Positives tests among Premier League players/staff

[ClaretTony]

What about anybody he has been in contact with ? Them too?

You would assume that will be done

[nil_desperandum]

What about anybody he has been in contact with ? Them too?

Well theoretically that's just people in his household and as I understand it they have to isolate for 14 days. (He only gets a a 7 day "ban")

[cricketfieldclarets]

6 out of 748 is a worry .
Thats a very high percentage.

It would mean that if that is a representative sample then there are 440,000 people in England infected.

If as some have suggested there were 150,000 infected on any given day last week these figures would suggest footballers have not been following social distancing rules aswell as the general population

Or on the contrary it’s less of a worry as it shows it’s less symptomatic or problematic for more people.

[Roosterbooster]

Hi Rooster, I admit I'm struggling with your accuracy maths. Can you take me through it?

748 tests
98.8% accurate
1.2% inaccurate
1.2% of 748 = 748 x 1.2 / 100 = 897.6/ 100 = 9 inaccurate results

Positive tests are very unlikely to be inaccurate due to the nature of the test. It is a PCR test which essentially matches genetic code. Very unlikely to have genetic code which matches that of the coronavirus, unless you have the virus. But for the sake of fairness, let's assume one of the positive tests in this sample was inaccurate. So 5 people were correctly identified as having the virus, or true positives.

So we have 8 inaccurate tests left to find. These must, by process of elimination, be false negatives. These are much more likely, because the test most places are using are swabs taken from the upper aerodigestive region. i.e. nose or throat. You can have the virus, but not pick up virus, if the swab is not done correctly. Or there just happens to be little virus in this region.

8 false negatives means 8 people had the virus, but had a negative test.
Add in the 5 true positives. 8+5=13. 13 people have the virus.
Only 5 were identified.
The hit rate = true positives/ (true positives + false negatives), aka sensitivity.
5/13 = 38%

My original calculation left the false positive in: 6/14 = 43% (I was being nice)

You could reasonably assume there are no false positives. That leaves 9 false negatives of 15 total cases = 6 true positives. 40%

Either way, its a fairly shoddy test.
This is a disease where we want to pick up every case. i.e a sensitivity close to 100%. This test appears to be nowhere near that

[huw.Y.WattfromWare]

The concern being did he come into contact with any players on Sunday?
It doesn’t have to be physical contact remember. Following a victim through a door with a metal handle might be enough?

[NewClaret]

The concern being did he come into contact with any players on Sunday?
It doesn’t have to be physical contact remember. Following a victim through a door with a metal handle might be enough?

Players haven’t returned to training, have they, so presumably no contact with players?

[Chester Perry]

apparently Woan is asymptomatic

https://twitter.com/RobHarris/status/12 ... 4509370368

Great for his personal wellbeing but this could be difficult because the virus could just hang around on him as in cases we have heard of in Italian football

[NewClaret]

apparently Woan is asymptomatic

https://twitter.com/RobHarris/status/12 ... 4509370368

Great for his personal wellbeing but this could be difficult because the virus could just hang around on him as in cases we have heard of in Italian football

That’s the more worrying aspect - could be one of those folk who test positive for an extended period of time.

[dsr]

748 tests
98.8% accurate
1.2% inaccurate
1.2% of 748 = 748 x 1.2 / 100 = 897.6/ 100 = 9 inaccurate results

Positive tests are very unlikely to be inaccurate due to the nature of the test. It is a PCR test which essentially matches genetic code. Very unlikely to have genetic code which matches that of the coronavirus, unless you have the virus. But for the sake of fairness, let's assume one of the positive tests in this sample was inaccurate. So 5 people were correctly identified as having the virus, or true positives.

So we have 8 inaccurate tests left to find. These must, by process of elimination, be false negatives. These are much more likely, because the test most places are using are swabs taken from the upper aerodigestive region. i.e. nose or throat. You can have the virus, but not pick up virus, if the swab is not done correctly. Or there just happens to be little virus in this region.

