Defense attorneys point out breath test flaws
NEWARK -- Would you trust a machine that is banned in other states for
unreliability to test your blood-alcohol content?
The state of Ohio does.
In 2011, Licking County law enforcement agencies started using the Intoxilyzer
8000, a breath-alcohol testing machine manufactured by CMI Inc. in Owensburg,
Ky., Newark Law Director Doug Sassen said. Most used it along with the
BAC Datamaster, its predecessor manufactured by National Patent Analytical
Services in Mansfield.
Defense attorneys have raised several concerns about the reliability of
the machine from its record-keeping to the possibility a smartphone could
alter a blood-alcohol content test.
To make the Intoxilyzer 8000 portable, CMI Inc. used a different light
source, detector and filters. The 8000 uses a pulsed light beam that measures
at four points per second, compared to 40 points per second on the Intoxilyzer
5000 or about 100 points per second on the BAC Datamaster, defense attorney
Robert Calesaric said.
"Because of the longer pauses, it does not allow as accurate of a
reading," Calesaric said.
The Intoxilyzer 8000 is more prone to picking up mouth alcohol or acid
reflux that have higher levels of alcohol. Its limited pulses might not
pick up a spike indicative of something other than deep lung air.
Before and after two samples are taken from an individual's breath,
the machine tests a dry gas control. Differences in the dry gas should
indicate an improper test, Calesaric said.
In a March breath test for a Newark man, the machine was off by .013 for
dry air, according to the subject test report. He was tested again 13
minutes later despite the discrepancy and cited for operating a vehicle
Mary Martin, program administrator for the Bureau of Alcohol and Drug
Testing, said she thought the department had removed a couple machines
for similar issues.
"If it's out of range, we'll remove that instrument,"
The machine from the March test still was in use at the Ohio Highway Patrol
as of June.
The Intoxilyzer 8000 also allows for a 20 percent difference between the
two samples before rejecting the test. That means a .080 result could
be between a .06 and .10 -- a 40 percent difference.
Martin said .02 was a difference accepted by national scientific research.
Calesaric thought a .005 difference would be fair.
"If the machine is accurate, it's accurate. A 5 percent error
rate is high enough," Calesaric said.
In addition to being portable, the Ohio Department of Heath touted the
Intoxilyzer 8000's ability to transmit data electronically as one
of its greatest benefits.
Agencies using the Intoxilyzer's predecessor, the BAC Datamaster,
would record the results of each test and keep a log at the department,
said Martin. Now, all data is transmitted to the Bureau of Alcohol and
Drug Testing, where it aggregates reports in a searchable database on
This will allow the state to target areas with a large number of operating
a vehicle while intoxicated citations, said Martin, adding that no research
has been started yet because not all machines have been distributed.
Defense attorneys are concerned data can be easily deleted and there is
no policy in place to store records necessary for trial as required by
Ohio Administrative Code, Calesaric said.
In result on the website, a tested listed a blood-alcohol content of 23
-- more than 38 times the lethal limit. It was removed from the website.
The CMI software allows anomalies, such as a 23 BAC, to be replaced with
other data and thus hides "inconvenient information," Athens
County Municipal Court Judge William A. Grim wrote. ODH switched software
in May to alleviate any problem, Martin said.
"It was a software issue," Martin said.
Grim found that the disappearance of data was, at best, an indication
the website is still a work in progress, and at worst, a manipulation
to hide adverse information.
"If it is the purpose of ODH to have a comprehensive database, that
purpose has not been achieved," Grim said.
Calesaric said no policy is in place to store these records. Martin said
the department is required to keep them for three years but plans to store
the information indefinitely.
"We have no plans to get rid of them," Martin said.
Another concern is how the machine operates around radio frequency interference
from cell phones or Blackberries. Thomas Workman Jr., who has testified
for defense attorneys as an Intoxilyzer 8000 expert, said interference
might skew results from .09 to .20. Ward acknowledged interference from
a Blackberry, but former Chief Toxicologist John Kucmanic found it "impractical"
to test all frequencies of available smart phones.
CMI did testing that found cell phones do not interfere with the Intoxilyzer
8000, Martin said. A representative testified the company didn't test
smartphones or PDAs.
Another vulnerability is the machine will detect more alcohol the longer
a person breathes into it. The person should stop when the progress bar
reaches 100 percent, but the sample is not collected until the person
Defense attorneys are concerned officers could manipulate the system to
yield a higher blood-alcohol content sample.
"It's not consistent between each person," Calesaric said.
Inconsistent does not necessarily mean inaccurate, Martin said.
"You really can't blow higher than what's in your lungs," she said.
Local law enforcement said they hadn't had any problems with the Intoxilyzer
8000. Sassen said the arguments from defense attorneys are typical of
any new technology.
"I don't think there are problems with it," Sassen said.
Martin said she trusts the extensive research the Ohio Department of Health
did into the Intoxilyzer 8000 to determine its reliability.
"It has time and time again shown that it is reliable, it is accurate,"
Grim did find that the Intoxilyzer 8000 could be used to determine blood-alcohol
content and have those findings accepted in court.
"There is no such thing as a perfect person, a perfect machine or
a perfect computer operating system. All have limitations or vulnerabilities,"