Coronavirus testing

[ratchers]

I don't really understand the purpose of testing-

to see whos got it?

to see whos had it?

how does this help?

[Dave-G]

Those who have had it can mix with others is my understanding - and those who have it know to self isolate in another room rather than just stay home?

Edited April 17, 2020 by Dave-G

[ratchers]

Just been reading that Germany says their outbreak is under control and one of the reasons is all the testing they have done

Also read they you can get re-infected with corvid 19?

[Raja Clavata]

It's still very much an inexact science and nobody really knows the answers definitively.

It seems like the swab test is not 100% reliable - on that basis everything else falls down from a definitive viewpoint.

My understanding is if you've tested positive in hospital and admitted you can't be discharged until you test negative (given the above this could involve more than one test).

There is talk of re-infection but nobody yet knows if the original test result was wrong or if your test after "apparently" recovering was wrong or if the virus can lay dormant and re-appear later on or perhaps other scenarios.

With regards to immunity they don't really know the score there either - initial studies suggest length of immunity varies between individuals.

This all leads to further uncertainty over the effectiveness of a vaccine.

In summary, we potentially don't yet know the full extent of all the things we don't know about this virus.

What we can be pretty sure of is that Germany has done a lot more testing than we have but we don't know to what extent that has been to their advantage - each countries numbers differ in the way they are put together so it's unclear, beyond total number of tests completed, what valid comparisons can be drawn.

By no means definitive but I hope the above helps, at least a bit. ATB.

[Dougy]

Its very much unlike a pregnancy test, with that you either are or your not, none of the "your not quite pregnant". 

 

Tests have shown that the Covid19 test is not 100% reliable, giving folk a false sense of security, not really a good idea under the present conditions

[Jaymo]

7 minutes ago, Dougy said:

Its very much unlike a pregnancy test, with that you either are or your not, none of the "your not quite pregnant". 

 

Tests have shown that the Covid19 test is not 100% reliable, giving folk a false sense of security, not really a good idea under the present conditions

Correct

[eddoakley]

So you test negative and have to isolate so you don't get it. Until when?

You test positive and have to isolate until you can get retested as having had it, you are fine but are now immune?

So you are immune and are ungrounded.  But you can still spread it as it can survive on any surface for a million billion years and you can pass it on to the next person who touches something that you have touched or they have had it but the tenth person hasn't and they can get it?

The guy on tv says tomorrow that it's all good now and you are allowed to get back to normal....is that it? All done and just forget about it?

But wait...apparently you can be re infected. And second time around you can have worse symptoms. Unless you believe the other version that you can't be re infected and you can't pass it in and testing will fix everything.

Seems to me that there is a lot of hype, propognda, misreporting and misinformation, not to mention scare mongering and plain simple bulls%!/.

Also seems to me that more and more people are starting to ignore being grounded and are trying to get in with life.

Just my thoughts.

Edd

 

Oh yeah and tests are so far unreliable and innacurate. One story I read was that they were 40% accurate!?!?! 

Is that even possible? Surely it can't be less than 50%. Either it's right or wrong? Like tossing a coin.

 

Edited April 17, 2020 by eddoakley

[eggy74]

italy and germany have carried out the same number of tests (per million of population) 

 

Quote

Quote

 

 

Quote

 

 

[Smokersmith]

2 hours ago, ratchers said:

I don't really understand the purpose of testing-

to see whos got it?

to see whos had it?

how does this help?

The only test widely used is on people showing symptoms, to see who has it.

It’s useful for identifying, who has it, is therefore a risk to others and needs to self-isolate......, and who is just feeling a bit rough and can get back to work.

[Blackstone]

The nucleic acid test (swab test) accuracy depends a lot on the quality of the swab, and where you swab. Best sample type is from a bronchoalveolar lavage (a thoroughly unpleasant procedure where they flush your lungs with fluid and then collect it. you have to be sedated for this procedure, and it's very dangerous to the people carrying out the procedure due to the risk of splashing and aerosols). Most swabs are taken from the nose/throat, and those samples will only give around a 70% accurate result.

