Roy Motteram ☂️🎪💙 Profile picture
Jul 4, 2021 25 tweets 7 min read Read on X
I heard someone today assert that we needn't worry too much about the rising number of Covid hospital admissions, because they're only staying in for a day.

That's interesting, I thought, & reassuring if true.

So I looked into it.
1/7
2/7
I couldn't find any stats on length of stay, so the assertion appears to be based on anecdote.

But what I did find is not quite so reassuring.

The number of Covid patients in hospital are not climbing as steeply as infections, but they ARE climbing
3/7
So I thought I'd look further, & see how many of those Covid patients are severely ill. The only measure of this that I have is the number on ventilators.

After all, they don't put a patient on a ventilator for the lolz.
4/7
OK. That's not so terrible. The "red" Mechanically Ventilated patients is much smaller. It's climbing, but not much.

But then I wondered if the scale was hiding a problem
5/7
So I looked at those same figures but as a %age.

After all, if the vaccination programme is reducing the severity of infections, then we should see a decline in the %age of hospitalised patients getting as far as needing ventilation.

I don't see a drop. Do you?
6/7
Actually, I do see a decline from the beginning of April until the 12th or 13th of May.

Then I see it climbing again.

From memory, that's more or less the time that the Delta variant took over.
I'm no epidemiologist. I'm no expert at all.

My only skill is the ability to look at data & produce some graphs.

But ignore anyone telling you that we needn't worry about the hospitalisation figures because "they're only in for a day"

The data tells a different story
/End
Thread extension:
I've had a number of helpful comments from people I know & whose opinion I value, as well as from people I don't know well (yet).

A number of you have identified that this is a very simplistic & limited view of the data ...
But remember ... I was only interested in the "you can ignore the hospitalisation statistic, because they're only in for a day" narrative, not a learned analysis

We know that there are a number of factors that will impact the %age of patients that need mechanical ventilation ...
For example, the older the patient, the more likely they will be severely ill & need ventilation.

On the other hand, if the patient has been vaccinated, they are less likely to get very ill.
Counterintuitively, if the NHS can get patients back on their feet & out of the door more quickly, then the patients staying longer are the ones needing ventilation, driving the %age up

I don't believe there are any public stats on duration of hospital stay
Having said all of that, I took your advice, & found the data (not very difficult) for earlier dates, & this is the coresponding chart, make of it what you will:
Bear in mind that I was looking into the argument that we should ignore hospital admissions, because stays are short. But if some patients are able to be discharged earlier, then as stated, this might push the %age on ventilators up ...
So, while the data might be far from reassuring, it doesn't directly address the question.

Should we be worried about the Hospital Admissions statistic?

I tried another approach, superimposing the number of admissions with the number on ventilators
I'm still not reassured?

Are you?

For those that didn't come in at the top of this thread, the data can be found here: england.nhs.uk/statistics/sta…
So I thought I'd look at a comparison of the number of patients in hospital versus the number of patients admitted over the previous "n" days. If the avg stay is "n" days, then patients admitted "n+1" days ago will no longer be patients, but all admitted since then still will be
If we start with 12 days, we get a picture that shows, on average, since mid Aug 2020, Covid patients have been staying in hospital for fewer than 12 days
If we add the line showing the preceding 10 days of admissions to that chart, we see that at times the average has been above 10 days, & at other times, it has been less
Adding a line for the preceding 8 days of admissions shows us that during the peaks of the 2nd & 3rd waves last autumn & winter, the average stay was mostly between 8 & 10 days
Adding a line for the preceding 6 days of admissions, & focussing in on the timeframe from 1st April, we see that during the tail of the 3rd wave, avg stay was still between 10 & 12 days, but now, as the 4th wave is kicking off, the average stay has dropped to between 6 & 8 days
Why might this have happened?

For 1 thing, more of us are vaccinated, which, according to the scientists, means that if we get Covid, the symptoms will, in general, be less severe.
The second reason why the average stay might be reducing is that the average age of patients is also dropping (probably also due to the effect of the vaccination of us old folk)

Apologies for the lairy colours
That yellow section of 18-54 is quite a broad range, but it's the only data I have.

What I find quite worrying is the relatively small but growing section of 0-5 & 6-17 yr olds.

