Yes - in a country like the UK where there is very little testing, the official number of cases will massively understate the actual number who have it, but the death rate as a proportion of those who are tested will be high, because the sample of people who are tested is not representative of the population of people who contract the disease - it will be biased because it contains only the more severe cases who reach hospital treatment.

There will no doubt be some unscrupulous media source who runs a story that claims a super high mortality rate based on this.

Data on caseloads will be more accurate in countries like China/Korea where there is more extensive testing.

In the UK where so few people are tested, the best proxy to use for whether we are getting on top of the virus or not will be the number of deaths per day, but the death rate today is an indicator of the case load two weeks or so ago.

If for instance, the measures that the government announced today for additional social distancing are effective, the number of deaths per day may still rise rapidly over the next two weeks. This will be because the number of cases was rising rapidly in the two weeks before today, when we didn't have the measures in place, then if those are effective, it will start to level off in a couple of weeks and hopefully fall.

But again, people who don't understand the dynamics of the data, would see death rates rising in the next two weeks and say "social distancing is not making any impact".

The other factor that will start to drive deaths per day up rapidly (as found in Italy) is when the NHS gets to overcapacity, where people who could have survived with the right treatment start to die because of the lack of capacity. The reason measures like today are so important is to keep the NHS below that level.

So Johnson signs death warrant on the pub and restraunt industry. If you tell ask their customers to stay away but don't instruct them to close they cannot claim on insurance but cannot pay their bills which will lead to bankruptcy and mass unemployment.

Shame on whoever voted for this cretin to lead the country.

I don't think Boris is up to the job of PM, but it's unfair to criticise him over taking measures like this. If he said carry on as normal, he would be signing the literal death warrant of a lot of vulnerable people. Many other heads of state in other countries are taking measures like this and its sensible.

Boris knows the consequences to the economy and he's had to make an unpleasant trade-off. He's given it longer than some other PMs would. Theresa May or Gordon Brown for instance would probably have had this level of lock down last week.

Nope I’m not for one minute saying ‘carry on as normal’ but if you’re telling people not to go to pubs or cafes then why not shut them so at least the owners can claim on their insurance. He’s looking after his pals and protecting the profits of insurance companies.

Why is my wife going to teach a primary school full of kids tomorrow, but I shouldn't go and have a coffee in a cafe? It makes zero sense. He’s hanging pubs, cafes, restraunts, theatres etc out to dry.

Leader my arris.

Ah i didnt realise that and yes you make a good point.

I wonder how a) they detect new cases and b) determine deaths from corona virus if they are no longer testing? Or is it that they are only testing those admitted to hospital.

Cutting through all the various modelling estimates, there are only two factual numbers, in that 55 deaths in 1500 tested positive.

What is not clear is the demographic age breakdown of those two numbers.

From news reports it appears the 55 souls are predominantly elderly or with some underlying health issue,

So 40 of those could be from 400 elderly/infirm tested positive, or from 1000 elderly/infirm tested positive.

10% or 4%

If it is 10% then final numbers will be determined how far this virus transmits around the elderly or infirm population.

Indeed. He'll close the pubs and cafes when he has to but by then he'll be blaming a rising death toll on those who enjoy a coffee or a pint for not following advice. Much as the French govt. has done.

No, I would start from those factual numbers and calculate the factual death to tested positives weekly ratio every Sunday night for each age / lnfirmity range , and then produce an ongoing set of week to week trends graphs.

All tested positive and all deaths/ belated deaths would become included at some point, and a rolling average figure calculated per age group.

You could see where the big numbers are happening, and pick up any spikes.

Just been listening to the bafoon who wants to close pubs clubs restaurants etc to curb the spread of the virus...but I've got to go to work where I'm in contact with 150 plus other mugs...hypocrite springs to mind...stroll on Friday 4 pm I'm off to local thru back door for a well earned beer...suck on that doris

It depends what happens with the untested, whether they recover, or deteriorate Further to become tested, where they will enter the ongoing weekly calculation.

But any changes to the scope of who gets tested will distort those ratios over time. The best rule of thumb is to track how the raw number of deaths is trending over time. Assume the mortality rate stays fairly constant, the deaths per day will give you an indication of the underlying caseload, although bear in mind when the NHS reaches capacity point the mortality rate will probably shift to a higher level and then stay there.

In the next few weeks the daily death count will likely rise rapidly; the point at which we know we are gaining some control is where that levels off and starts to fall. That will show us that the underlying case load started to fall a couple of weeks earlier.