Alaska News

Calibration error puts DUI tests in question

Nearly 2,500 breath-alcohol tests conducted in drunken-driving cases across the state over a period of nearly four years used equipment with flawed quality control standards, which some defense attorneys say call into question the accuracy of the tests themselves.

At least one attorney is preparing to challenge his client's conviction based on the new information, arguing the mistake means the breath test that played a central role in his client's case can't be trusted.

State officials acknowledge the error but say it did not affect the breath-alcohol level DUI suspects blew, only the quality-control checks that accompany every test. Only two of the cases involving the equipment may have been affected by the error and they are under review by the Department of Law, according to state officials.

But while the breath-alcohol content levels may not have been affected, some attorneys say problems with the instruments' quality testing could raise doubts about the accuracy of the BAC results.

"Every good forensic system has to have quality-assurance systems in place," said defense attorney Fred Slone, who specializes in drunken-driving cases. "There's a reason for that quality assurance: to increase the confidence that you have in the result you get. Garbage in, garbage out is another way to look at it."

The problem involves dry gas tanks, also known as "alco bottles," used in confirming the accuracy of the DataMaster instruments, which drunken-driving suspects blow into to determine breath-alcohol content. The tanks contain a known sample of alcohol that the instrument measures before and after every test to ensure it is functioning properly.

When the state crime lab prepares the bottles, the alcohol levels are measured 10 times and the results are entered into a spreadsheet, which calculates the average value. Because air pressure can affect the test results, the average value is then adjusted to a standard barometric pressure from what it was on the day the tank was tested.


But back on Feb. 16, 2006, a trainee and an analyst thought the correction factor didn't appear right and decided to invert the fraction, said Orin Dym, director of the crime lab. And because it was in a spreadsheet that calculated the values automatically, no one noticed until Dec. 8, 2009.

"It's a long period of time," said Capt. Dave Koch, captain of patrol at APD. "And it is part and parcel to every prosecution, other than if someone refuses (to take a breath test)."

By the time the error was discovered, 48 of the 663 tanks prepared in the years involved had been affected, according to the state. That means 2,465 tests were conducted with equipment that was prepared with the inverted fraction and that had been corrected for the barometric pressure in the wrong direction, according to the Department of Public Safety.

"That information was discoverable to every defendant who blew into the machine and ended up convicted or plead out to a DUI," said Rex Butler, an Anchorage defense attorney. "The defendant has the right to call into question the validity of the breath test. But if you don't know that there's a problem, you can't."

According to a July 26 letter from the attorney general to the directors of the Public Defender Agency and the Office of Public Advocacy, the lab determined the error was significant in only two cases, both from 2006, in which the quality assurance test fell .001 outside the correct target value.

"The BAC results that come out of the DataMasters are what they are. They're not affected by the incident here," Public Safety commissioner Joe Masters said. "We're confident in those results and that they're accurate. And we're also confident that the potential affecting of cases is limited to two."

Michael Logue, of the firm Gorton, Logue & Graper, represents one of those defendants and said Thursday that he was preparing to file a motion to withdraw the man's no contest plea based on the new information.

Logue asserted that the breath-testing instrument never would not have measured the sample from his client if the tank were properly calibrated the day he blew into the instrument. Since then, his client has had his license revoked and had to pay high-risk insurance, Logue said.

"I'm expecting that the city will allow us to get a fresh start on that case," Logue said. "I know the city prosecutor's office and APD are very concerned about this and are not happy with what's going on. I've dealt with them for 20 years or more and they really put a high price on integrity. They're the ones who kind of discovered this problem and took it to the Department of Public Safety."

Al Patterson, the municipal prosecutor, said an Anchorage police officer going to court on a DUI case noticed changes on a calibration document for a DataMaster machine, so prosecutors brought it to the attention of Alaska State Troopers. Troopers investigated the matter and determined nothing illegal had happened but that the changes were made because of the error at the lab, he said.

That was in June, said Dym, who wrote a memorandum about the situation to Robert Gorder, deputy commissioner of the Department of Public Safety, that is dated July 9. But the lab first noticed the problem on Dec. 8, 2009, a day with pressure low enough that a tank failed a test, he said. The lab's staff realized it was a calculation error on Dec. 23.

"They did bring it to my attention," Dym said. "They said that they had a problem with the barometric pressure calibration change."

The calculation was corrected and the affected tanks were recalled, but word never got out about the error because lab officials didn't realize the potential legal significance, he said.

"Is this a scientific issue? No," Dym said. "I mean, we're dealing in these really tiny numbers that didn't make a difference on anyone being tested or anyone's breath test results. But, I think there was a failure to recognize that we touch the legal environment and that it's not about, 'Is it a scientific issue?' It's a legal issue."

Patterson said the error to him did not appear to be legally significant, though he thought prosecutors would likely end up dismissing the case against Logue's client in light of the recent revelation. Eventually someone is going to have to challenge the matter in court and let a judge decide how serious the mistake was, he said.

"It's a question of confidence, you see. I can't tell you where it's going to go," Patterson said. "It could be miniscule. It could be nothing at all or just a tempest in a teapot or it could cause a lot of paperwork, which will result in really nothing happening, too. It just depends on how it's approached."

Butler, the defense lawyer, questioned how anyone can trust the lab's own assessment that only two cases were affected by it.


"Look at this potential scenario: Suppose all those people turned around and sued the city or the state of Alaska for that in a class-action," he said. "Because why? All the money I had to pay, you took people's cars and sold them. You put me in jail without me having the opportunity to challenge my test because I didn't know that somebody had messed up in their calculations for calibration."

Find James Halpin online at or call him at 257-4589.