Particularly given the growing evidence that long covid is a problem we should be worried about
So ... my final conclusion

Yes, the average hospital stay has definitely come down.

No, I'm not reassured that Covid-19 can now be dismissed as a trivial illness.

Are you?

Many thanks to all of you that prompted me to take the analysis further.
PS: For those of you thinking that comparing avg hospital stay at the beginning of a wave with avg stay during the peak might not be sound, this is what the picture is for the beginning of the 2nd wave, when numbers were similar to today.

Between 6 & 8, just as it is now. 🙄

• • •

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

Keep Current with Roy Motteram ☂️🎪💙

Roy Motteram ☂️🎪💙 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 @RoyMotteram

Oct 30, 2022
I had an interesting conversation yesterday with a group of #Tory voters about the #NHS.

The consensus was that "we can't afford to keep pouring money into the NHS. It's inefficient, & needs a good manager to wade in & sort it out."

A 🧵 1/11
They went on to discuss nurses (mixed reviews between very caring, & spending their shift at the nurses' station huddled around a computer, rather than being on the ward), & the difficulty of getting a face-to-face meeting with a GP.
2/11
They also compared the NHS with healthcare in France, where apparently there is a similar shortage in GPs, but no problem in arranging an appointment.

The general view was that healthcare in France is superior to that in the UK.
3/11
Read 11 tweets
Mar 2, 2019
Mini thread
I was volunteering at a stall this afternoon.

I like working on People's Vote stalls. You get to hear all sorts of stories & perspectives, from both sides of the divide.

1/8
Today, I spoke to a leave voter that couldn't wait to get out. He was 70, a retired businessman in the cosmetics industry who traded world-wide, & didn't like the EU.

With EU regulations on cosmetics safety being amongst the toughest in the world, I can imagine why.

2/8
I asked him to look at what Patrick Minford said would happen to our manufacturing industry & agriculture (he'd not heard of Minford, nor his expectations).

His wife promised that she would make sure he did.

3/8
Read 9 tweets
Jan 31, 2019
Like many, I signed a petition calling for a referendum on Mrs May's proposed Brexit, & today I received the response from the govt.
For those that didn't sign up to this particular petition, I reproduce it here, with my response
My MP, @JHowellUK probably hasn't seen it
1/14
The response came from the Department for Exiting the European Union.

It starts by insisting that the govt will "respect the result of the 2016 ref"
This is such a key point that they say it twice, using "clear" on both occasions to emphasise it.
2/14
But look at that 3rd paragraph.
Apparently the 2016 ref gave a very clear instruction to the govt.
If it was so clear, why is there so much discord amongst Leave voters? Not one of those I met outside Parliament on 29/01 were happy
3/14
Read 14 tweets
Aug 6, 2018
#MondayMotivation
I recently published a thread looking at #LeaveLies. Not surprisingly, I got responses that "both sides lied". After a bit of to & fro, @Jamiepembs kindly sent me some links which demonstrated #RemainLies
1/18
The 1st link was, on the surface, a little disappointing. It appeared to just make the assertion that "Campaigners on both sides of the EU Referendum made false claims" with no evidence of either #RemainLies or #LeaveLies
2/18
If you look carefully in that page, you'll see a link to a comparison of the Vote Leave & Stronger In leaflets. The 1st thing I noted was that the Leave list was longer, but I'm here to look into #RemainLies. This is an image, but the individual links are included below 3/18
Read 18 tweets
Aug 1, 2018
I thought I'd have a look at the FB targeted ads that have recently been published. We know that VL told porkies, but I wanted to see how much of their campaign was untruthful.
It seems that it was pretty much all of it. #VoteLeaveLied
Here's a sample of the worst offenders 1/17
We know that they targeted groups with their own special trigger messages & we know that they lied about the EU budget contribution (gross of £225m/wk, net £203m/wk), but £350m wasn't just on the bus ... it was pretty much everywhere in their ads 2/17
It started in their data capture con they used to get your information (Trump had Kogan & CA, VL used their £50m competition)
£50m/day = £350m/wk 3/17
Read 17 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

Don't want to be a Premium member but still want to support us?

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

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us!

:(