2

A friend of mine was detected COVID-19 positive through PCR one month back. He was asymptotic as per doctors and had no symptoms. He quarantined himself and increased his intake of water. After 1 month, he got himself tested and the result was negative. His antibodies value was +7.0.

My question: Is it possible to get COVID-19 false positive ? 2) Is it possible that a person comes as COVID-19 positive due to bad sampling technique or something ? 3) Given the antibody value as 7.0. Is it possible that the same person gets COVID-19 positive again ?

(My theory is that COVID-19 will try to come to his body but his antibodies value is so high that it won't affect him.)

Improve this question

1 Answer 1

Reset to default
1

The rt-PCR tests for Covid-19 are not 100% specific so yes it's possible to get a false positive.

From the China PCR test

The sensitivity of the RT-PCR diagnostic test was estimated to be 0.777 (95% CI: 0.715, 0.849), while the specificity was 0.988 (95% CI: 0.933, 1.000). The confidence intervals include sampling error in addition to the error due to probabilistic knowledge of the data.

This means that out of 100 true negative patients, the test will correctly identify 98.8 as negative. Out of 100 true positive patients, the test will correctly identify 77.7 as positive. The positive predictive value (PPV, Given a positive test, what are the chances that the patient is actually positive) and negative predictive value however depend on the population sample that is being tested. This answer explains it fairly well.

We don't have enough data on antibodies yet to interpret them.

https://www.medrxiv.org/content/10.1101/2020.04.24.20078949v1

Improve this answer
17
  • 5
    "this suggests 1:100 positives could be false positive" is not true, it depends on the ratio between true positives and true negatives. If, for example, you did 1,000,000 tests on a population with 0 true positives, you would expect to return 12,000 positives, all of which would be false. The correct interpretation is instead that "this suggests that 1/100 people who are actually negative will nonetheless test positive".
    – Bryan Krause ♦
    Commented May 27, 2020 at 21:45
  • 2
    You have stated it as 1/100 positives are false positive; that is not the same as 1/100 negatives will test positive. If the test population contains a lot of no-CoV subjects then the number of positives that are false positive will be much much higher than 1/100. I gave maybe an extreme example of 100% false positive, but let's say you test a population of 1000, 10% are CoV cases and 90% are virus-free. 1.2% (1-specificity) of the 900 (=11) will test positive incorrectly, as will 77.7% (sensitivity) of the 100 (=78). Total 89 positive tests, 11/89 = 12% are false positive, rather than 1.2%.
    – Bryan Krause ♦
    Commented May 28, 2020 at 1:51
  • 2
    Exactly! Also, this is because most screening tools (and the PCR Test as well) are only used on risk groups. The sensitivity and specificity are not dependant on the population sample, but the rate of true to false positives and true to false negatives is. This is why it doesn’t make sense to do any medical test without an indication (regardless of the cost). Only very very few screening tools are used on a whole population sample.
    – Narusan
    Commented May 28, 2020 at 6:07
  • 1
    @GrahamChiu I am solely arguing from the statistical facts of the sensitivity/specificity numbers you reported in your own answer, and the correct interpretation of those numbers. Maybe data from China should be questioned or maybe the specificity is way better than .988. If the specificity is truly .988 you would expect 120,000 positive tests in a sample of 10 million if no one is infected. I believe they used a bulk screening approach which might reduce sensitivity but also improve specificity.
    – Bryan Krause ♦
    Commented Jun 3, 2020 at 3:32
  • 2
    @BryanKrause (and GrahamChirlu) One possible cause of false positives is contamination of negative samples with positive samples. This particular source of false positives would lead to false positive rate (1 - specificity) to be correlated with the number of true positives...
    – cbeleites
    Commented Jun 5, 2020 at 18:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.