Austin attorney questions breathalyzer accuracy

A local criminal defense attorney is concerned about the accuracy of breathalyzer devices used to determine whether a driver is intoxicated. 

This comes after a New York Times investigation found 30,000 breath tests have been thrown out by courts in New Jersey and Massachusetts because of human error and lax governmental oversight. 

“Innocent people wind up getting convicted,” said George Lobb, a criminal defense attorney in Austin.  

Lobb doesn't trust that Texas isn't having those same issues. 

“Here in Austin, we have two different labs that have two different sets of standards, criteria, and protocol that they use to test breath and blood,” Lobb said.  

Police use breathalyzer machines to determine if someone's blood alcohol content is over the legal limit, which is .08 in Texas. In the Lone Star State, the Intoxilyzer 9000 has been used to analyze breath samples since 2015. 

“I've been asking for the raw data and the code used not only in the intoxilyzers but in the gastomata graphs used by the two labs here in Austin,” said Lobb.  

The code can be used to determine if programming compiled the right data, while raw data uses graphs to show what substances were detected. Lobb said his requests for that information have been continually denied in court. 

“Why don't they release the code? Why don't they release the raw data? If they're not doing anything wrong, why are they hiding it from the public?” said Lobb.  

The Texas Department of Public Safety operates the Texas Breath Alcohol Testing program, which established how breath alcohol testing is to be administered and regulated. 

DPS’ communications department sent this statement on the accuracy of their results:  

"Breath alcohol tests are conducted at evidential testing locations throughout the state. These tests are conducted by certified breath test operators under the supervision of qualified forensic scientists who calibrate the breath testing instruments in accordance with accredited procedures utilizing certified reference materials." 

Lobb isn't convinced. 

“DPS opted to leave out the histogram function of the Intoxilyzer 9000. Why's that important? Because the histogram tells you the rate of flow, the volume of flow and whether any mouth alcohol got into the machine that would skew the results,” Lobb said.  

He also worries crime lab workers could use different protocols, and food, other substances, or even health conditions could affect breath tests. 

“Those with diabetes can produce a blood alcohol concentration up to .03, so you're already a third of the way to .08 completely sober because of the way your body processes sugars,” said Lobb.