With breath testers in doubt, Vermont prosecutors to toss dozens of DUI cases

Editor’s note: This story is by Andy Bromage, a staff writer at Seven Days. It was first published on Seven Days’ website.

Judging from the cruiser-cam video, the case against accused drunk driver Brian Grenier looked like a slam dunk.

When a Washington County sheriff’s deputy stopped him for a busted headlight on North Main Street in Barre last May, Grenier admitted he’d been watching a Celtics game with a buddy, drinking Bud Lights.

Asked by the officer to stand on one leg and count backwards, Grenier swayed like a tree in the wind.

Back at the police station in Montpelier, Grenier blew into the DataMaster DMT breath tester. His blood alcohol registered .172 — more than twice the legal limit of .08.

Grenier, a 50-year-old electrician with four prior drunk-driving convictions, was cited for his fifth DUI — a felony punishable by up to five years in jail. On the video, you can hear him muttering, “I’m screwed.”

Under normal circumstances, Grenier likely would be. But his case is anything but normal. Grenier is one of three codefendants in a court case that’s blowing the lid off Vermont’s alcohol-testing practices.

Grenier’s lawyer is David Sleigh, a tenacious and crafty criminal defense attorney based in St. Johnsbury. Together with Burlington lawyer Frank Twarog, Sleigh is using Grenier’s case and those of two other DUI clients to attack the credibility of the DataMaster breath testers used by police and the state health lab that certifies and maintains them.

Their star witness is Darcy Richardson, a former state chemist turned consultant who has supplied the lawyers with reams of internal health-department memos and potentially damning testimony about malfeasance she claims to have witnessed inside the Vermont Department of Health Lab. Richardson and state chemist Amanda Bolduc, who remains employed at the health lab, complained to supervisors last year that a coworker, lab technician Steven Harnois, used unorthodox methods to repair DataMasters and to get them to “pass” routine performance checks over a period of years [1]. The health department investigated Harnois last year and cleared him of wrongdoing.

Since Seven Days first reported the story last month, the problem has snowballed. On the last day of the legislative session, lawmakers in Montpelier passed fast-tracked legislation, at Gov. Peter Shumlin’s urging, that will move the alcohol-testing program from the Department of Health to the accredited forensic lab at the Department of Public Safety — in an attempt to restore the program’s credibility.

Meanwhile, Sleigh and other criminal-defense attorneys report that DUI cases are settling “fast and furious” as a result of questions raised by Richardson about the breath-testing program.

Why? Sleigh now has proof the DataMaster used at the Vermont State Police barracks in Royalton was improperly installed and had been malfunctioning for almost a year [2] before it was fixed last month. As a result, dozens of drunk driving cases in Windsor and Orange counties are now being reopened or tossed. Prosecutors say some drivers whose licenses were suspended may get them back immediately.

There’s more. A Seven Days investigation shows that numerous other DataMasters deployed to police departments around the state were set up incorrectly or passed routine performance checks when they should have failed. In each case, multiple individuals — including police and health-lab employees — signed off on the inspection.

Public records show that last summer in Winhall and Manchester, breath testers were installed without activating the sensors that would detect interfering compounds that could skew results. The breath tester at the Northfield Police Department was installed in December without properly setting up its temperature-monitoring function. Each of those problems was fixed within a day or two, before any breath samples were taken.

But at the state police barracks in Derby, a breath tester that failed a routine performance check was not pulled from service. Using that slip-up to his advantage, Sleigh recently argued on behalf of three DUI clients. The prosecutor tossed their civil license suspensions.

It’s unclear how many other machines around Vermont have been similarly compromised. The Vermont Department of Health did not answer questions about the scope of the problem by deadline.

To Sleigh, the revelations are evidence of the state’s “incompetent, arrogant and unethical” breath-testing program. They raise legitimate questions about whether innocent drivers have been convinced of DUI based on faulty evidence.

Equally troubling, though, is the prospect of dangerous drunk drivers getting off the hook and back behind the wheel.

