Red Herring
middle-class woman of a certain age
- Joined
- Jun 9, 2010
- Messages
- 7,916
- MBTI Type
- INTP
- Enneagram
- 5w4
- Instinctual Variant
- sp/sx
Since you occasionally hear people talk about false positive tests, I anted to share some numbers.
There was a study done in June which seemed to suggest that there PCR-tests had a false positive rate of somewhere between 1.4 and 2.2%, depending on whether the tests samples contained a closely related virus. That would be relatively high. However, that study was based on a relatively low number of tests (7000 samples sent to 463 labs in 36 countries). Also, quite a few of those were down to human error, i.e. mixed up samples etc and varing quality of the labs involved.
BUT if you look at some more current numbers from Germany, they suggest a much, much lower false positive rate. First of all, there is a threshold below which the virus is har to detect so some sampes considered false positive because a second test came back negative much actually have been right but the second test didn't detect the low concentration. In the second week of July German did some 510.00 tests and only 0.59% of those came back positive, so even if somehow all of them were false positives, the false positive rate must be well below 1% or have been well below 1% in that week for that half a million of tests. In a imilar vein, there is a lab with 500 employees that tested them on a regular basis. They have done roughly 20.000 tests on these people all together and only 3 of them ever came back positive - from people who did show symptoms!
That suggests that while false positives are obviously always something to keep an eye on, the rate of false positive tests seems to be fairly low. False negative PCR-tests seem to be much more of an issue as there is only a relatively limited time window for them to make any sense.
There was a study done in June which seemed to suggest that there PCR-tests had a false positive rate of somewhere between 1.4 and 2.2%, depending on whether the tests samples contained a closely related virus. That would be relatively high. However, that study was based on a relatively low number of tests (7000 samples sent to 463 labs in 36 countries). Also, quite a few of those were down to human error, i.e. mixed up samples etc and varing quality of the labs involved.
BUT if you look at some more current numbers from Germany, they suggest a much, much lower false positive rate. First of all, there is a threshold below which the virus is har to detect so some sampes considered false positive because a second test came back negative much actually have been right but the second test didn't detect the low concentration. In the second week of July German did some 510.00 tests and only 0.59% of those came back positive, so even if somehow all of them were false positives, the false positive rate must be well below 1% or have been well below 1% in that week for that half a million of tests. In a imilar vein, there is a lab with 500 employees that tested them on a regular basis. They have done roughly 20.000 tests on these people all together and only 3 of them ever came back positive - from people who did show symptoms!
That suggests that while false positives are obviously always something to keep an eye on, the rate of false positive tests seems to be fairly low. False negative PCR-tests seem to be much more of an issue as there is only a relatively limited time window for them to make any sense.