8 false negatives means 8 people had the virus, but had a negative test.
Add in the 5 true positives. 8+5=13. 13 people have the virus.
Only 5 were identified.
The hit rate = true positives/ (true positives + false negatives), aka sensitivity.
5/13 = 38%

My original calculation left the false positive in: 6/14 = 43% (I was being nice)

You could reasonably assume there are no false positives. That leaves 9 false negatives of 15 total cases = 6 true positives. 40%

Either way, its a fairly shoddy test.
This is a disease where we want to pick up every case. i.e a sensitivity close to 100%. This test appears to be nowhere near that

How do they work out the 98.8%?

I'm not saying it's wrong, because I don't know the methodology. But what sort of population was the 98.8% based on? If it was based on people suspected of having coronavirus, with say 25% of the people tested having the disease, then it's not a bad record; if it's based on an apparently healthy population like footballers, then like you say it's pretty rubbish. But is it possible that it's based on finding 98.8 out of every 100 positive cases that come before it? Because if it;s that, and it would be a logical way to test, then it's pretty darn good.

I just don't know how they measure the 98.8%.

[Zlatan]

Hi Rooster, I admit I'm struggling with your accuracy maths. Can you take me through it?

The maths is simple Paul 748 x (1-0.988) = just under 9 - it surprised me too.

We can kiss goodbye to any life at all, permanently!

In the entire history of medicine only 1 disease has ever been wiped off the face the Earth - smallpox!

Don’t upset anti-vaxers - telling them a vaccine actually worked

[Zlatan]

748 tests
98.8% accurate
1.2% inaccurate
1.2% of 748 = 748 x 1.2 / 100 = 897.6/ 100 = 9 inaccurate results

Positive tests are very unlikely to be inaccurate due to the nature of the test. It is a PCR test which essentially matches genetic code. Very unlikely to have genetic code which matches that of the coronavirus, unless you have the virus. But for the sake of fairness, let's assume one of the positive tests in this sample was inaccurate. So 5 people were correctly identified as having the virus, or true positives.

So we have 8 inaccurate tests left to find. These must, by process of elimination, be false negatives. These are much more likely, because the test most places are using are swabs taken from the upper aerodigestive region. i.e. nose or throat. You can have the virus, but not pick up virus, if the swab is not done correctly. Or there just happens to be little virus in this region.

8 false negatives means 8 people had the virus, but had a negative test.
Add in the 5 true positives. 8+5=13. 13 people have the virus.
Only 5 were identified.
The hit rate = true positives/ (true positives + false negatives), aka sensitivity.
5/13 = 38%

My original calculation left the false positive in: 6/14 = 43% (I was being nice)

You could reasonably assume there are no false positives. That leaves 9 false negatives of 15 total cases = 6 true positives. 40%

Either way, its a fairly shoddy test.
This is a disease where we want to pick up every case. i.e a sensitivity close to 100%. This test appears to be nowhere near that

To start off, I’m no expert in this but I just want to question the logic a little if I may.

If 98.8% are classed as accurate then that leaves 1.2% as inaccurate. Now inaccurate could be positive or negative as far as I can understand - or am I getting this wrong? If the 1.2% are inaccurate because the swab wasn’t done correctly; or the test doesn’t work properly I suppose would it be classified as invalid as opposed to inaccurate? I hope I’m making sense...!

Ta

[MrTopTier]

Tests positive for Covid 19

[Zlatan]

Has he? Hadn’t heard that...

viewtopic.php?f=2&t=47324

[conyoviejo]

Already a thread mate.

[Zlatan]

Already a thread mate.

I prefer sarcasm...

[MrTopTier]

Apologies, delete this please,I’m off to spec savers

[conyoviejo]

I prefer sarcasm...

Your post wasn't on when I posted.

[timshorts]

No doubt the 3 clubs are Villa, Watford and the whining East Sussex club.

Hang on. The place in east sussex is Westham not West Ham. Westham is nice enough. The only big bowl with no atmosphere nearby is Pevensey Castle.

[Roosterbooster]

How do they work out the 98.8%?