3 hours ago, eddoakley said:

Oh yeah and tests are so far unreliable and innacurate. One story I read was that they were 40% accurate!?!?! 

Is that even possible? Surely it can't be less than 50%. Either it's right or wrong? Like tossing a coin.

 

That's not how it works. You can't really apply "accuracy" to a test. A test's performance is separated into two things: sensitivity and specificity.

A test's sensitivity is how good it is at correctly identifying a positive result. A low sensitivity results in more false negatives.

A test's specificity is how good it is at correctly identifying a negative result. A low specificity results in more false positives.

"Accuracy" is just a measure of how much a test agrees with the results of a "gold standard" test that it's being compared to. But it's really not a useful metric beyond trying to explain the test's effectiveness to a layperson.

[Mice!]

3 hours ago, eddoakley said:

So you are immune and are ungrounded.  But you can still spread it as it can survive on any surface for a million billion years and you can pass it on to the next person who touches something that you have touched or they have had it but the tenth person hasn't and they can get it?

Probably best to stay home for now, its almost like there not really sure, better to stay safe and hope someone clever comes up with something clever,  or carry on as you like and see what happens. 

I'm not in any at risk groups,  not old and back at work but I'm avoiding any unnecessary contact,  if I'm not shopping or at work then I'll be at home, it's not the hardest thing you'll be asked to do, stay at home.

[Longbower]

There is a company called 'Surescreen ' based in Derby. Just have a look at their website , and make your own minds up, if accurate tests are available. 

[Raja Clavata]

9 hours ago, Blackstone said:

The nucleic acid test (swab test) accuracy depends a lot on the quality of the swab, and where you swab. Best sample type is from a bronchoalveolar lavage (a thoroughly unpleasant procedure where they flush your lungs with fluid and then collect it. you have to be sedated for this procedure, and it's very dangerous to the people carrying out the procedure due to the risk of splashing and aerosols). Most swabs are taken from the nose/throat, and those samples will only give around a 70% accurate result.

That's not how it works. You can't really apply "accuracy" to a test. A test's performance is separated into two things: sensitivity and specificity.

A test's sensitivity is how good it is at correctly identifying a positive result. A low sensitivity results in more false negatives.

A test's specificity is how good it is at correctly identifying a negative result. A low specificity results in more false positives.

"Accuracy" is just a measure of how much a test agrees with the results of a "gold standard" test that it's being compared to. But it's really not a useful metric beyond trying to explain the test's effectiveness to a layperson.

Good post. Just reading that the WHO are advising that testing positive for the antibodies is no guarantee of immunity, that has to be a worry...

[Blackstone]

2 hours ago, Raja Clavata said:

Good post. Just reading that the WHO are advising that testing positive for the antibodies is no guarantee of immunity, that has to be a worry...

It sounds like the WHO are being overly cautious because of a lack of peer-reviewed studies. The fact is that there HAS to be some degree of immunity, otherwise you wouldn't be able to recover from the infection. There are two forms of long-term immunity - humoral (antibody-based) and cell-mediated (memory T-cells).

The easiest to study is humoral immunity, because it's easier to measure how much antibody someone has. Long-term immunity is provided by Immunoglobulin G (IgG), which is produced about 2 weeks into an infection. IgG levels peak at 4 weeks and then slowly decrease over time. Their presence in blood should provide immunity for at least a year. In the case of the 2003 SARS virus, a followup study found that most people did not have detectable IgG to SARS after 6 years. However, this doesn't mean they aren't immune anymore, because you should still have your cell-mediated immunity.

