After the field sobriety test and the ride to the police station, comes the make-or-break moment for the thousands of Connecticut motorists arrested each year on suspicion of drunken driving: blowing into the Intoxilyzer 5000.

A reading from the state's breath-testing device is the chief weapon for police officers in proving that someone was driving under the influence.

But in a daylong hearing Tuesday at the Department of Motor Vehicles in Wethersfield, three defense lawyers argued that the device is fundamentally imperfect in measuring blood-alcohol content and can even be biased against women, small men and people with respiratory conditions such as emphysema and asthma.


The state's chief toxicologist defended the accuracy of the test in testimony Tuesday.



But a University of Washington expert said physiological factors can cause readings to vary widely for a single person — for example, a higher percentage of alcohol for suspects who hold their breath before exhaling and a lower one for suspects who have hyperventilated.

First offenders who refuse to submit to a chemical test face an automatic six-month license suspension under state law.

If the lawyers ultimately convince a Superior Court judge that the Intoxilyzer 5000 is inaccurate — a charge state officials and DMV attorneys reject — the ruling could affect DUI cases pending across the state. It also could force Connecticut police departments to take blood or urine samples from suspected drunken drivers, a significant change in policy and everyday practice.

"The state will have to throw out 169 $5,000 machines," said defense lawyer Jay Ruane of Ruane Attorneys at Law, a Connecticut firm that focuses on DUI cases and takes on about 500 new ones a year.

Ruane considers blood samples to be "the gold standard" in intoxication testing: "Because if you want to know what's in somebody's blood, why would you test their breath when you can test their blood?" he asked.

Judge George Levine ordered the DMV hearing a year ago in Superior Court in New Britain while presiding over an administrative appeal of two men who had their driver's licenses suspended after failing Intoxilyzer 5000 tests. The two men, represented by the Ruane law firm, challenged the legitimacy of the machine, which the state certified for law enforcement use in 2001.

Levine said he wanted more evidence and sent the case — now involving four plaintiffs — back to the DMV and DUI attorneys to argue six points before a hearing officer. Among the major issues Levine seeks to determine are what the machine measures, how it does that and whether the method falls within state law and public safety regulations.

On Tuesday, the state's chief toxicologist, Dr. Robert Powers, testified that not only is the breath test a fair and widely accepted way of determining blood-alcohol concentration — 0.08 percent or higher is illegal in Connecticut — but that the Intoxilyzer 5000's readings are "probably an underestimate" of someone's true alcohol level.

The machine measures the amount of ethyl alcohol in a person's air sample, which is believed to be equivalent to the amount of alcohol in the person's blood if both were tested at the same time, said Powers, who served as the DMV's expert witness.

But Dr. Michael Hlastala, a professor and researcher in the School of Medicine at the University of Washington in Seattle, disputed that equivalency, calling it an "old paradigm" of thought held by the nation's forensic toxicologists.

Hlastala argued that the alcohol recorded in a breath test does not come from a deep part of the lungs, as Powers testified, but from the surface tissue of a person's airway. Hlastala depicted breath-alcohol testing as inherently flawed after citing research stating that the level of alcohol rises as a person exhales and continues to exhale.

"If you have alcohol in your system and you breathe in different ways, either before the test or when you breathe in for the test, you'll change the alcohol reading," Hlastala said. "For example, hyperventilation. That'll reduce the amount of alcohol," because breathing rapidly "flushes away" the alcohol that was stuck on the airway's surface, he said.

Holding one's breath will elevate the alcohol reading, he contended, because the alcohol is being charged before the breath is released.

Hlastala also argued that a people who have a smaller lung volume — such as women or people with asthma and chronic bronchitis — will yield higher blood-alcohol readings because they need to exhale a greater fraction of available breath for the machine to register a reading.

The hearing is expected to wrap up this morning. After compiling all the evidence, Hearing Officer William D. Grady will have 90 days to make his recommendations to Levine.

The judge will then decide on the license suspensions and, ultimately, the Intoxilyzer 5000's validity.