I'm not saying it's wrong, because I don't know the methodology. But what sort of population was the 98.8% based on? If it was based on people suspected of having coronavirus, with say 25% of the people tested having the disease, then it's not a bad record; if it's based on an apparently healthy population like footballers, then like you say it's pretty rubbish. But is it possible that it's based on finding 98.8 out of every 100 positive cases that come before it? Because if it;s that, and it would be a logical way to test, then it's pretty darn good.

I just don't know how they measure the 98.8%.

I suspect they know the sensitivity and specificity of the test, and have used this to calculate how "accurate" this specific subset is. From experience, I know repeat tests do come back as positive, after previously being negative. A sensitivity of over 95% is very unlikely based on the cases I've seen. If they've got 98.8% sensitivity on their test, then they need to share how, and quickly. Specificity is likely to be close to 100%, but not as low as 98.8%. So I dont think they are quoting either of these.

[Zlatan]

Apologies, delete this please,I’m off to spec savers

Sorry MrTopTier - sarcasm is my elixir these days - lockdown fuel

[huw.Y.WattfromWare]

Players haven’t returned to training, have they, so presumably no contact with players?

All the staff were in for the test but I would hope they were scheduled separately.

[Roosterbooster]

To start off, I’m no expert in this but I just want to question the logic a little if I may.

If 98.8% are classed as accurate then that leaves 1.2% as inaccurate. Now inaccurate could be positive or negative as far as I can understand - or am I getting this wrong? If the 1.2% are inaccurate because the swab wasn’t done correctly; or the test doesn’t work properly I suppose would it be classified as invalid as opposed to inaccurate? I hope I’m making sense...!

Ta

Correct that 1.2% are inaccurate. They COULD be positive or negative, but based on the methodology and nature of the test, false positives are exceedingly unlikely. Therefore most,
if not all, inaccurate tests will be false negatives.
Some tests will be classed as invalid. But these will be tests that were, for example, labelled incorrectly, not sent within a specific timeframe, out of date etc. These will not be tested, or the test will not be included in the results.
The lab testers will unlikely know if a swab was done correctly (although even if done correctly there is a chance it inadvertently samples an area with a low viral load in a covid positive patient). Therefore these will not be classed as invalid, and will still be tested. Data from large samples gives us an idea how likely these tests come back as negative incorrectly, based, amongst other things, on retesting with different results. We also don't know if the test result is accurate just based on the lab test itself, rather than the quality of the swab taken. But these tests are rarely wrong. PCR is often the gold standard.
So the inaccurate tests, logically, are almost certainly false negatives. That is, positive cases we have missed.
By manipulating the stats, you can still make a low hit rate look good. That's why the real sign of a good test depends on the sensitivity and specificity being reported. You dont need both a high sensitivity AND specificity. If you want to rule out a disease, such as COVID, you need a high sensitivity (or hit rate). 40% is not high. It's abysmal.

[Spijed]

apparently Woan is asymptomatic

https://twitter.com/RobHarris/status/12 ... 4509370368

Great for his personal wellbeing but this could be difficult because the virus could just hang around on him as in cases we have heard of in Italian football

The same goes for anyone else who has no symptoms.

[Wile E Coyote]

I love my football team, and its great that sophisticated and regular testing for players is being undertaken, shame there's hardly any available for the local nursing and care teams in the care homes in our county. losses have been staggering recently.

[KateR]

while it's fine working out the math's and saying 1.2% inaccurate, that is surely only for one test and I believe those tested are going to have repeated tests (twice a week?) prior to playing, (June 12?) and also throughout the playing program. Therefore this inaccuracy is reduced drastically over this period and the fact the test tend to default to a negative false test, this helps in the "players wont be isolating due to a false positive result".

[huw.Y.WattfromWare]

the fact the test tend to default to a negative false test, this helps in the "players wont be isolating due to a false positive result".

On the downside they could come into Contact with a “positive” believing them to be “negative”.