Cell-mediated immunity makes use of memory cells, who can remember the particular pathogen that originally invaded the body. When reinfection occurs, they can immediately reactivate the body's defences against that pathogen (phagocytes, killer T-cells, cytokines etc...). There are also memory B-cells that when activated, will produce IgG antibodies again. Cell-mediated immunity is a little harder to measure, because you have to take someone's blood, introduce bits of the virus to it, and see if there is any response to it. (As an aside, this cell-mediated immunity test is called an Interferon Gamma Release Assay, and is one of the only ways of diagnosing people who have a latent tuberculosis infection. That is, they have TB, but do not show any signs of it, and do not even have antibodies to it)

So the reason why WHO is saying that there is no guarantee of immunity, is only because there is little peer-reviewed research into this. Nobody knows how long the initial antibody response is good for. And nobody knows the nature of the cell-mediated immunity to it.

Edited April 18, 2020 by Blackstone

[Blackstone]

10 hours ago, Longbower said:

There is a company called 'Surescreen ' based in Derby. Just have a look at their website , and make your own minds up, if accurate tests are available. 

Their test looks suspiciously like a Chinese test that we've seen, so I don't think they are actually making it...

[fatchap]

Horrible test, swipe your tonsils with the cotton bud, I came very close to projectile vomiting all over my steering wheel and dashboard, Then shove it up your trumpet both sides causing uncontrollable sneezing for about 10 minutes.

[Raja Clavata]

13 minutes ago, Blackstone said:

It sounds like the WHO are being overly cautious because of a lack of peer-reviewed studies. The fact is that there HAS to be some degree of immunity, otherwise you wouldn't be able to recover from the infection. There are two forms of long-term immunity - humoral (antibody-based) and cell-mediated (memory T-cells).

The easiest to study is humoral immunity, because it's easier to measure how much antibody someone has. Long-term immunity is provided by Immunoglobulin G (IgG), which is produced about 2 weeks into an infection. IgG levels peak at 4 weeks and then slowly decrease over time. Their presence in blood should provide immunity for at least a year. In the case of the 2003 SARS virus, a followup study found that most people did not have detectable IgG to SARS after 6 years. However, this doesn't mean they aren't immune anymore, because you should still have your cell-mediated immunity.

Cell-mediated immunity makes use of memory cells, who can remember the particular pathogen that originally invaded the body. When reinfection occurs, they can immediately reactivate the body's defences against that pathogen (phagocytes, killer T-cells, cytokines etc...). There are also memory B-cells that when activated, will produce IgG antibodies again. Cell-mediated immunity is a little harder to measure, because you have to take someone's blood, introduce bits of the virus to it, and see if there is any response to it. (As an aside, this cell-mediated immunity test is called an Interferon Gamma Release Assay, and is one of the only ways of diagnosing people who have a latent tuberculosis infection. That is, they have TB, but do not show any signs of it, and do not even have antibodies to it)

So the reason why WHO is saying that there is no guarantee of immunity, is only because there is little peer-reviewed research into this. Nobody knows how long the initial antibody response is good for. And nobody knows the nature of the cell-mediated immunity to it.

Thanks for this. Very insightful, cell-mediated immunity is a new one for me. Yes I figured WHO could just be being cautious but I guess we have no way of knowing yet.

Do you have a view on the different strains - could this be a cause of the reported re-infections? 

[7daysinaweek]

33 minutes ago, Blackstone said:

It sounds like the WHO are being overly cautious because of a lack of peer-reviewed studies. The fact is that there HAS to be some degree of immunity, otherwise you wouldn't be able to recover from the infection. There are two forms of long-term immunity - humoral (antibody-based) and cell-mediated (memory T-cells).

The easiest to study is humoral immunity, because it's easier to measure how much antibody someone has. Long-term immunity is provided by Immunoglobulin G (IgG), which is produced about 2 weeks into an infection. IgG levels peak at 4 weeks and then slowly decrease over time. Their presence in blood should provide immunity for at least a year. In the case of the 2003 SARS virus, a followup study found that most people did not have detectable IgG to SARS after 6 years. However, this doesn't mean they aren't immune anymore, because you should still have your cell-mediated immunity.

