dcd2103 said...
You sure about this g-man?
A 1:320 means they dilute it 320x and the antibody still needs to be detected to test positive. That’s a much higher hurdle than 1:64, where it’s diluted only 64x, so the antibody is easier to detect
yep - i didn't word that too well - but i think we have the same basic understanding - just tripping up over terminology - to try to clarify
a test that reports positive on a sample at or above a titre dilution ratio of 1:320 is reporting positive for a more dilute mixture of the original sample (and more added saline) - and as such this test is reporting positive with less antibodies per volume - and is therefore more likely to report positive than a similar test reporting positive for lower titre dilution ratios eg 1:64 ( all other things being equal - which of course they rarely are - see the rest of my original post )
they don't pick these numbers out of the air on a whim though - each lab should follow a well trodden procedure for arriving at what titres it uses for its thresholds based on a bit of math called Receiver operating curves (ROC) to try to optimise the "diagnostic value" of a given test - by picking the best compromise between more false positive vs false negative rates
there is a decent explanation of how this is done here - and in the linked page in this article to ROC
https://www.msdmanuals.com/en-gb/professional/special-subjects/clinical-decision-making/understanding-medical-tests-and-test-results for info - serological tests for lyme and co-infections are rarely anywhere close to the ideal
ie 100% sensitive (zero false negatives ) and 100% specific ( zero false positives)
so you effectively have to draw a line in the sand and say - "i accept this many false negatives and false positives" -
but there are a bunch of other variables that mean in the real world a test with a threshold of 1:64 may be more sensitive than on with a threshold of 1:320