• Dear Guest, and welcome to RailUK Forums. Our non-railway discussion forums are currently restricted until members have five or more posts, and you will not be able to make a new thread or reply to an existing one in this section until you have made five or more posts elsewhere on the forum.

How many people have symptomatic Covid in the UK right now?

Status
Not open for further replies.

yorkie

Forum Staff
Staff Member
Administrator
Joined
6 Jun 2005
Messages
55,832
Location
Yorkshire

218,694​

people are currently predicted to have symptomatic COVID in the UK
However..

Experts warn that infection levels remain high however, with about 553,000 people found to have the virus.
The way this is worded is confusing; are they saying 553,000 have the virus right now?

From the same page:

One in 115 people in England has the virus (compared to one in 80 the week before)

So by that measure it is estimated that around 489,000 have the virus, in England alone.

How does this add up?

Is the Zoe app massively underestimating? Or are other measures massively overestimating?

Or do over 550k people have the virus but only around 40 per cent are actually symptomatic?

The Government tell us only a third of people are asymptomatic; is that a huge underestimate and the true figure is around 60 per cent?

Or is there something else going on that I can't figure out?
 
Sponsor Post - registered members do not see these adverts; click here to register, or click here to log in
R

RailUK Forums

Bayum

Established Member
Joined
21 Mar 2008
Messages
2,508
Location
Leeds

However..


The way this is worded is confusing; are they saying 553,000 have the virus right now?

From the same page:



So by that measure it is estimated that around 489,000 have the virus, in England alone.

How does this add up?

Is the Zoe app massively underestimating? Or are other measures massively overestimating?

Or do over 550k people have the virus but less than 50 per cent are actually symptomatic?

The Government tell us only a third of people are asymptomatic; is that a huge underestimate?
Don’t forget that the numbers of people using the Zoë app will be very different to those being tested in the community. Particularly where you have users who have slowly begun to stop reporting signs and symptoms regularly.
The ONS statistics are based on representative samples - out of x numbers y were positive, a were symptomatic and b asymptomatic - scale up using some very clever maths and you have the various ratios presented.
 

yorkie

Forum Staff
Staff Member
Administrator
Joined
6 Jun 2005
Messages
55,832
Location
Yorkshire
Don’t forget that the numbers of people using the Zoë app will be very different to those being tested in the community. Particularly where you have users who have slowly begun to stop reporting signs and symptoms regularly.
The ONS statistics are based on representative samples - out of x numbers y were positive, a were symptomatic and b asymptomatic - scale up using some very clever maths and you have the various ratios presented.
Maybe but the Zoe team try to scale up their figures to estimate for the whole population; they have been incredibly accurate at detecting changes in prevalence and so I would say their data is likely to be highly representative.

I note that they are estimating symptomatic Covid whereas the Government figures are estimating prevalence of the Sars-CoV-2 virus which causes Covid in some (but by no means all) people.

Hence I'm wondering if the discrepancy could be due to a higher proportion of people being asymptomatic than the Government have stated.
 

Domh245

Established Member
Joined
6 Apr 2013
Messages
8,082
Location
nowhere
Don’t forget that the numbers of people using the Zoë app will be very different to those being tested in the community. Particularly where you have users who have slowly begun to stop reporting signs and symptoms regularly.
The ONS statistics are based on representative samples - out of x numbers y were positive, a were symptomatic and b asymptomatic - scale up using some very clever maths and you have the various ratios presented.

ONS just works on random PCR samples and does not do any sort of symptomatic breakdown. ZOE extrapolate their data out to be representative of the UK, but decreasing numbers just means larger confidence intervals. Of note though is that as more people get vaccinated, they are getting more and more false-flags where people with post-vaccine effects are falsely indicating as symptomatic covid, but a fix for this is coming in the next couple of days.

The Government tell us only a third of people are asymptomatic; is that a huge underestimate and the true figure is around 60 per cent?

