A summary of the data I presented at this afternoon's @IndependentSage briefing taking in the latest on Cases, Hospitalisations and Death statistics.
On cases and hospital admissions, it's largely been a good news week.
A thread. 1/13
Cases first.
Across the whole of the UK reported cases have been falling for a couple of weeks now, and fairly steeply, although from a high peak.
We're down to a 7-day average of around 15,000 cases per day.
Across the country cases (per 100K) in most Local Authorities have fallen compared to where we were on lockdown day (3 weeks ago 5/11/20).
A few areas have gone backwards, but for most cases are coming down.
Maps by @TravellingTabby
The number of new confirmed cases (per 100K) across all nations are coming down, as a result of the various restrictions that have been imposed. (Circuit Breaker in NI, Lockdown in England, 5-tier system in Scotland, firebreak in Wales).
Wales looks like it's increasing, but...
...when you look at percentage positivity, all four nations are decreasing (including Wales), which suggests Wales' rise in cases is due to more testing.
Over all a good picture for all four nations.
Delving deeper into cases, we can see drops across all of England's regions. Big drops in some cases, which is encouraging.
To be honest though, if we weren't seeing this sort of movement by now we'd be a bit worried that the lockdown wasn't working.
Moving on to hospitalisations (England). We're starting to see consistent falls in the number of new admissions, which again is good news, although hospital occupancy with COVID patients is still high.
Looking by English region again we can see the biggest drops in new hospital admissions in the regions we were most worried about (with the highest levels of admissions - i.e NW, NE, Midlands).
Zooming in on some of the regions with lower levels we can see that London and the South West are falling, but the South East and the East look to be rising slowly or levelling off.
But overall a good picture for the trends in new hospital admissions across the country.
Deaths are continuing to increase (as we knew they would), but there are signs their increase is slowing.
Weekly data from the ONS is the gold standard (based on Death certificate), but is lagged.
Shows that deaths across all four nations are no longer doubling every two weeks.
For a more up-to-date picture, these are figures from PHE. They also show deaths continuing to rise (more obvious once the fluctuations are averaged over using the 7-day average), but at a slower rate. No longer doubling every two weeks.
Just because these large numbers of deaths are happening now at a time when cases are coming down, this shouldn't in anyway diminish their importance.
We are seeing the highest rates of COVID deaths now since early May.
The worst thing is that we knew it was coming too.
So while it's good news that lockdown has brought cases and hospitalisations down.
Earlier action would have prevented many of these unnecessary deaths.
It's a lesson we surely have to learn if we are to keep COVID at bay.
/ENDS
• • •
Missing some Tweet in this thread? You can try to
force a refresh
In light of the announcements about Xmas, many people are asking me whether meeting up with friends and family over the holidays is 'safe'.
The honest (and unpopular) answer is that, simply because the government have sanctioned it, that doesn't make meeting up indoors safe.
In these situation there is only managed and mitigated risk.
If you are going to meet up with relatives then it's best to try to meet outdoors, or, if indoors, to try to ventilate the space as well as possible and take other precautions like wearing masks and social distancing.
If you can pay for a test before going to see loved ones then this may help, but being aware that you can test negative and still be incubating the virus is important.
Isolation beforehand might also help to reduce risk, but the length of isolation changes its effectiveness.
Our letter in the Lancet today - outlines key recommendations for a sustainable COVID-19 strategy within the UK. thelancet.com/journals/lance…
Thanks to @dgurdasani1 for pulling this all together.
In the UK we are currently in the middle of the second wave of the epidemic with cases over 20K per day and deaths averaging over 300 per day.
We are undergoing a second lockdown which is indicative of a failure of public health strategy.
We need to use the time bought by this lockdown wisely in order to prevent ourselves from being in the same situation in a few months time.
Here are our recommendations: 1. Urgently reform the find, test, trace, isolate and support system.
A summary of the slides from today's @IndependentSage briefing.
Slides prepared by the brilliant @chrischirp.
First up testing has stayed relatively flat for the last couple of weeks. Not ramping up like we might hope. Only rising Significantly in NI.
The proportion of tests being returned promptly is dropping in almost all settings (bad) but rising in care homes (good).
Confirmed cases still rising steeply. Cases delayed by Excelgate earlier in the week have been spread evenly over the days when data was missing.
I've noticed a consistent trend in people who like to pretend that all is well on the COVID front plotting graphs in a way which suits their purpose. This is the sort of graph they will show you for hospital admissions. Looks pretty good.
It's understandable they want to demonstrate that cases are low, but if you zoom in, you see a small, but appreciable rise in admissions over the last couple of weeks.
Here's the same trick but for patients curently in hospital. Looks like a rosy picture on this graph...
Antibody tests have the potential to be a game-changer. But many of them aren't all they're cracked up to be. In this article (and this thread) I explore at the impact of false positve and false positives in COVID-19 tests. theconversation.com/coronavirus-su…@ConversationUK@UniofBath 1/10
Cellex have had a test approved by the FDA. If you have antibodies against COVID-19 their test will tell you this correctly 93.8% of the time (this is the test’s “sensitivity”). If you don’t, it will get this correct 95.6% of the time (this is the test’s “specificity”). 2/10
Getting the correct result more than 90% of the time sounds pretty good. But there's a catch. To find out why lets consider testing 10,000 hypothetical individuals. The WHO suggested recently that as few as 3% of the global population may have had COVID-19 and recovered. 3/10