[Chester Perry]

3 of the positives are at Watford - 1 of which is a player - a number of other players refusing to train as a result

https://www.dailymail.co.uk/sport/football/index.html

which means 1 other club had 2 positives

[ClaretTony]

3 of the positives are at Watford - 1 of which is a player - a number of other players refusing to train as a result

https://www.dailymail.co.uk/sport/football/index.html

which means 1 other club had 2 positives

And yet big mouth Barber at Brighton has already said they've had three tested positive. Interestingly, the most they can now have is two.

[KateR]

On the downside they could come into Contact with a “positive” believing them to be “negative”.

I'm positive but that's just being negative

[Devils_Advocate]

And yet big mouth Barber at Brighton has already said they've had three tested positive. Interestingly, the most they can now have is two.

This was over a week ago and two of the players were reported as contracting the virus earlier in the pandemic and already being recovered when the story went out.

With that in mind what has the three Brighton players got to do with the 6 that have tested positive from the last few days testing?

[paulatky]

Or on the contrary it’s less of a worry as it shows it’s less symptomatic or problematic for more people.

It depends how long they have been infected. Most people dont show symptoms for 3 to
4 days .
Will depend on how many of those 6 go on to develope symptoms.

I player and 2 staff at Watford make up 3 of the others

[huw.Y.WattfromWare]

One club only tested today but I can’t find which one. Results to follow.

[Roosterbooster]

while it's fine working out the math's and saying 1.2% inaccurate, that is surely only for one test and I believe those tested are going to have repeated tests (twice a week?) prior to playing, (June 12?) and also throughout the playing program. Therefore this inaccuracy is reduced drastically over this period and the fact the test tend to default to a negative false test, this helps in the "players wont be isolating due to a false positive result".

Yes, I agree repeat testing does improve the hit rate. But a low sensitivity will lead to a substantial delay in diagnoses. And therefore spread will still potentially occur.

As for avoiding players isolating due to a false positive, i'd rather they had a false positive, isolated, and then had a follow up test later on. As it is, it just leads to the virus spreading unknowingly, which really makes a mockery of the whole thing.

In fairness, I'm not sure there's much we can do to improve things. It's probably the virus tself that makes it difficult to test for. I suspect a person can be infective, but early on, doesn't have enough of a viral load spread homogenously in the nose / throat, meaning we don't necessarily pick it up on single swabbing.

All I'd really ask for is transparency when it comes to testing accuracy: PUBLISH THE SENSITIVITY AND SPECIFICITY

[KateR]

RB,
like you say not much any of us can do about it, I would hope/think the clubs are getting the best tests available that money can buy, the other thing I hope is that they are all using the same test kits and keep measuring the results and tracking the who/where & when. If it defaults to more false + rather than - we/they have to understand that issue and try to work around it and ensure safest possible outcome.

For me the fact that 0.8 were positive is not that bad out of the diverse sample number and that this, number of how many test positive over time, should continue to reduce. Will be interested to see the results from the same people over a given time and see if there is a spike after reductions before starting, while hoping and praying it does not.

I still think a World Cup style environment would have been safest with everyone away from family, the 99.2% tested negative will still be subject to catching the virus from family, friends, routines etc. and bringing it back to the larger group. Yet again nothing we can do about that either.

From some of the German football pictures there was still some goal celebrations where they were all over each other, yet the thought process is if none of them have it, then none can spread to the other but there are weak links in the chain that don't make it foolproof.

[HunterST_BFC]

re Ian Woan...

Highlights the benefit of testing. He may otherwise have been unknowingly spreading the virus on visits to supermarkets etc.

Exactly this ^^^^^

We the UK have ignored the science and allowed this to spread.
We still are.

I still can't tested weeks down the line of being told I most likely have it.
Simply told to phone an ambulance if /when it gets too much.

The "isolate for a week" being told to players is stupid.
The two week isolate mantra is wrong also. Two weeks then testing again is what it should have been including for everyone the person identified as having put at risk without knowing at the time. Personally I'd say 3 or best 4 weeks isolate.

Sorry this is not a rant, not political or anything other than the frustration at some of the decisions made, and those being made.

Football is not that important.
Health apparently not either?
Money apparently is.


Hope Ian Woan is OK and hope he's not expected to be at work again in a week . . .