I think this is the closest to reality, although no doubt it's a combination of factors. The 1/3 asymptomatic seems like it may be on the low side, although worth noting that it seems difficult to pin a number on this, particularly as it depends on asymptomatic v presymptomatic. If I were being cynical I would say that the government would deliberately try to aim for the lower side of the range together with the fear-focused messaging ("Gosh, 2/3rds of people getting covid suffer horrific symptoms, better hermit up") but that's a little tin-foily.

It's certainly not out of the question that the ZOE data is a bit dubious too. As I noted they're currently having an issue with overrepresentation because of vaccine effects, and I've never seen any confidence intervals so quality of the data isn't guaranteed

The other thing that just pops into my mind is if they're using the same definitions of symptomatic - ZOE have been pushing for a wider definition recently IIRC and have generally done a lot of work trying to identify symptoms, so it's also quite plausible that they're including some people as symptomatic covid that the goverment wouldn't consider symptomatic going by the big 3 (fever, cough, taste)
 

yorkie

Forum Staff
Staff Member
Administrator
Joined
6 Jun 2005
Messages
55,832
Location
Yorkshire
...ZOE have been pushing for a wider definition recently IIRC and have generally done a lot of work trying to identify symptoms, so it's also quite plausible that they're including some people as symptomatic covid that the goverment wouldn't consider symptomatic going by the big 3 (fever, cough, taste)
Yes, I have seen Tim's updates on the Zoe app YouTube channel, but wouldn't that mean the gap is even wider?

I do remember hearing about cruise ships where huge numbers of people were asymptomatic and so I am wondering if the number of asymptomatic cases has been massive underestimated; this could certainly explain this discrepancy.
 

philosopher

Member
Joined
23 Sep 2015
Messages
749
The way this is worded is confusing; are they saying 553,000 have the virus right now?

From the same page:


So by that measure it is estimated that around 489,000 have the virus, in England alone.

How does this add up?

Is the Zoe app massively underestimating? Or are other measures massively overestimating?

Or do over 550k people have the virus but only around 40 per cent are actually symptomatic?

The Government tell us only a third of people are asymptomatic; is that a huge underestimate and the true figure is around 60 per cent?

Or is there something else going on that I can't figure out?

The REACT study yesterday estimated that 1 in 196 were infected, which would be approx. 350,000 for the entire UK. So this figure is between the ONS and Zoe estimates.

One of the largest and most authoritative coronavirus surveys has found that infections are quickly falling in England, confirming that lockdown is working to suppress the virus.

Imperial College London's REACT study found that infections had fallen by more than two-thirds since the last time it reported in mid-January.


Last time, REACT found that 1 in 63 people currently had the virus. This time, it estimates that 1 in 196 people are infected.

The odd thing is that the REACT study seems to like the ONS be doing random sampling PCR tests, so I would have thought it would get prevelance results very similar to the ONS, but this does not appear to be the case.
 

brad465

Established Member
Joined
11 Aug 2010
Messages
3,285
Location
Maidstone
The REACT study yesterday estimated that 1 in 196 were infected, which would be approx. 350,000 for the entire UK. So this figure is between the ONS and Zoe estimates.



The odd thing is that the REACT study seems to like the ONS be doing random sampling PCR tests, so I would have thought it would get prevelance results very similar to the ONS, but this does not appear to be the case.
Ironically, I think the REACT study is the least reactive of the lot, partly in the sense it's the slowest, but also because it doesn't extrapolate I think as much as ZOE and the ONS do. When it comes to trustworthiness and reliability I rank measures in the following order:

1) ONS
2) ZOE
3) Actual confirmed cases
4) REACT
 

Mojo

Forum Staff
Staff Member
Administrator
Joined
7 Aug 2005
Messages
18,711
Location
0035
ONS just works on random PCR samples and does not do any sort of symptomatic breakdown.
As part of the ONS survey they do however ask participants whether they are suffering any symptoms (and ask them to say what from a list) - so the data would exist as they are collecting it.
 

