Coronavirus outbreak

[kingrollo]

So what will happen here then do you think? Given that we basically can't do test and trace controls if we lift the lock down.

We will use the app
And come up with some figures showing that most of those who died would have died anyway....

 

[marinyork]

We will use the app
And come up with some figures showing that most of those who died would have died anyway....

Unfortunately in Singapore only 17% are estimated to have used the app, whereas much higher %s are needed.

 

[marinyork]

So what will happen here then do you think? Given that we basically can't do test and trace controls if we lift the lock down.

Eventually some other country, probably Germany, will do some kind of robust study on antibody immunity and harder figures on % of the population had it in wave-1. This informs policy and what can actually be done beyond about a month's time in the vaguest of terms. If the 15-50%ers are right then we'll have a horrid second wave that's even bigger than now in the winter. If the 2-3%ers are right they'll be horrid wave after horrid wave of this thing until a vaccine.

Track and trace could be deployed to very dense multi-generational housing such as London or the West midlands or care homes.

Eventually the UK and other governments may settle up and plump for an antibody test kit. What else can they do apart from do more robust antibody tests in UK labs? We already know there are validated test kits that are 50% accurate with few false positives (as opposed to truly useless tests that had 30% detection and high false positives and high false negatives). It's just the US and probably UK threw these down in a tantrum and binned them. Unfortunately there may come a time eventually when they have to swallow their pride and use them. It's not ideal, but there's pretty much literally nothing else.

 

[roubaixtuesday]

.

So what will happen here then do you think? Given that we basically can't do test and trace controls if we lift the lock down.

Nobody knows. There is no plan at the moment, it seems. Indeed, the govt seems unwilling to even acknowledge that there should be a plan, lest its own disagreements, jockeying for position and lack of strategy is exposed.

Our government is made up of ideologues and incompetents appointed purely for their willingness to pledge obeisance to Johnson and hard Brexit. All the remaining capable people having were purged. Now, faced with a fast moving problem totally impervious to rhetoric, they are entirely unable to respond and are to be found blinking in the headlights. They spend their time between making unkeepable promises and trying to rewrite the history of their U-turns along the way to our present state, having delivered amongst the very worst epidemics in the world despite having more time to respond than many.

New Zealand, by contrast, has been highly effective and has a prepared level of alerts which can be expected to be stepped through.

https://covid19.govt.nz/alert-system/covid-19-alert-system/

 

[roubaixtuesday]

We will use the app
And come up with some figures showing that most of those who died would have died anyway....

Beware of simple answers to complex problems.

See for example
https://www.bbc.co.uk/news/technology-52353720

No-one thinks apps are the whole solution.

But their defenders say they are a useful weapon in an "armoury of epidemic control measures" and if enough people adopt them, a second wave of cases could be suppressed, saving many lives as a consequence.

 

[icowden]

Why hospital deaths only and how is it used to calculate R0?

Essentially the data from hospital deaths is available immediately. It is recorded as soon as the patient has died and is available to the hospitals reporting function within 24 hours. That data can then be reported to NHS England pretty much straight away.

Data from Care Homes etc can take quite a few days after a person has died. When someone dies in hospital death is certified immediately. When someone dies in a care home a GP is called. They will attend eventually and report the death and apparent cause of death. They may record cause of death inaccurately in that they may go for something generic such as heart failure or pnuemonia rather than being able to test for Covid 19. There is then a delay before that death is entered on the GP system and pushed to the NHS Spine. Then the data has to be extracted and added to the stats by NHS England.

 

[tom73]

Use of an App opens a whole new can of worms and lays bare yet again a whole bunch of social and economic inequalities.
Without even getting started on the whole issues of privacy and data ownership.
The NHS don't have a great record with tech and NHSx approach to building it's own App and owning and controlling the data.
Don't exactly fill you with confidence and the idea of 80% download and use is total pie in sky.
As it stands it won't even get past Apple/Google joint trancing App protocol or data protection measures.

The Ada Lovelace institute did rapid review yesterday on contract tracing apps. With a number of key findings
https://www.adalovelaceinstitute.org/our-work/covid-19/covid-19-exit-through-the-app-store/

 

[lane]

We will use the app
And come up with some figures showing that most of those who died would have died anyway....

