Page 16 of 27
Re: Coronavirus - Numbers and Statistics
Posted: November 26th, 2021, 8:43 pm
by jfgw
The rate of increase in new cases is going down for school children. Case rates are looking better for older people, probably (I suggest) due to the boosters.
One thing that concerns me about these data is that they do not include reinfections. This is likely to be more of an issue if a new variant more effectively circumvents existing naturally acquired immunity. Hopefully, "New Cases" will be redefined if this becomes significant.
My graph. Contains public sector information licensed under the Open Government Licence v3.0. : https://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/Julian F. G. W.
Re: Coronavirus - Numbers and Statistics
Posted: December 6th, 2021, 10:02 pm
by jfgw
Cases do not include reinfections where people tested positive previously. I think that this needs to change, especially if reinfections by the Omicron variant become significant.
My graph. Contains public sector information licensed under the Open Government Licence v3.0. : https://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/All age groups under 55 are showing a sharp rise over the previous few days.
55—59s are showing a less pronounced rise.
60—79s are still falling.
80+ age ranges are not so clear, probably due to the low numbers involved.
I do not think that the boosters alone are sufficient to explain the differences in the age groups, especially the 60s.
My graph. Contains public sector information licensed under the Open Government Licence v3.0. : https://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/Julian F. G. W.
Re: Coronavirus - Numbers and Statistics
Posted: December 6th, 2021, 11:13 pm
by servodude
jfgw wrote:Cases do not include reinfections where people tested positive previously. I think that this needs to change, especially if reinfections by the Omicron variant become significant.
Hi Julian
Is this from the cases description at
https://coronavirus.data.gov.uk/metrics/doc/newCasesBySpecimenDateCases definition
COVID-19 cases are identified by taking specimens from people and testing them for the SARS-CoV-2 virus. If the test is positive this is referred to as a case. Some positive rapid lateral flow test results are confirmed with lab-based polymerase chain reaction (PCR) tests taken within 72 hours. If the PCR test results are negative, these are no longer reported as confirmed cases. If a person has more than one positive test, they are only counted as one case. Cases data includes all positive lab-confirmed polymerase chain reaction (PCR) test results plus, in England, positive rapid lateral flow tests that are not followed by a negative PCR test taken within 72 hours.
- as when I read that back in the mists of time I presumed it was so that immediate corroborating re-takes didn't skew that days figures
But if it applies to historical tests also it's a bit of a big red flag!
The deaths within 28 days of one's first test figure has been raised before as a misleading metric
- 10,515,239 cumulative cases in the UK to date most of whom can never now official die of covid
- but at least there are other data streams to qualify that
- sd
Re: Coronavirus - Numbers and Statistics
Posted: December 6th, 2021, 11:36 pm
by jfgw
servodude wrote:Is this from the cases description at
https://coronavirus.data.gov.uk/metrics/doc/newCasesBySpecimenDateCases definition
COVID-19 cases are identified by taking specimens from people and testing them for the SARS-CoV-2 virus. If the test is positive this is referred to as a case. Some positive rapid lateral flow test results are confirmed with lab-based polymerase chain reaction (PCR) tests taken within 72 hours. If the PCR test results are negative, these are no longer reported as confirmed cases. If a person has more than one positive test, they are only counted as one case. Cases data includes all positive lab-confirmed polymerase chain reaction (PCR) test results plus, in England, positive rapid lateral flow tests that are not followed by a negative PCR test taken within 72 hours.
- sd
Yes. Note also the definition of cumulative cases,
Total number of cases (people who have had at least one positive COVID-19 test result) since the start of the pandemic. Data are shown by the date the sample was taken from the person being tested.
https://coronavirus.data.gov.uk/metrics/doc/cumCasesBySpecimenDate
Julian F. G. W.