In Vermont, drivers charged with DUI face two penalties — a criminal misdemeanor or felony charge, and a civil license suspension. License suspensions are based solely on the breath test and a state chemist’s sworn statement affirming that a breath-testing machine is accurate and reliable.

A breath test isn’t necessary to secure a criminal conviction if there’s sufficient evidence of impairment — for instance, a video showing the driver stumbling, or empty beer cans littering the car. Criminal proceedings can often take several months to resolve. Vermont adopted the civil suspension system to get dangerous drivers off the road immediately.

The DataMaster has been effective in that regard. At least Windsor County State’s Attorney Bobby Sand still believes in the machine. “In my experience, the instrument is really good,” he says. “It is human error that has caused many of the problems.” Although Sand believes he could establish the Royalton instrument’s reliability in a court of law, he says that would be “unfair.”

“I don’t want to fight that fight because I think government has an obligation to do things the way they say they’re going to do them and that didn’t happen in this case,” Sand says. “We missed the mark on this instrument, and there should be some consequence for that.”

Because the breath tester wasn’t set up properly, Sand intends to reopen 40 DUI cases where the driver was prosecuted using the Royalton DataMaster and toss the civil license suspensions. In addition, he has sent letters to those individuals with criminal convictions, informing them of their right to appeal.

Drunk driving is a huge problem in Vermont, and one that has eluded easy fixes. Aside from probation violation, no single offense consumes more of the criminal justice system’s time than DUI, says Defender General Matthew Valerio.

According to crime statistics published by the Vermont Center for Justice Research, there were 3287 drivers arrested for driving under the influence in 2009 — a rate of nine per day — and more than a third of those were repeat offenders.

Every year, roughly a third of all traffic deaths in Vermont are determined to be alcohol-related. In 2007 that number was 20. For young drivers ages 15 to 20, the percentage of fatalities that are booze-induced climbs even higher.

For more than two decades, the state’s weapon in the war on drunk driving has been the DataMaster, a breath-testing instrument that uses infrared light to measure a driver’s blood-alcohol level. Vermont first started using the DataMaster, manufactured by Ohio-based National Patent Analytical Systems, in the late 1980s. The machine was a workhorse and, despite a series of legal challenges brought by defense attorneys, was repeatedly upheld by the Vermont Supreme Court as a reliable and accurate instrument. And unlike the old “crimper” breath testers, which captured a sample of breath and physically shipped it to a lab, the DataMaster produced results instantly, making it faster and easier to disable dangerous drunks.

But by 2005, the DataMaster fleet was aging, and the Vermont Department of Health began testing new instruments to replace it. At the time, Darcy Richardson was a chemist in the state health lab in Burlington tasked with testing and evaluating various instruments for evidentiary use. According to Richardson, that’s when the trouble began.

Internal memos indicate Richardson and fellow state chemist Amanda Bolduc documented all kinds of problems with the machines. The first batch of 10 DataMasters was returned to the manufacturer because they wouldn’t turn on, Richardson says. Others arrived with loose screws rattling around in boxes. In another instance, the health lab returned a DataMaster that emitted a plume of smoke after it was turned on.

Richardson says she encountered so many problems with the new DataMasters that she ultimately recommended her bosses purchase another instrument. Bolduc, who is presently the state’s sole expert witness on the DataMasters, concurred.

But the supervisors overruled the two chemists, in part because Vermont courts trusted the DataMaster brand; it was the only breath-testing instrument they accepted for evidentiary purposes. So the state purchased a fleet of new, improved touch-screen DataMaster DMTs, with many more bells and whistles than the old DataMasters. They’re currently in use at 67 police agencies around Vermont.

Sand describes the device as “gold-standard technology.”

Richardson stayed at the health department until last September, when she left to start her own consulting firm. As sole proprietor of Milton-based Vermont Forensic Services, she’s been an expert witness in numerous drunk-driving cases. Richardson, who is married to a South Burlington cop, shared her detailed accounts of alleged problems in the health lab with any defense lawyer who hired her.