Bayum

Established Member
Joined
21 Mar 2008
Messages
2,508
Location
Leeds
As part of the ONS survey they do however ask participants whether they are suffering any symptoms (and ask them to say what from a list) - so the data would exist as they are collecting it.
This.
 

Domh245

Established Member
Joined
6 Apr 2013
Messages
8,082
Location
nowhere
Yes, I have seen Tim's updates on the Zoe app YouTube channel, but wouldn't that mean the gap is even wider?

I do remember hearing about cruise ships where huge numbers of people were asymptomatic and so I am wondering if the number of asymptomatic cases has been massive underestimated; this could certainly explain this discrepancy.

I was thinking that it'd explain the difference between the governments 33% and the 60% that the figures seem to indicate. Using the 1in3 figure and the ONS prevalence you'd expect ~184,000 people to have the big-3 symptoms, and then there are a further 34000 people with some other symptom that ZOE uses and Gov doesn't.

There's a CEBM article from early in the pandemic which puts the asymptomatic figure between 5% and 80%!

As part of the ONS survey they do however ask participants whether they are suffering any symptoms (and ask them to say what from a list) - so the data would exist as they are collecting it.

I had vaguely remembered that they did but checking the latest ONS release there wasn't any mention of this. Checking again this morning, it turns out they publish this but out of sync with their prevalence study. Worth a read through, but highlighting the key points:

  • In England, 47% of people who tested positive for COVID-19 (95% confidence intervals: 46% to 48%) reported having any symptoms.
  • In Wales, 55% of people who tested positive for COVID-19 (95% confidence intervals: 49% to 62%) reported having any symptoms.
  • In Northern Ireland, 38% of people who tested positive for COVID-19 (95% confidence intervals: 31% to 46%) reported having any symptoms.
  • In Scotland, 47% of people who tested positive for COVID-19 (95% confidence intervals: 41% to 54%) reported having any symptoms.
  • Cough, fatigue and headache were the most commonly reported symptoms from people who tested positive between 1 October 2020 and 30 January 2021 for all four countries of the UK.
  • Abdominal pain, diarrhoea and nausea were less commonly reported symptoms.

Which goes to show both how variable asymptomatic-ness is, and that the government's figure doesn't quite match up!
 

yorkie

Forum Staff
Staff Member
Administrator
Joined
6 Jun 2005
Messages
55,832
Location
Yorkshire
Good analysis again @Domh245 and yes, that makes perfect sense.

As I was saying to someone a while back I did think the proportion of asymptomatics may be a lot higher than claimed and the Government probably likes to say a third because it is high enough to make people think there is a realistic chance they could be unknowingly infected but low enough to make people think the majority of people 'suffer' from it, thus being the optional proportion to spread fear.

Maybe I am just being cynical though?

But I think at this stage it would be good news if 60 per cent of people are asymptomatic as it means the virus is less severe than previously thought

Indeed I think it's probably going to be no more severe than any of the existing seasonal Coronaviruses,the only difference probably being our immunity levels rather than anything relating to the virus itself.

But that's not based on any actual research and is just a guess
 

Domh245

Established Member
Joined
6 Apr 2013
Messages
8,082
Location
nowhere
As I was saying to someone a while back I did think the proportion of asymptomatics may be a lot higher than claimed and the Government probably likes to say a third because it is high enough to make people think there is a realistic chance they could be unknowingly infected but low enough to make people think the majority of people 'suffer' from it, thus being the optional proportion to spread fear.

Maybe I am just being cynical though?

But I think at this stage it would be good news if 60 per cent of people are asymptomatic as it means the virus is less severe than previously thought

I had a similarly cynical thought last night!

It would be good news if it was that high, but I suspect that asymptomaticness is closely linked to age and other underlying factors (as is tending to be the case with most other things covid) and so just tends to be reflective of the current distribution of infections rather than anything else. The fact the UK currently is something like 60% asymptomatic might just indicate the bulk of infections being among the younger population less susceptible to have symptoms, but that is just a hunch though
 
Status
Not open for further replies.

Top