Cell-mediated immunity makes use of memory cells, who can remember the particular pathogen that originally invaded the body. When reinfection occurs, they can immediately reactivate the body's defences against that pathogen (phagocytes, killer T-cells, cytokines etc...). There are also memory B-cells that when activated, will produce IgG antibodies again. Cell-mediated immunity is a little harder to measure, because you have to take someone's blood, introduce bits of the virus to it, and see if there is any response to it. (As an aside, this cell-mediated immunity test is called an Interferon Gamma Release Assay, and is one of the only ways of diagnosing people who have a latent tuberculosis infection. That is, they have TB, but do not show any signs of it, and do not even have antibodies to it)

So the reason why WHO is saying that there is no guarantee of immunity, is only because there is little peer-reviewed research into this. Nobody knows how long the initial antibody response is good for. And nobody knows the nature of the cell-mediated immunity to it.

Very informative and excellent post Blackstone 👍

[bluesj]

I was talking to someone yesterday that has a sort of medical background (shes a vet) about testing etc.

The antigen test tests if you have got the virus now and the antibody test test if you have had or been exposed to the virus.

the sensitivity of the tests is how good it is at say if you have or haven't got something and specificity is how good it is at checking what you are looking for. she said a test that can get the right positive or negative result 100% of the time is still no good if it can't tell the difference between the virus and a broken leg.

Also once you have had the virus you will have some immunity otherwise you would not recover but no one can know how long the immunity will last as the virus has only been known about for a few months, It could be weeks, months, years or it could be for life.

[Blackstone]

1 hour ago, Raja Clavata said:

Thanks for this. Very insightful, cell-mediated immunity is a new one for me. Yes I figured WHO could just be being cautious but I guess we have no way of knowing yet.

Do you have a view on the different strains - could this be a cause of the reported re-infections? 

I don't know much about this theory of two different strains. However, there does seem to be reports that the Spike protein (how the virus attaches to our cells) is undergoing significant mutation. This would be a problem as it means that our immune system will no longer recognise that spike protein. This means the antibodies will no longer be able to bind to them to "deactivate" them. I believe the original SARS virus had a pretty genetically stable spike protein. Coronaviruses in general are much more stable than Influenza viruses. I'm not too worried about this right now, but this could change as more research is done.

I would take the reports of re-infections with a grain of salt. In many of these cases, it is likely that "reinfection" was really just the PCR test detecting bits of dead virus. Or the patient had a previous false-negative test, and hadn't actually recovered (I refer back to my earlier post about the "accuracy" of PCR tests).

[Blackstone]

Just wanted to make another post about test accuracy. This isn't really directly applicable to coronavirus testing though, for reasons I'll explain.

It's not simply enough to know a test's sensitivity and specificity. What is a far more useful figure is the 'positive predictive value' and the 'negative predictive value'. The PPV is the probability that someone who tested positive actually has the disease. Conversely the same for the NPV. The PPV and NPV depends heavily on how prevalent a disease is in a given population.

This is easier with a worked example:

Let's say we have a test that is 99% sensitive and 99% specific. That is to say, if it is used to test 100 positive people, it will return 99 positive results, and one false negative result. And if it tests 100 negative people, it will return 99 negative results, and one false positive result. On paper, this is a great test.

But, let's say that in a population of people, this disease only affects 1 in 1000 of them. If you randomly screened 1000 people with this test, then you would expect 11 of those results to return positive. One of those is the actual true positive case. The other 10 are false positive results. That means that for every true positive, you have 10 false positives. That's a 9% positive predictive value.

This is why blindly screening for rare diseases can be a waste of time and money, and can actually be harmful. And this is why you only perform certain tests on high-risk people who do have a high chance of actually having the disease. Now, this isn't applicable to coronavirus because of how widespread COVID-19 is, I just wanted to give an example of why accuracy of tests is actually quite a complex subject.