Wednesday, August 06, 2008

Charlie Pavitt: what's the false positive rate for MLB steroids tests?

This guest post is by Charlie Pavitt. Take it away, Charlie ...


I have a question about drug testing, which for the sake of this blog I will restrict to that for steroids by professional baseball leagues although it is just as relevant for all performance enhancing drugs (PEDs) by all sports organizations in general. I’ve had this question for awhile, but this is a good time to ask it, given that the Tour de France was recently completed (with the usual disqualifications for PEDs along the way) and the Olympics is about to begin (with, leading up to the Games, the same).

No drug test is perfect. All lead to some percentage of false negatives (users whom the test misses) and false positives (innocents whom the test implies are guilty). The latter is what specifically concerns me. In our court of law, one is supposedly innocent until proven guilty. But every few weeks I read of another minor leaguer suspended for a couple months for a positive test. Given that every player knows about the tests, I wonder how many are really still using. Yeah, I can see the possibility that some of the more pampered of them, who have never been punished for anything they’ve done in the past decade because their athletic skill has made them sacrosanct in their communities, probably imagine they can use and get away with it. But I’m not at all convinced that they are all guilty. And, given the publicity it generates, a reported false positive would result in a drug-free player pretty much permanently tarred-and-feathered (Rafael Palmeiro comes to mind; I’m not saying that he wasn’t a user, I have no idea either way, but talk about a positive image permanently crashing down to earth in an eyeblink…).

I just did a bit of web searching (googled “drug testing baseball false positive”) and didn’t find a ton of helpful information on it.
One website mentions a study with a huge 14% false positive rate. I’m not the only one concerned (see here, among others).

So here’s my specific, two-part question: What is the false positive rate for the test (assuming there is only one) used in professional baseball? And what precautions are there against false positives (at the very least, there should be a blood sample divided in two, with the second tested in case the first comes up positive)? If anyone has these answers, it is not I’d like to know. More importantly, I’d like the general baseball-fan public to know also.

--Charlie Pavitt

Labels: , ,


At Wednesday, August 06, 2008 9:14:00 AM, Blogger Don Coffin said...

Note that second tests using a divided sample will nopt necessarily be very useful in the case of naturally produced substances like testosterone. In the case of testosterone, there are two false positive results:

1. The test shows a higher concentration of testosterone than actually exists. In this case, a retest is useful.

2. The test accurately measures the testosterone level, but it's a result of natural vairation in testosterone levels. In this case, a second test will not be useful.

(These is also the problem of botched tests, but let's set that aside.)

Let's look at case 1. David Pinto's post cites a study that suggests that the false positive rate can be as high as 14%. So what's the probability of two consecutive false positives? 1.96%, which is still fairly high. If we test 1000 players once, we'd get about 2 players false-testing positive twice in a row. And, given that we're testing major-league players (1200 on the 40-man rosters) and minor league players (what, around 2000?), that's around 6 false positives per test. If we test more than once per year, just multiply that.

But the problem is worse because there is natural variation. Here' Pinto's post suggests a upper-bound of 0.8% of the population producing testosterone above the Olympic cutoff. Or 8 out of 1000. That's really huge, and retesting would not change that at all, becuase the test would be accurate, but, for these players, maeningless.

The problem still exists for substances that the body does not naturally produce (all we have is case 1), but the issue of false positives remains of extreme importance.

And here I fault the media. I haven't ever seen the question raised; the reporting (and opinionating, which has been a zillion times worse than the reporting) apparently assumes the tests are 100% accurate all the time.

Rant, rant, rant.

At Wednesday, August 06, 2008 12:10:00 PM, Blogger Nanker said...

Poeple don't generally care about false positive rates -- if they did, I wouldn't have to put up with the constant wail of car alarms going off outside my window whenever a truck drives by.

It will take some kind of large financial incentive for baseball to even look into the issue -- maybe someone could sue them for defamation or something.

At Wednesday, August 06, 2008 6:23:00 PM, Blogger Phil Birnbaum said...

I'd think that if a player were falsely accused and suspended, and he knew he was clean, he would raise so much bloody hell that Bud Selig would fear for his life and those of his children.

No major league player is going to take an unjust conviction lying down.

At Wednesday, August 06, 2008 10:14:00 PM, Blogger Don Coffin said...

But, Phil, how's he going to "prove" his innocence? MLB hols up this positive test. No one talks about false positives. Everyone--except someone who knows he never used, or maybe knowing he never knowingly used, any prohibited substances--believes he's guilty.

Or look at Barry Binds. No positive tests for steroids, right? But what does "everyone" believe? What do people believe about Sammy Sosa? Or look at cycling. Lance Armstrong? No positive tests, but what do people believe? Or Floyd Landis, who did use the false positive defense? What happened to him?

You know you're clean, but everyone believes you're dirty. What di you DO about it?

At Wednesday, August 06, 2008 11:53:00 PM, Blogger Phil Birnbaum said...

>You know you're clean, but everyone believes you're dirty. What do you DO about it?

Well, suppose in your job, you got tested for steroids, and came up dirty when you're really clean, and because of that, you got fired. What would YOU do? Yell, scream, threaten to sue, demand another test immediately. If that didn't work, you WOULD sue.

As a major-leaguer, you complain to every reporter around. You say colorful things to get quoted. You show up on every sports talk show you can to discuss false positives and you're getting screwed by Bud Selig.

I can't imagine any innocent player just sitting back and taking a suspension without raising hell.

At Thursday, August 07, 2008 1:21:00 PM, Blogger Tangotiger said...

If he does have a false positive as a result of something natural in his body, then wouldn't further tests continue to confirm this? He can get quarantined every day for two weeks, tested every day, and then prove that yes with no changes in the testosterone levels, then that is his natural level.

I agree that drug testing is not the same as DNA. And it's not as bad as lie detector testing.

At Friday, August 08, 2008 11:51:00 AM, Anonymous Anonymous said...

Minor leaguers: Each major league team has about 200 players (major and minor combined) under contract, so the total is about 6,000.

At Tuesday, August 19, 2008 7:18:00 AM, Blogger Unknown said...

Thanks for all your responses to my post. Phil - what if the player is a minor leaguer and has no interested media to scream to and no resources to pay for a lawsuit?
Anyway, my point is that I don't trust the test results in any given case; maybe it is accurate, maybe it is not.

At Tuesday, August 19, 2008 8:06:00 AM, Blogger Phil Birnbaum said...

>Phil - what if the player is a minor leaguer and has no interested media to scream to and no resources to pay for a lawsuit?

That's a good point.


Post a Comment

<< Home