Share this page!

John Roberts Profile picture
Twitter logo
20 Oct, 4 tweets, 2 min read
A release from the CDC that I suspect one presidential candidate will find rather untimely. 300,000 excess deaths, a third of which are not directly attributed to COVID - that's a higher proportion than in the UK. 1/4 cdc.gov/mmwr/volumes/6…
Media comment already noting that the 25-44 group has been hit hardest. Maybe in terms of % increase, but given very low death rates at younger ages, I'm not sure I'd agree with that view. 2/4 Image
The disparity by ethnicity that we see here is repeated, but just looking at the graphs even more so. 3/4 Image
Coming just two weeks before the election, I can't imagine the Trump campaign will be pleased with the body the administration sidelined early on in the pandemic producing such a set of figures. It'll be interesting to see whether it has an effect on the final days' debate. 4/4

• • •

Missing some Tweet in this thread? You can try to force a refresh
　

Keep Current with John Roberts

John Roberts Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @john_actuary

John Roberts Profile picture
22 Oct
The latest weekly PHE Flu & COVID Surveillance Report is out. assets.publishing.service.gov.uk/government/upl… At first glance there's been a welcome flattening off of both cases and infectivity, but the age split shows that at older ages there are still increases. Case rates first.... 1/5 Image
Now infectivity levels, for pillar 2 (community testing). Most are still rising steeply, except for younger ages. 2/5 Image
Admissions next, using rates per 100k, to allow for differing populations by region. It's still consistent with the picture @COVID19actuary has been sharing for some time now. The age split shows the concern over older age groups. 3/5 ImageImage
Read 5 tweets
John Roberts Profile picture
1 Oct
The @imperialcollege #REACT Survey has published an interim report of its 5th round. The headline is that it suggests R has fallen sharply in recent weeks, and could now be around 1.1 . Let's take a more detailed look... 1/9
The interim report covers 85,000 samples over the period 18th-26th Sept, so is very up to date. It found 363 positive samples, an infectivity of 0.55% (over 1 in 200) , consistent with the latest @ONS report. That's the highest since the survey started in May. 2/9
Round 4 put R at 1.7 for late Aug/early Sep. These results show a clear slowing, and the report notes other indicators (eg calls to the NHS) are consistent with this direction. The CI is wider though as this is an interim report, so the sample size is reduced. 3/9
Read 9 tweets
John Roberts Profile picture
28 Sep
The @ONS is producing some terrific analysis regarding COVID-19. Today's release provides some maybe predictable, and some maybe surprising, findings regarding the recent spread. Let's start with deprivation. 1/7 Image
We know that in the first wave, the virus took a greater toll on those from more deprived areas, both in terms of incidence and outcomes. But there's a very clear reverse now for incidence, with the most affluent areas driving the increase. 2/7
Digging a bit deeper, it's also clear that it is those under 35 and are more affluent for whom infectivity is rising quickest. For over 35's there's not much difference by deprivation. 3/7 Image
Read 7 tweets
John Roberts Profile picture
25 Sep
The latest admission for England has topped 300 today, and again a question raised is are these "WITH" COVID or "FROM" COVID? Just like yesterday's analysis of deaths, let's dig into this question a bit... 1/7
First of all, note that any elective admissions, ie for planned surgery, is preceded by a COVID test. If it's positive, the admission doesn't happen. So the only cases that might by chance test positive are emergency admissions for what are believed to be non COVID reasons. 2/7
There are approx 15k such admissions a day. What proportion of them might test positive, yet be totally unrelated? The latest ONS figures give a clue, suggesting that 1 in 500 now has the virus. That would suggest 30, out of the 300 cases. 3/7
Read 7 tweets
John Roberts Profile picture
24 Sep
A common question by those who may feel that the effects of COVID-19 are overstated is to ask how many deaths have been WITH COVID as opposed to being DUE TO COVID.
Here's why I think these questions and underlying beliefs are misguided. 1/6
The actuarial profession, through the CMI, has monitored mortality in the UK for many years, so it's in a good position to understand the level of overall excess mortality we've seen in E&W since March. Here's a graph showing recent years in comparison with each prior year. 2/6 Image
It currently estimates 55,000 excess deaths since March. It uses age standardised mortality rates, so changes in population distribution (such as an ageing pop'n) are allowed for. There's very little reduction since June, which would happen if people had "just died early". 3/6
Read 6 tweets
John Roberts Profile picture
20 Sep
A short thread with some key points on the latest PHE surveillance report issued Friday. I'll start with the age breakdown and trend - the colour coding highlights how the level of infection is gradually spreading to older, more vulnerable, ages. 1/5
The regional split shows a now familiar picture. There's not the same concept of spreading as for age of course. The E/SE/SW regions are not showing the same growth - possibly an argument against national measures where they are particularly damaging economically? 2/5
We know that outcomes are worse for ethnic minorities once infected, but the difference in infectivity levels is startling. The correlation with other factors (eg socio-economic) needs to be understood before jumping to conclusions, but surely more research is needed here? 3/5
Read 5 tweets

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!

Follow Us on Twitter!

Like this author's thread?

Get email notifications when new thread is out from this Author!

Check out other threads from this Author!

Read all threads by @john_actuary