·  Legal News, Analysis, & Commentary

News & Politics

Drug Recognition Experts Are Accurate, But Maybe Not Accurate Enough

— June 11, 2018

Drug Recognition Expert (DRE) Program is similar to Alcoholics Anonymous and Narcotics Anonymous, which are 12-step programs designed to assist substance abusers into recovery (and many rehabs incorporate them, such as California’s Chapters Capistrano 12 Step Program). DRE is a 12-step program to detect drugged drivers.

There’s been a lot of concern that police don’t have an easy, accurate way to test drivers for impairment due to marijuana like they do for alcohol. What’s less well known is that they don’t have a similar test for opioids, methamphetamines, or many other intoxicating substances, either.

In several states—including California, Colorado, and Michigan—trials are underway using mouth swabs and may be able to determine levels of a half-dozen drugs, not just the presence of drugs. A variety of other methods are in various stages of development, including breathalyzers.

Until then what may be the best method available originated, as so many things do, in Los Angeles, California. It may not be good enough.

According to the International Drug Evaluation and Classification (DEC) Program, in the early 1970s “two LAPD sergeants collaborated with various medical doctors, research psychologists and other medical professionals to develop a simple, standardized procedure for recognizing drug influence and impairment,” with formal recognition in 1979.

They came up with the Drug Recognition Expert (DRE) Program. While Alcoholics Anonymous and Narcotics Anonymous are 12-step programs designed to assist substance abusers into recovery (and many rehabs incorporate them, such as California’s Chapters Capistrano 12 Step Program), DRE is a 12-step program to detect drugged drivers.

The National Highway Traffic Safety Administration (NHTSA) came on board in the 1980s. Now DREs are certified by the International Association of Chiefs of Police (IACP), which sets uniform standards for these DREs.

A DRE is an officer trained to observe and evaluate driver behavior behind the wheel and outside the vehicle using 12 steps. Though some deride them as “drug whisperers”, their track record seems to be around 85 %.

They conduct a 12-step exam, sometimes called a Drug Influence Evaluation (DIE) or, confusingly, a Drug Recognition Exam (DRE), which includes checking blood alcohol content, examining eyes for pupil size and other conditions, typical drunk driver balance and coordination tests (walking, standing on one leg, finger-to-nose), looking for track marks indicating intravenous drug use, and urine, blood, or saliva tests, then form an opinion.

Roadside intoxication test; image by Jeffrey Smith, via Flickr, CC BY-ND 2.0, no changes.
Roadside intoxication test; image by Jeffrey Smith, via Flickr, CC BY-ND 2.0, no changes.

Three supportive studies are often cited, two funded by the NHTSA and the Department of Transportation (DOT), and one by the Arizona Governor’s Office of Highway Safety. IACP and NHTSA support the International Drug Evaluation and Classification (DEC) Program. There are DREs in all 50 states plus Canada.

But Atlanta, Georgia TV station WXIA Channel 11’s 11 Alive Investigators pointed out that NHTSA was behind two of those supportive studies – “the same federal agency pushing the drug recognition expert program” – while “independent studies” rated a DRE’s accuracy as only about 50-50.

The main critic of DRE is Denver malpractice expert witness Greg Kane, MD, though he claims “being an expert witness in DUI and DUID cases is not what I do for a living.” He regards his criticism as a civic duty.

In a paper for the Journal of Negative Results in Biomedicine, Kane found the three cited studies were subject to multiple biases and don’t validate that the results can be generalized. He doesn’t deny that some form of DIE could be of value but says the current test isn’t correctly conducted or interpreted.

Most of the time they seem to get it right. But what about when they get it wrong? One DRE in Cobb County, Georgia, had a success rate reported at 85 out of 90.

The American Civil Liberties Union of Georgia has filed suit on behalf of four (originally three) people who were arrested and accused of driving under the influence based solely on the analyses of that same Cobb County DRE.

These analyses were not borne out by blood tests, so the suspects were released, but the police department seemed to regard this as letting them off on a technicality. It claimed that the DRE was right and the tests were wrong, that the cannabis left the blood before the testing could be gone. One suspect was told a urine test would have shown she was using marijuana.

DREs are trained to believe that their opinions are better than lab results. According to 11 Alive, DRE instructors teach that “The laboratory is not perfect and the toxicologists won’t always be able to corroborate your opinion, even though your opinion may be accurate.”

That stuff is nonsense. Cannabis remains in the body long after intoxication ends – days, weeks or even months afterward. True, THC does remain in urine longer than in blood, but in that case, why didn’t they test urine? In some European countries, urine first and then blood are tested, with any trace evidence enough for a charge of driving under the influence.

This false confidence in the face of the facts may be the best argument against the use of DREs.

The ACLU charges that Cobb County doesn’t even use the full “12 step protocol,” which it says already is “riddled with flaws, based on discredited studies, and irresponsibly entrusts police officers with performing essentially medical or scientific tests.” (One reason might be that only full evaluations are reported to federal databases.)

The Canadian Broadcast Corporation’s The Fifth Estate also found the tests are prone to bias and false positives, further noting that six US states’ courts have refused to let DREs testify as experts, including California, the birthplace of the DRE. A Maryland court stated that DRE training “does not enable DREs to accurately observe the signs and symptoms of drug impairment,” so DRE conclusions aren’t accurate or reliable.

In Canada, where recreational marijuana is expected to be legalized later this year, the Supreme Court says DREs are allowed to testify as experts. A 2013 article from the Canadian Society of Forensic Science Journal concludes that “drug evaluations conducted by DREs in Canada are accurate.” Our neighbor to the north is planning to greatly increase its DRE presence.

Few tests are 100 % accurate, drugged driving is a bad and increasing problem, and these suspects were released. Police and politicians could argue that no harm is done.

They would be wrong.

  • The DIE test can last 45 minutes and is extremely subjective. Told to walk an imaginary line, the DRE gave one suspect demerits for an “improper number of steps.”
  • It is scary to be arrested, especially if you’re innocent, even if it’s only overnight. You feel like a character in a Kafka novel, especially if the police won’t believe the blood or urine test results. Sandra Bland committed suicide while in police custody when a traffic stop escalated. The officer wasn’t a DRE, and drugs weren’t suspected, but it’s still an object lesson.
  • It can be expensive. One of the plaintiffs in the federal lawsuit had legal fees in the thousands of dollars because of her arrest.
  • One of the Cobb County suspects lost her job because she was arrested on a drug charge, costing her an alcohol server’s license.

Part of the problem is that the suspects were arrested for minor traffic violations such as briefly crossing or touching the lane line. That wouldn’t normally even rate a ticket unless it was part of an observed pattern of behavior.

In their panic to cope with new, unfamiliar, and difficult to detect intoxicants, law enforcement is putting too much faith in a questionable test. Better put on the brakes.

  • several states:
  • other methods:
  • 85 %:
  • 12-step exam:
  • Greg Kane, MD:
  • paper:
  • filed suit:
  • shown she was using:
  • any trace:
  • noting:
  • article:
  • Sandra Bland:

Join the conversation!