BFC can you get me a test?

[IanMcL]

BBC report:

Burnley assistant manager Ian Woan has tested positive for coronavirus, the Premier League club has confirmed.

The Clarets' first-team squad and staff were all tested for Covid-19 at the weekend, before a return to training.

Burnley say former midfielder Woan, 52, is "asymptomatic" and will now self-isolate for seven days.

There were six positive tests across three Premier League clubs on Sunday and Monday, as the English top flight prepares to resume in June.

A total of 748 players and staff from 19 clubs were tested. Norwich City did their tests on Tuesday so will be included in Saturday's results.

Squads started non-contact training from Tuesday.

The Premier League has been suspended since 13 March because of the coronavirus pandemic, with 92 fixtures remaining.

Burnley say Woan will be retested in the week commencing Monday, 25 May.

The club added: "Ian is asymptomatic and is currently safe and well at home. He will remain in close communication with club personnel regarding his re-engagement in training once he is clear of the virus."

Copyright © 2020 BBC. The BBC is not responsible for the content of external sites.

[mdd2]

The PCR test has a specificity of around 99% which means only 1 in 100 who test positive will not have the viral rna and in fact PCR is probably even better than that but it is not a very sensitive test in that only at best 75% of those with the virus will test positive and a lot of that is due to problems getting good nasal and throat samples and getting those samples to the labs. As I have posted before and those on the front line will confirm quite a few where Covid-19 is suspected need more than one test to pick it up and for those who get Covid-19 pneumonia the radiologists reporting the X-rays are better than the PCR test.
It will therefore under diagnose Covid-19 when it comes to picking up carriers of the virus who have no symptoms or those with just a cough or snotty nose.
When it comes to contact tracing therefore there will be those who have it who test negative, hence the need to self isolate and I could not understand how NHS staff who had a few symptoms who tested negative were allowed back without completing isolation.
As for the footballers not wanting to play, that is fine and understandable but they need to forgo their salaries.

[Roosterbooster]

The PCR test has a specificity of around 99% which means only 1 in 100 who test positive will not have the viral rna and in fact PCR is probably even better than that but it is not a very sensitive test in that only at best 75% of those with the virus will test positive and a lot of that is due to problems getting good nasal and throat samples and getting those samples to the labs. As I have posted before and those on the front line will confirm quite a few where Covid-19 is suspected need more than one test to pick it up and for those who get Covid-19 pneumonia the radiologists reporting the X-rays are better than the PCR test.
It will therefore under diagnose Covid-19 when it comes to picking up carriers of the virus who have no symptoms or those with just a cough or snotty nose.
When it comes to contact tracing therefore there will be those who have it who test negative, hence the need to self isolate and I could not understand how NHS staff who had a few symptoms who tested negative were allowed back without completing isolation.
As for the footballers not wanting to play, that is fine and understandable but they need to forgo their salaries.

Where are you getting the sensitivity data from?

[paulatky]

The PCR test has a specificity of around 99% which means only 1 in 100 who test positive will not have the viral rna and in fact PCR is probably even better than that but it is not a very sensitive test in that only at best 75% of those with the virus will test positive and a lot of that is due to problems getting good nasal and throat samples and getting those samples to the labs. As I have posted before and those on the front line will confirm quite a few where Covid-19 is suspected need more than one test to pick it up and for those who get Covid-19 pneumonia the radiologists reporting the X-rays are better than the PCR test.
It will therefore under diagnose Covid-19 when it comes to picking up carriers of the virus who have no symptoms or those with just a cough or snotty nose.
When it comes to contact tracing therefore there will be those who have it who test negative, hence the need to self isolate and I could not understand how NHS staff who had a few symptoms who tested negative were allowed back without completing isolation.
As for the footballers not wanting to play, that is fine and understandable but they need to forgo their salaries.

Didnt Inchy post that information about the difficulty of doing the test about 6 weeks ago on the covid-19 thread.

Inchy is a frontline worker in ICU in Leeds.

