Commendations to comrade scott for posting about testing and the pitfalls, fallacies and assumptions in correctly determining illegal drug use. However I believe sending the samples to different labs would be an improvement over the current regime, but not because it would necessarily produce more accurate results (i.e., fewer false negatives or false positives). Rather, splitting the samples between facilities might solve other problems and answer other questions:

1. Why is it that the press seems to get wind of A sample test results well before the protocol dictates? How is it that the test results from the LNDD are matched to the athlete control numbers and reported by L'Equipe so quickly?
2. How can a B sample test be considered reliable if the testers already know the desired outcome? These tests are supposed to be double-blind, are they not?
3. Why is it so easy for an accused rider to claim "error" or "conspiracy"?

I think these questions at least could be addressed by sending the samples to separate labs, accredited by different governing bodies. Test the samples simultaneously and report the results at the same time. Known positive samples should be intermixed and sent along to the labs as well, and perhaps even not all of the samples taken should go, for example: 10 riders sampled, 11 samples go to be tested, 2 are dummies (other 9 selected at random from the 10 collected samples), 1 of the 2 dummies is a known positive. This could be "blinded" further by randomly changing the number of dummies and number of known positives in the dummies.

It ought to be a pretty rare occurrence for all of the test results between the labs not to match. (If it's common, there are other issues.) Yes, there will continue to be false results, but the likelihood of undetected error due to human intervention or equipment problems should be reduced. I won't believe you if you tell me the results are never affected by testing error. And having both samples tested at the same lab in succession after the first results are known makes no sense; this is not double-blind testing. The lab and the whole of professional cycling have a vested interest (more on that in another post) in the second test matching the first. With this kind of procedure in place, it will be much more difficult for a rider to claim that there was a problem with the test itself. It will also be more difficult for labs to collude with the press to leak results.

-Michael