Re: Coronavirus - Numbers and Statistics
Posted: December 6th, 2021, 11:49 pm
by servodude
jfgw wrote:servodude wrote:Is this from the cases description at
https://coronavirus.data.gov.uk/metrics/doc/newCasesBySpecimenDateCases definition
COVID-19 cases are identified by taking specimens from people and testing them for the SARS-CoV-2 virus. If the test is positive this is referred to as a case. Some positive rapid lateral flow test results are confirmed with lab-based polymerase chain reaction (PCR) tests taken within 72 hours. If the PCR test results are negative, these are no longer reported as confirmed cases. If a person has more than one positive test, they are only counted as one case. Cases data includes all positive lab-confirmed polymerase chain reaction (PCR) test results plus, in England, positive rapid lateral flow tests that are not followed by a negative PCR test taken within 72 hours.
- sd
Yes. Note also the definition of cumulative cases,
Total number of cases (people who have had at least one positive COVID-19 test result) since the start of the pandemic. Data are shown by the date the sample was taken from the person being tested.
https://coronavirus.data.gov.uk/metrics/doc/cumCasesBySpecimenDate
Julian F. G. W.
Thanks
That does seem "quite important" to bear in mind
I wonder if it's considered in their prevalence stats?
- sd
Re: Coronavirus - Numbers and Statistics
Posted: December 7th, 2021, 7:56 pm
by jfgw
servodude wrote:I wonder if it's considered in their prevalence stats?
By definition, it wouldn't. However, I haven't seen any government prevalence stats. The "7-day" case rate data are new cases, not prevalence.
The ONS and ZOE publish prevalence data.
Prevalence data are not perfect; It is possible to test positive after having fully recovered.
Patients who have recovered from COVID-19 can continue to have detectable SARS-CoV-2 RNA in upper respiratory specimens for up to 3 months after illness onset. However, replication-competent virus has not been reliably recovered and infectiousness is unlikely.
https://www.cdc.gov/coronavirus/2019-ncov/hcp/duration-isolation.htmlJulian F. G. W.
Re: Coronavirus - Numbers and Statistics
Posted: December 10th, 2021, 9:43 pm
by jfgw
The December monthly NHS publication has just been published.
https://www.england.nhs.uk/statistics/w ... -2021.xlsxI have plotted the admissions data for the different age groups.
My graph. Contains public sector information licensed under the Open Government Licence v3.0. : https://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/Julian F. G. W.
(Boosted today!)
Re: Coronavirus - Numbers and Statistics
Posted: December 19th, 2021, 8:22 pm
by jfgw
Increases in cases seem to be starting in the 20—39 age groups, then affecting other ages.
Things are only just starting to happen in the North East,
There is a bit more happening in Yorkshire and The Humber, the West Midlands, and the South West,
The trend continues,
London cases rising rapidly,
All images: My graph. Contains public sector information licensed under the Open Government Licence v3.0. :
https://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/Julian F. G. W.
Re: Coronavirus - Numbers and Statistics
Posted: December 19th, 2021, 10:24 pm
by GrahamPlatt
Difficult to decipher, those graphs. e.g. that last one, London cases; the third highest line seems to have a green + sign as identifier, but I can’t see that in the legend. Is it 30-34?
Re: Coronavirus - Numbers and Statistics
Posted: December 19th, 2021, 10:29 pm
by servodude
GrahamPlatt wrote:Difficult to decipher, those graphs. e.g. that last one, London cases; the third highest line seems to have a green + sign as identifier, but I can’t see that in the legend. Is it 30-34?
yes that's how I read it
- i think the horizontal bar of the cross gets hidden in the legend by the line
- sd
Re: Coronavirus - Numbers and Statistics
Posted: December 19th, 2021, 11:07 pm
by jfgw
GrahamPlatt wrote:Difficult to decipher, those graphs. e.g. that last one, London cases; the third highest line seems to have a green + sign as identifier, but I can’t see that in the legend. Is it 30-34?
Yes, it is 30—34.
I have wondered about merging age groups but it would mean losing information as well as requiring the effort of looking up populations.
I think that the London graph is worth keeping an eye on.
Julian F. G. W.