It is hoped that within those 6 weeks testers have leant a lot more about doing the test to achieve far greater accuracy. The lack of accuracy was in positive people being recorded as negative.

[paulatky]

while it's fine working out the math's and saying 1.2% inaccurate, that is surely only for one test and I believe those tested are going to have repeated tests (twice a week?) prior to playing, (June 12?) and also throughout the playing program. Therefore this inaccuracy is reduced drastically over this period and the fact the test tend to default to a negative false test, this helps in the "players wont be isolating due to a false positive result".

What a total irresponsible post.

For the good of everyone it is better for a false positive result than a false negative one.

If others like you have the same attitude no wonder the USA have the problems they have.

One of the most stupid and insensitive posts I have seen during this whole crisis.

To sum up your post” Better for a footballer not prevented from training than to saves lives “

[2 Bee Holed]

Get well soon Ian..

??? He isn't ill, he didn't know he had it.

[cricketfieldclarets]

And yet big mouth Barber at Brighton has already said they've had three tested positive. Interestingly, the most they can now have is two.

Not necessarily. Hudson Adoi was positive, Arteta, Mendy possibly. And these wont be included.

[paulatky]

??? He isn't ill, he didn't know he had it.

But may well become ill.

Remember Boris was first reported with very mild symptoms but 2 weeks later ended up in ICU.

Good luck Ian

[Zlatan]

But may well become ill.

Remember Boris was first reported with very mild symptoms but 2 weeks later ended up in ICU.

Good luck Ian

But Boris was absorbing Covid-19 at every hospital handshake increasing his vital load with every shake - and he’s also in the high BMI risk group. Mr Woan is less likely to have had a massive viral load, and is also fit and healthy - you can’t compare Apples with Bananas

[Spijed]

But may well become ill.

Remember Boris was first reported with very mild symptoms but 2 weeks later ended up in ICU.

Good luck Ian

The chances are still very small of that happening when no-one in football has been reported as having any problems from the virus.

[cricketfieldclarets]

But Boris was absorbing Covid-19 at every hospital handshake increasing his vital load with every shake - and he’s also in the high BMI risk group. Mr Woan is less likely to have had a massive viral load, and is also fit and healthy - you can’t compare Apples with Bananas

They’re both one of your five a day though.

[mdd2]

Where are you getting the sensitivity data from?

Read this

BMJ 2020;369:m1808 doi: 10.1136/bmj.m1808 (Published 12 May 2020)

You will see that when it comes to testing those with mild symptoms or none at all a negative result means not a lot. A negative result in the Burnley players who are perfectly well means some may have the virus or none of them. A negative test in someone with a snotty nose does not mean they do not have it but they may not have it as they may have a cold.
A negative test in someone with a bad pneumonia where doctors think it is Covid would make the clinician re test and perhaps three or four times as the pre-test probability of this patient having Covid is high. The pre test probability of any Burnley player having it is low.
We are presently hearing about the lack of testing early on and care homes being sacrificed in favour of NHS but bad as it was the reality is because of the vulnerability of people in care homes the residents should have been managed as if they had the virus and staff too should have managed them as if they, the staff, had the virus, the latter hard to do due to a shortage of PPE. Bog standard infection control which Nursing homes would be familiar with but not necessarily residential homes should have been in place at the outset and one of the first cases we had came from a care home after a fall and the person developed chestiness after 48hr and the young doctor who had her swabbed was ticked off as the patient was low risk until the result came back followed by many apologies to the young lady.

[dougcollins]

I think apples and bananas are relatively safe.

[cricketfieldclarets]

I think apples and bananas are relatively safe.

Not in Tanzania.

[UnderSeige]

There is one big question relating to tests and Premier League Players.

Will Premier League Players be subject to the same 'Track and Trace' regulations as the rest of the population?

If the answer is yes then there is likely to be disruptions to the league program when 'full contract training' and 'full contact matches' are in place. Just one positive test could disrupt 4 matches.

If the answer is no (e.g. just isolate the one player infected without isolating those who he has come into close contact with) then there is possibly a moderate risk of the virus spreading between clubs and through squads.