Standards

Back in the day, when I worked for a national health magazine, the gold standard for any scientific study that wanted to find its way into our pages was that it had to be a controlled, double-blind, peer-reviewed study.

Controlled means that along with testing the therapy in question—a new drug, a medical device, a technique—the test also included a control set: a set of subjects who received no therapy at all. You can’t measure improvement accurately without an accurate baseline.

Double-blind means that the people in the test must not know whether they are receiving the treatment or receiving a placebo, and the people evaluating the data must not know whether they are seeing results of someone who got the placebo or the treatment that’s being tested. This makes some treatments difficult—or impossible—to evaluate. If the treatment is massage or acupuncture, you generally know whether you’ve been rubbed or stuck with needles. If the treatment is a pill, it’s easier to keep the study blind.

Peer-reviewed means that the study has been subjected to the scrutiny of other specialists in the same field, besides the folks who performed the study and wrote the paper.

If a lab is testing blood or urine samples from Barry Bonds, they might well have an agenda—either proving that he’s using steroids or proving that he’s not. So it’s important that the lab should not know whose samples they’re testing.

At the arbitration hearing in Malibu this spring where U.S. cycling arbitrators heard the case of Floyd Landis from last year’s Tour de France, it became clear that the lab technicians probably knew exactly whose samples they were testing, and their significance, as they tested them.

It also became starkly clear that the lab technicians did not use the first set of results they got from the samples. They retested and retested until they got a set of results they liked.

That’s uncomfortable for everyone. The rider being accused can’t be sure whether the lab cooked the results to make him look bad—because he’s an American, because they like some other rider, because a different team paid them to falsify results.

The lab is in an awkward position, because no matter how honest they are and how earnestly they believe in the accuracy of their results, it still smells bad.

The lab technicians claimed that they threw out early results because they were pre-tests used to calibrate the machine. Nobody can tell whether they’re being honest, because they did delete all the early results. Nobody can really tell how many times they ran the test before they got the results they ended up with. Nobody gets to see what results they threw away.

Floyd Landis showed during the arbitration that he’s a poor sport willing to stoop to a pretty scummy level to win. But that doesn’t make his point wrong. A single lab is used for all the Tour de France testing; that’s a weakness in the system. That lab chronically—daily during the Tour—illegally leaks test results to the press. Lab technicians seem to regularly know exactly whose results they’re handling. If the results are called into question, the same lab is used to verify the original results.

If we’re going to hold riders to high standards of credibility—ditching them from a race just because they miss a test or lie about their whereabouts—we owe them the dignity of being tested through a process of unimpeachable integrity. Yes, cycling needs to continue surging ahead of all other sports in eradicating dopers from its ranks. Cycling has done a great job so far. But its testing procedures need a complete overhaul in order to protect the riders who are going to every effort to keep the sport clean:

1) Samples need to be identified with a mark that does not identify what rider—or what sport—the sample comes from.
2) At least a half-dozen labs in several countries need to be used for sample evaluation, in random rotation.
3) Labs need to be certified by independent third parties.
4) No lab should be allowed to test a B sample to verify its original results from the corresponding A sample. A second lab must be used to verify results.
5) Any lab whose results are made public except through the institution governing the event must be given a single warning. The second time results are leaked, that lab must be disqualified for a two-year period. A lab that can’t keep its staff from leaking results certainly can’t keep its staff from falsifying results.

Finally, after the testing is done, a record of results should be kept for riders, in case a baseline is needed to evaluate future samples. When Tyler Hamilton was accused of blood doping, he claimed his blood might have been that way (“chimeric”) all his life. A data bank could support a claim like that or shut it down swiftly. A data bank would help sort out who has naturally high testosterone levels after a tough race and who’s been wearing testosterone patches. Riders’ test results could be measured accurately against their own historic baselines, rather than against baselines drawn from the general population.

The UCI, which governs most cycling events worldwide, has asked its riders to sign a “voluntary” pledge that they will give up a year’s income and let themselves be suspended if a lab accuses them of cheating. It’s fair to ask, if the accusations are credible. But until the lab testing system is completely revised, it’s a nervous rider indeed who signs such a pledge. All it takes today is one lab slip-up—or one round of recalibration—for a clean rider to lose a year’s income.

Considering how much of themselves these riders put into each grueling day on the road, the Dick Pounds and Christian Prudhommes who sit behind desks should be held responsible for writing better rules.