In January, Richardson penned a three-page affidavit for one of Twarog’s cases that chronicles all the DataMaster’s failings she had observed, both during the test phase and after they were being used to collect evidence. In one such example, Richardson observed that the reported alcohol value on the instrument’s screen was not the value always printed on the ticket.

Prosecutors and judges around Vermont have been disclosing Richardson’s statement to defense lawyers as possible “exculpatory” evidence — that is, discoverable evidence that could be helpful to the defendant in a criminal trial.

Numerous defense attorneys tell Seven Days that information supplied by Richardson has helped them settle DUI cases for lesser charges.

“I’ve had several cases in which prosecutors have resolved cases with an understanding that there is a significant question regarding the accuracy and reliability” of the instrument, says Paul Volk, a well-known criminal defense attorney from Burlington.

Health department spokesman Robert Stirewalt previously told Seven Days there were software issues with the new DataMasters, but nothing that would affect the accuracy of a subject’s breath sample. And while the manufacturer continues to make “hardware modifications,” Stirewalt said, all instruments in service for evidential use “have met and continue to meet” the performance standards established by Department of Health rules and regs.

That was on April 4. On April 27, Department of Health Commissioner Harry Chen was summoned before the Senate Judiciary Committee in Montpelier and revealed that a DataMaster at the state police barracks in Royalton has been malfunctioning for almost a year — and no one in the health department had noticed.

“I’m just wondering,” said Senate Judiciary Committee chairman Dick Sears (D-Bennington) at the hearing, addressing Chen. “Are there other problems out there that aren’t going to come to light until some employee or former employee speaks up?” Days after that, lawmakers amended the drunk-driving bill — now law — to move the alcohol-testing program from the Department of Health to the Department of Public Safety.

To understand what went wrong with the Royalton DataMaster, it helps to understand how the machines work. Before a suspect blows into the tube, the machine conducts a series of self-tests — first, a “blank” test using ambient air from the room, followed by a simulator test using a premixed solution of known alcohol content.

If everything checks out, and the driver supplies an adequate breath sample, the DataMaster will produce a printout that shows results of the self-tests and the breath test. However, if the alcohol content of the simulator solution falls outside an acceptable range — which could indicate a problem — the machine is supposed to shut down and abort the test. What that happens, the DataMaster printout indicates “simulator out of range.”

That failsafe is referred to as the “tolerance” function, or “accuracy test.” It has to be activated as part of the software setup before a DataMaster can begin collecting breath samples.

In Royalton, the tolerance function was not activated during setup in May 2010. Apparently, no one from the health lab or the state police noticed until April 17, 2011, when Sleigh pointed it out.

Sleigh became aware of it because his law partner, Corby Gary, had a DUI client who was prosecuted using that instrument last year. When Gary reviewed the DataMaster printout, he noticed the simulator solution had been outside the acceptable range — but the breath test was accepted, anyway. The machine had failed to abort. The case ended up settling last September for other reasons, Sleigh says, so he never brought up the technical problem.

“We didn’t understand the significance of it at the time,” Sleigh says. Even if he had understood, Sleigh maintains he was under no legal or ethical obligation to reveal the problem to authorities. “It’s the state’s burden of proof. My ethical obligation is to the client. The government is in charge of taking care of itself.”

Public records show that three different health-lab employees signed off on the Royalton DataMaster: State chemist Amanda Bolduc certified the instrument and Richardson signed off on Bolduc’s paperwork. Technician Steve Harnois, the health lab’s go-to guy for trouble shooting the breath testers, installed it at the barracks.

Richardson describes it as a “perfect storm of dropping balls” by lab employees under the gun to get the newly purchased DataMasters tested, calibrated and deployed into the field to meet a federal grant funding deadline. Health department officials have since said new procedures are in place to ensure such an oversight couldn’t happen again — essentially a checklist of set-up options.

