Ed Brayton is a journalist, commentator and speaker. He is the co-founder and president of 
Dozens of convictions for drunk driving are suspect now that a routine audit of a crime lab in Colorado Springs showed that the results of those tests were typically overstated by the lab's testing.
Prosecutors have begun contacting lawyers for 82 defendants whose drunken driving charges were based in part upon incorrect blood alcohol tests by the Colorado Springs police crime lab.In each of the cases, test results reported by the forensic chemist unit of the Metro Crime Lab were higher that the actual results, police officials disclosed Friday.
This will likely lead to many of those convictions being overturned or, at the very least, the sentences reduced. The good news is that the local police and prosecutors seem to be taking this seriously and looking to do the right thing:
"We're not going to be relying upon any questionable blood alcohol content results," he said. "The District Attorney's office and the Colorado Springs Police realize how serious it is and we're acting accordingly.""We don't want to treat anybody as guilty if they're not," he added...
At the Police Department's request, the Colorado Bureau of Investigation will conduct an external and independent investigation of what went wrong.
The Police Department's Internal Affairs Unit also is conducting its own investigation.
"If there are corrections we need to make, we'll make them," said interim Deputy Chief Rod Walker.
Whitlock said the on-going investigation is trying to determine if they were the result of a human error or a procedural error.
As a result of the discovery, the crime lab is re-analyzing about 1,000 blood alcohol test results taken since January 2009.
A perfect example of why such audits should be mandatory, not only for blood alcohol tests but for a whole range of other forensic lab tests as well. People's lives are on the line and these tests are treated as the final word; it is absolutely crucial that they be done correctly. Kudos to the Colorado Springs law enforcement community for being on the ball and doing the right thing.
Politics
Comments
Of course, DWIs are more likely to catch better class of people than say a no knock warrant searching for marihuana, that drunk is respected member of the community. (snark)
Posted by: The Pale Scot | December 19, 2009 9:12 AM
Although this is ancient history by now, we should remember the O. J. Simpson case where his defense attorneys pilloried the LAPD crime lab for sloppy practice. Anybody remember how those attorneys were trashed by the media and accused of being shysters? Maybe Barry Scheck and company were just a little ahead of their time!
Posted by: SLC | December 19, 2009 9:33 AM
What tests are they using? Why can't they back it up with Cr6+-Cr3+ testing? (Cr6+ is oxidized by ethanol to Cr3+)
Posted by: Katharine | December 19, 2009 10:40 AM
The saddest thing about this is how rare the opinion "We don't want to treat anybody as guilty if they're not," seems to be in the law enforcement community. Kudos to Colorado Springs for getting it right.
Posted by: penn | December 19, 2009 11:30 AM
Given that at least one breathalyzer manufacturer (CMI) refuses to allow examination of its device's source code, is it any wonder that some results are inaccurate?
The best defense against DUI is, of course, not to drive while impaired. We should however expect that the devices used for measuring BAC and recording the results as legal evidence are properly designed, programmed, calibrated, and operated. Too many assume the magical blinking box is always right.
Anyone having experience with software and test & measurement equipment understands the difficulty of getting reliable & repeatable results. Compound those problems with making a durable device for field use as well as the requirements for producing & handling evidence, and you have a very very difficult design problem on your hands. It's not impossible, but it requires far more transparency to assure than most old-school small hardware vendors are willing to pony up.
We see the same resistance to transparency and defenses of trade secrecy and appeals to righteousness ("Trust us - we're the good guys!") with electronic voting machines. Somehow when the skeptics can get in to do a decent test, they manage to find critical problems.
So yes, it's vital that we pay attention to these vendors, the men behind the metaphorical curtain. I am convinced that prohibiting examination of source code does more damage to society by shielding incompetence from scrutiny than it helps by protecting real innovation and trade secrets. I base this on my experience reverse-engineering and examining the source code of vendors and by reviewing my own employer's proprietary code, not out of any FSF dogma.
Posted by: Bob | December 19, 2009 11:54 AM
But if it saves the life of one child, isn't it worth punishing 1,000 innocent adults?
Posted by: soboco | December 19, 2009 12:02 PM
One of the problems that people overlook when sloppy police work is exposed is that people who were actually guilty will benefit from the reduced or overturned sentences.
This is certainly not as shocking as the idea of innocent people who were wrongfully convicted, but it's another type of injustice for which shoddy police work is responsible.
Posted by: barry21 | December 19, 2009 12:12 PM
I used to run legal blood alcohols ... and we were regularly blind tested by the state. We never knew if the sample was from a real arrest or was a QA sample sent in the same kind of packaging.
It wasn't a difficult test, except for the speed you had to work because the EtOH drops quickly after you open the sample tube.
Posted by: Tsu Dho Nimh | December 19, 2009 12:49 PM
If a person gets behind the wheel of a vehicle with any alcohol in his system whatsoever, he poses a risk to others and deserves punishment for his reckless behavior. I don't really care about arguments over minuscule fractions of a percent of BAC levels on one side or another of an arbitrary line.
Posted by: Pat Donohue | December 19, 2009 2:46 PM
Mr. Teetotaling Homophobe doesn't seem to understand the fact that you don't get drunk from one beer.
Posted by: Katharine | December 19, 2009 10:00 PM
Or that you don't necessarily. If you're a small, thin woman you may well; a guy built like Ed or I, not so much.
Of course, if this "Pat Donohue" is that same "Pat" Ed's been kicking around, he's about 6 inches tall, so HE would...
Posted by: Azkyroth | December 20, 2009 1:00 AM
Wouldn't the blood tests have only been done after the driver failed a breathalyzer? What this means is that guilty people will be getting off because of improper lab work. Don't be happy because "innocent" people are being freed.
Posted by: Tom | December 21, 2009 12:51 PM