Studies of the rates of false +ve's are all well and good but the rate
changes depending on who is being tested. For example - a +ve pregnancy
test on a male is pretty much guaranteed to be false +ve. A +ve HIV
test on a 90 yo nun is much, much more likely to be a false +ve than a
+ve result on a Nairobi prostitute. Yet the sensitivity & specificity
of the test are identical regardless of who it is being tested upon. It
is affected by the pre-test probability in the subject being tested. So
there is generally no standard rate of false +ve's for a test.
>From WADA's website, they did ~4229 urine tests for Epo in 2003 and got
28 +ve. Some reasons for false +ve's are known from the Beke situation:
exertion-induced loss of alpha 1 ACT & bacterial contamination of the
sample due to improper storage. Both of these are relatively easy to
test for even in a stored sample so someone unlucky enough to be
falsely +ve can prove these without much difficulty.
So for an elite athlete competing at the highest level, which is more
likely (a) doping like maybe 90% of the peloton as Benjo believes or
(b) a rare renal condition plus improper storage of the urine specimen
in a very unlucky athlete? The chances of (b) are very slim, especially
if they've tested +ve on, say, 6 occasions.
Many studies have been/are being done regarding rEpo detection. From a
PubMed search, you can see some of these:
http://www.ncbi.nlm.nih.gov/entrez/...=Display&dopt=pubmed_pubmed&from_uid=15228161
Just because you & I don't know the sensitivity, specificity & positive
predictive value doesn't mean that hasn't been studied.
