthcare workers.
So we still don’t have an accurate picture of the total cases.

I guess they are assuming that hospital cases are around 20% of all cases and working it out from there, but we don’t really know.

Actually I’m hoping our hospital cases are much lower than 20 % of the total, as that would make our CFR shockingly high given yesterday’s news that that one third of hospital patients are dying, but that’s a different discussion.

My understanding is that there is a starting value, R0 , which is relatively well known from data across the world. Varies a bit with population density etc etc, but gives a ball park starting point ...but I don't think you really even need that, but the more boundaries you have, the mor accurate the estimate

Once you know that and your starting rate of people going into hospital then you can estimate how it's changed from the change in the rate of people going into hospital

You can add in other data sources, such as the self reported cases, 111 queries etc and watch how they change over time also

Basically, if the numbers are going down , then R is less than 1
If R is very much less than 1 then the numbers will go down quicker
You can do graphs, make a starting number eg 1000, and then draw graphs of how that number changes for different values of R