What’s more likely...a false positive or a false negative??

[Hermanfromguesswho]

4 of us in the house... 3 tested and we got 2 negatives and 1 positive test back.
We are isolating for the full period regardless, just thinking out loud here...
2 children have no symptoms. (One tested negative and the other didn’t test)
Third child has temperature, bad cough and loss of taste (two negative tests!)
Adult has tight chest, sore throat, cough snd mild temperature. (Positive test)
Is it more likely we are both positive but the child tested negative twice or that the adults test was a false result and we both have a normal cold/flu type thing?
It just doesn’t seem possible that we accurately have different results when we clearly have the same illness...

I think false negative is more likely

I think false negative is more likely.
Did the children do the tests themselves?

I also think a false negative is more likely, I read somewhere that it is possible to do a swab but miss the area where the infection lies .

False negatives are more common

It depends on the test and also symptoms and some other variables:
PCR test, high chance of false positive on asymptotic people (97%)
LF, high chance of false negatives (50%)

% are based on conditional test results not number tested

@Taciturn

It depends on the test and also symptoms and some other variables: PCR test, high chance of false positive on asymptotic people (97%) LF, high chance of false negatives (50%)

% are based on conditional test results not number tested

That’s what I’ve understood to happen also

Approximately a third of people with Covid will test -ve on PCR test. False positives in people with symptoms are low. Loss of sense of taste/smell is pretty specific to Covid. I'd assume you all have it, sorry. Hope it passes quickly and uneventfully.

False positives are rare. Mostly they occur in people who have just recovered from coronavirus but are still shedding bits of viral dna (so the virus is there but the person is not infectious).

False negatives are really common. What you have is COVID. Great well soon x

We don't know the false negative and false positive rates of either type of test. Nor, therefore, do we know the true numbers of people testing positive (so we can't work out the numbers of false results - it's circular). SAGE suggests that the false positive rate is between 0.8% (8 per thousand tests, ie 800 per 100,000) and 4% (4000 per 100,000 tests). The higher figure is probably wrong because there were too few positive tests when cases were low in summer to support that.

The Lancet says that there are reports of false negative rates between 2% and 33% in repeated sample testing. We do know that the time since infection is a very important factor in how likely a negative is to be false. The American College of Cardiology reports that a day 8 test, which they say is 3 days after onset of symptoms on average, is the least likely to produce a false negative.

What we do know is that if cases are low and test numbers are high we are more likely to get false positives, as false positives are a proportion of the tests carried out. False negatives are a proportion of the number of cases, so there will be more of them when there are a lot of people who should be testing positive.

We also know that the PCR test is set up to pick up Covid RNA (to put it simply). It does not differentiate between potentially infective RNA or RNA that is inactive, for example because the body's immune system has dealt with it. The immune system works by destroying the body's cells with the Covid in (simply, again) so a lot of 'debris' will be left, including 'scraps' of deactivated RNA. This is one reason why it is possible to get a positive test result months after the original infection. (We do not know the maximum number of months this is possible for, it's probably a range).

For you, none of this tells you much because your 'sample' (number of tests) is too small. We know that if you toss a coin 6 times statistically you 'should' get 3 heads and 3 tails and this 50-50 will be the outcome of thousands of tosses. But it's not uncommon to get 6 of one, or any other combination, from just 6 tosses.

I wonder the same.

My DS tested positive. Widespread rash, fever, fatigue, loss of appetite, headaches, red sore eyes. His test was 25th November, but he'd had the rash on its own with no other symptoms from around 12th November.

My DD began symptoms on 2nd December. Fever, headaches, red blotches on finger and toe joints, chest pains, joint pains, fatigue. She tested negative twice. The interesting thing was that DD had been nowhere except the garden since the 19th November due to cases in her year group, I also hadn't been anywhere except walks for ages. No one else in our household.

I have had no symptoms.

Wtf.

Possibly more helpful to you is the Diamond Princess. They were originally unaware of the risk of Covid spread, so everyone was mixing freely, mostly in an enclosed space, with a higher than usual average age. Nearly everyone was tested. About one fifth tested positive, of whom about half were asymptomatic. In couples sharing a cabin one testing positive did not mean that the other did.

@Taciturn

It depends on the test and also symptoms and some other variables: PCR test, high chance of false positive on asymptotic people (97%) LF, high chance of false negatives (50%)

% are based on conditional test results not number tested

This is correct

@hamstersarse no thats not correct. If you give 100 people with no symptoms a pcr test 97 will not come back positive for coronavirus!

Even if it was correct it refers to asymptomatic samples. And the positive case in this instance is symptomatic.

False positives are rare. If you are in a low covid area (v unlikely in the U.K.) a large proportion of the positives that you get back might be false positives but that’s just because there aren’t many actual positives.

I tried looking this up recently and couldn't find a definitive answer, I think it's simply not statistically possible to say for sure as there are too many variables.

But even if you did know the statistically correct answer I don't see how it would help you as it won't necessarily apply to your family.

From my experience it's difficult to get a really good test from a child, compared to an adult.

So false negatives are more likely in children just because of the quality of the sample.

False negative