But for prosecutors in Windsor and Orange counties, that horse has left the barn.

Last week, Orange County State’s Attorney Will Porter sent out letters to nine people in his jurisdiction that were charged with drunk driving based on tests from the Royalton DataMaster. Rather than tossing all the civil license suspensions, as Sand is doing in Windsor County, his deputy says they’ll evaluate each case individually and only grant reprieve to those where it’s clear the machine malfunctioned.

“If we don’t see a scientific basis for it being an inaccurate result, we’ll still use it,” says Orange County Deputy State’s Attorney Michael Kainen.

Practically speaking, winning the civil case after the fact will mean little for drivers who have already been inconvenienced by a suspended license. Those who are mid-suspension could get their licenses reinstated, Sand says. Others could be relieved of having to purchase special automobile insurance and participate in the Project CRASH driver-rehabilitation program — which together cost hundreds of dollars.

For drivers convicted on the Royalton DataMaster, that’s all the free justice they’ll get. Neither Sand nor Porter plans to offer any other form of restitution. Sleigh suggests that, with a lawyer’s help, affected drivers could petition the courts for redress. But he isn’t sure their awards would even be enough to cover attorney’s fees.

Prosecutors aren’t the only ones reopening old files in light of the DataMaster debacle. Public defenders such as Dan Sedon of Orange County are digging into old cases, too, and concluding that, in some instances, prosecutors are underestimating the number of impacted people. Drivers who don’t request a hearing on their civil suspension might not show up on the prosecutor’s rolls, says Sedon, in which case there could be “several dozen” impacted drivers in Orange County, rather than nine.

Sedon isn’t reassured by Bobby Sand’s assertion that the Royalton DataMaster was producing valid tests even with the tolerance function switched off. Sand maintains that if a DataMaster printout shows no “simulator out of range” warning, then it can be assumed it’s working correctly.

It’s crucial for the instrument to be able to distinguish naturally occurring compounds such as acetone from blood alcohol, Sedon says. “There’s likely acetone on your breath right now. If you’re fasting or have certain medical conditions, it can be significant,” he says. “We’re not talking about something insignificant here. This is the admissibility of evidence that people’s freedom can depend on.”

What’s more, Sedon recalls that when Vermont first started using DataMasters 20 years ago, officials reassured everyone by saying that the device self-tests, so it won’t produce a result if it’s not functioning properly.

“The fear in all defense attorneys at that time was, What if the malfunction is that the device is not telling us that it is malfunctioning?” Sedon says. “It appears we’ve found that malfunction.”

On a recent Friday morning, not far from where Brian Grenier was pulled over for DUI, Sleigh is making his case in a windowless courtroom in Barre. In a smooth baritone, Sleigh is pleading a motion to compel half a dozen health lab employees to testify under oath about the breath-testing program.

Seated at the opposing table is Stuart Schurr, Vermont’s traffic safety resource prosecutor, and the health department’s lawyer, Assistant Attorney General Margaret Vincent. Schurr says the state has turned over every discovery document Sleigh has asked for and doesn’t oppose him questioning the health lab officials.

The judge gives Sleigh until June 1 to get the depositions done and come back with motions to dismiss or suppress evidence. That’s the point at which Sleigh, depending what he finds, will move to throw out the breath tests against Grenier and his codefendants.

Schurr says he can’t really discuss the situation outside of the courtroom. But during a phone conversation, he suggests there’s a lot more to the story that will come out as the case proceeds. Plus, he says, the DataMaster has withstood many challenges before. It’ll survive this one.

Maybe so. But as Orange County public defender Dan Sedon points out, you don’t need to invalidate the entire breath-testing program to win. All you need is some well-placed reasonable doubt.

“Juries are going to hear about this,” Sedon says. “This really does a lot of damage to the state’s position in terms of prosecuting DUIs. Any way they want to spin it, in any juror’s mind now, there’s less scientific reliability around this device.”