No. Everyone who died would have died anyway. Eventually.

 

[lane]

So to sum up my understand so far:

1. We get the RO down as near 0 as we can due to the current lockdown
2. We get the number of infections down to a low enough level due to the current lockdown
3. We start lifting the lockdown
4. We now need to continue to keep the RO below 1 or we will be back where we started. We have no reliable strategy for doing this and it is risky
5. If R0 goes back above 1 we will need another lock down quite quickly

It is likely we will cycle through points 1 to 5 periodically

 

[kingrollo]

Unfortunately in Singapore only 17% are estimated to have used the app, whereas much higher %s are needed.

I didn't say the app would be useful !

 

[mjr]

They’re using Smartphone Apps in Asia I think, don’t know how keen the GB public will be to have such on their phones using data, even if they have a smart phone

If it becomes mandatory and the app wasn't audited by people I trust (security researchers, basically), I'd be resuscitating an old phone that I was about to recycle when we got locked down and keeping the government spyware well away from my real phone and its encryption keys. I would probably carry it enough for the spyware to work (in my saddlebag), but I bet enough people would put it on a spare phone they didn't carry except when required that it would unwittingly undermine the system.

 

[kingrollo]

Essentially the data from hospital deaths is available immediately. It is recorded as soon as the patient has died and is available to the hospitals reporting function within 24 hours. That data can then be reported to NHS England pretty much straight away.

Data from Care Homes etc can take quite a few days after a person has died. When someone dies in hospital death is certified immediately. When someone dies in a care home a GP is called. They will attend eventually and report the death and apparent cause of death. They may record cause of death inaccurately in that they may go for something generic such as heart failure or pnuemonia rather than being able to test for Covid 19. There is then a delay before that death is entered on the GP system and pushed to the NHS Spine. Then the data has to be extracted and added to the stats by NHS England.

I still don't get how they calculate R0 from hospital deaths - without contact tracing....

 

[kingrollo]

If it becomes mandatory and the app wasn't audited by people I trust (security researchers, basically), I'd be resuscitating an old phone that I was about to recycle when we got locked down and keeping the government spyware well away from my real phone and its encryption keys. I would probably carry it enough for the spyware to work (in my saddlebag), but I bet enough people would put it on a spare phone they didn't carry except when required that it would unwittingly undermine the system.

If they are using BT then some phones have a larger BT range than others - will therefore get more alerts than others
.*BT = Bluetooth

 

[roubaixtuesday]

I still don't get how they calculate R0 from hospital deaths - without contact tracing....

The rate of increase of deaths is directly related mathematically to Ro.

 

[mjr]

There's a bit of evidence that the deaths in some european countries deaths in nursing homes is between a third and half the hospital deaths. Some even reckon at a parity with.

More than a bit of evidence! And I'm sorry but it probably could be far worse than a third...

One of the reasons why Belgium's death rate looks so much worse than its neighbours on comparison sites is that the response there is led by seconded experts like Drs Emmanuel André and Marc Van Ranst - not politicians directly like here - and they have been including care home cases and deaths in the toll ASAP, much faster than the UK's week-plus lag. A few days ago, it was reported that over 60% of cases are linked to care homes, although some of those are of course transferred to hospitals as they deterioriate - and probably more than ever now the army are assisting in 10 care homes and they've started to roll out "massive" testing in them.

In case we doubted what politicians would do differently and aren't willing to look at what That Hancock is doing here, Health Minister Maggie de Block suggested that Belgium should change its counting method to be basically in line with others. The independent expert response was impressively scathing: "you don't lower the temperature by changing the thermometer" was the stinging rebuke by Geoffrey Pleyers of UCLouvain. So I don't think they'll change.

Belgian media has even started talking up that other countries will have to adopt "Belgian counting" if they want to have any hope of controlling this pandemic. Maybe the UK will finally change and speed up its fossilised essentially-mid-1800s system where we have 5 days to register a death outside hospital and something like 80 to complete all the reports, and pretty much any funeral director will probably have tales of doctors "identifying" the wrong corpse days later and errors like that.