Re: Coronavirus - Numbers and Statistics
Posted: December 21st, 2021, 4:37 pm
by jfgw
If you look at a short enough timeframe, there is evidence of an impending peak among the most heavily infected age groups in London. I think this is one to keep an eye on for the next few days (until Christmas disrupts the flow of data) to see which way it continues.
My graph. Contains public sector information licensed under the Open Government Licence v3.0. : https://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/Julian F. G. W.
Re: Coronavirus - Numbers and Statistics
Posted: December 22nd, 2021, 10:10 pm
by jfgw
CORRECTED GRAPHLondon cases by age. The previous graph showed absolute case numbers for the 80+ age groups. I have corrected this and included another day's data. My apologies for any confusion. This error exists in my other "cases by age" spreadsheets — I will correct these.
My graph. Contains public sector information licensed under the Open Government Licence v3.0. : https://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/Julian F. G. W.
Re: Coronavirus - Numbers and Statistics
Posted: December 24th, 2021, 5:11 pm
by MrFoolish
Maths question.
If 1 in 35 people have covid, how many people do you need in a room for it to be more likely than not that someone has covid?
Re: Coronavirus - Numbers and Statistics
Posted: December 24th, 2021, 5:22 pm
by monabri
390 million tests carried out in the UK. Where has the waste gone?
Re: Coronavirus - Numbers and Statistics
Posted: December 24th, 2021, 5:24 pm
by CryptoPlankton
MrFoolish wrote:Maths question.
If 1 in 35 people have covid, how many people do you need in a room for it to be more likely than not that someone has covid?
24 (assuming everyone is equally likely to have it - a big assumption!)
Re: Coronavirus - Numbers and Statistics
Posted: December 24th, 2021, 5:34 pm
by MrFoolish
CryptoPlankton wrote:MrFoolish wrote:Maths question.
If 1 in 35 people have covid, how many people do you need in a room for it to be more likely than not that someone has covid?
24 (assuming everyone is equally likely to have it - a big assumption!)
I've long since forgotten my stats. How do you work it out?
Re: Coronavirus - Numbers and Statistics
Posted: December 24th, 2021, 5:49 pm
by dealtn
MrFoolish wrote:CryptoPlankton wrote:MrFoolish wrote:Maths question.
If 1 in 35 people have covid, how many people do you need in a room for it to be more likely than not that someone has covid?
24 (assuming everyone is equally likely to have it - a big assumption!)
I've long since forgotten my stats. How do you work it out?
Assuming there is true independence any person is 34/35 probability free of Covid. So ask the question how many times do you need to multiply 34/35 by itself to be less than 0.5. That is the chance of at least 1 person being infected is greater than the chance all are free of Covid.
34/35 to the power of 24 is just below 0.5
Re: Coronavirus - Numbers and Statistics
Posted: December 24th, 2021, 5:50 pm
by Mike4
CryptoPlankton wrote:MrFoolish wrote:Maths question.
If 1 in 35 people have covid, how many people do you need in a room for it to be more likely than not that someone has covid?
24 (assuming everyone is equally likely to have it - a big assumption!)
Also to muddy the waters it depends on what is meant by '
has covid', I'd have thought. "
Infectious with covid" is what I'd be most interested in, were I in the room.
People in the earliest stages of incubating it are not necessarily infectious yet, and those who are in late stages of recovery might still return positive PCTR tests even though not infectious.
Re: Coronavirus - Numbers and Statistics
Posted: December 24th, 2021, 6:15 pm
by MrFoolish
dealtn wrote:MrFoolish wrote:CryptoPlankton wrote:24 (assuming everyone is equally likely to have it - a big assumption!)
I've long since forgotten my stats. How do you work it out?
Assuming there is true independence any person is 34/35 probability free of Covid. So ask the question how many times do you need to multiply 34/35 by itself to be less than 0.5. That is the chance of at least 1 person being infected is greater than the chance all are free of Covid.
34/35 to the power of 24 is just below 0.5
Thanks. I tried raising 1/35 to the power of 24 and obviously just ended up with a nonsense tiny number. It didn't occur to me to try something with 34/35. Is there a name for this sort of calculation?