This spring, 16 state patrol officers from Colorado and Wyoming took a couple days off their usual work schedule to do something special. They assembled in a hotel conference room in Denver. As instructed, they wore street clothes for their first assignment: going shopping at nearby marijuana dispensaries.
"It's a brave new world," said instructor Chris Halsor, referring to the years since Colorado legalized recreational marijuana.
There are now more marijuana dispensaries in Colorado than there are Starbucks shops, said Halsor, a Denver lawyer and former prosecutor. And though consuming cannabis is legal across the state, driving under its influence is not.
The cops in that conference room, with their buzz cuts and Mountain Dew, are all part of the force charged with keeping the roads safe. But first, they needed a formal pot education — to learn how to identify various marijuana products and paraphernalia when they pull over a driver they suspect is under the influence.
Here's the rub: Despite the increasingly legal use of cannabis in many states, cops still don't have the equivalent of a reliable alcohol breathalyzer or blood test — a chemically based way of estimating what the drug is doing in the brain. Though a blood test exists that can detect some of marijuana's components, there is no widely accepted, standardized amount in the breath or blood that gives police or courts or anyone else a good sense of who is impaired.
A number of scientists nationally are working hard to create just such a chemical test and standard — something to replace the behavioral indicators that cops have to base their judgments on now.
"We like to know the human error and the limitations of the human opinion," said Tara Lovestead, a chemical engineer at the National Institute of Standards and Technology in Boulder, Colo., who is working on setting standards for what a marijuana detection test might require.
It's actually really hard for Lovestead to do this kind of research because she works in a federal lab; federally, cannabis is considered a Schedule 1 substance, "a drug with no currently accepted medical use and a high potential for abuse." So even though Lovestead is in Colorado, getting hold of a sample for research purposes is just as hard as getting hold of heroin.
"We cannot use the stuff down the street," she said.
Aside from being a bureaucratic mess, coming up with a standardized blood or breath test is also a really tricky chemistry problem because of the properties of the main psychoactive chemical in cannabis: delta-9-tetrahydrocannabinol, or THC.
In states like Colorado, there is a THC blood test that law enforcement can use to show "presumed" impairment. If a person has more than 5 nanograms of delta-9-THC per milliliter of blood, a court or jury can infer that they are impaired, according to Colorado law (this is called "permissible inference" in legalese).
But Lovestead and others maintain that, scientifically speaking, that cutoff doesn't actually mean anything.
"We just don't know whether or not that means they're still intoxicated, or impaired or not," she said. "There's no quantitative measure that could stand up in a court of law."
Turns out it can be a lot harder to chemically determine from a blood or breath test that someone is high than to determine from such a test that they're drunk.
Ethanol, the chemical in alcoholic drinks that dulls thinking and reflexes is small and dissolves in water. Because humans are mostly water, it gets distributed fairly quickly and easily throughout the body and is usually cleared within a matter of hours. But THC, the main chemical in cannabis that produces some of the same symptoms, dissolves in fat. That means the length of time it lingers in the body can differ from person to person even more than alcohol — influenced by things like gender, amount of body fat, frequency of use, and the method and type of cannabis product consumed.
In one study, researchers had 30 frequent marijuana users stay at a research facility for a month without any access to drugs of any sort and repeatedly tested their blood for evidence of cannabis.
"And it shocked everyone, including ourselves, that we could measure, in some of these individuals, THC in the blood for 30 days," says Marilyn Huestis, a toxicologist with the University of Maryland School of Medicine who recently retired from leading a lab at the National Institute on Drug Abuse.
The participants' bodies had built up stores of THC that were continuing to slowly leech out, even though they had abstained from using marijuana for a full month. In some of those who regularly smoked large amounts of pot, researchers could measure blood THC above the 5-nanogram level for several days after they had stopped smoking.
Conversely, another study showed that people who weren't regular consumers could smoke a joint right in front of researchers and yet show no evidence of cannabis in their blood.
So, in addition to being invasive and cumbersome, the blood test can be misleading and a poor indicator of whatever is happening in the brain.
Recently, some scientists have turned their attention to breath, in hopes of creating something useful.
A number of companies, like Cannabix Technologies and Hound Labs, are in the process of developing breath detection devices. Tara Lovestead is providing the data that will help relate the concentration of THC detected in the breath to what's in the blood. Even though blood provides an incomplete and indirect inkling of what's happening in the brain, it's the measure law enforcement turns to as a benchmark.
That, too, is a chemist's nightmare. THC and other cannabinoids — the chemicals that cause a high — are really squirrelly. They degrade quickly and appear only in very tiny amounts in the breath.
Luckily, Lovestead's specialty is detecting tiny amounts of chemicals in the air. She and her colleagues have worked on methods to use tiny air samples to detect evidence of arson, buried bodies and hidden explosives. Marijuana is the next challenge.
In the future, she said, an accurate breath test would likely involve looking at a lot more than just THC — probably a whole combination of chemicals.
"One thing to look for would be metabolites — something that comes out of the breath that shows it actually went through your system," she said. Such a test would greatly reduce the possibility that someone might test positive from inhaling secondhand smoke, she said.
In the meantime, it's up to law enforcement officers like the ones in Chris Halsor's class to make the call, based on circumstantial evidence and their best guess.
"The whole point of this class is to get the officers to make correct decisions," said Halsor.
Many officers in his courses have never used marijuana — or haven't since some exploratory puffs in high school. These officers need training, he said, to boost their confidence — "confidence that they're making the right arrest decision and confidence that they're letting people go who really aren't impaired."
The cops attending his seminar in the spring paged through Dope Magazine, chuckled at a photo of an edible called "reef jerky" and watched a video together on how to dab — heating concentrated marijuana and inhaling the vapors. In their visit to a local marijuana dispensary, they examined gold-plated blunts — hollowed-out cigars filled with marijuana.
But the real test of these officers' ability to identify the signs of cannabis impairment faced them outside the hotel, in a parked RV that was plastered with bumper stickers.
Four volunteers for the project were inside the RV, legally getting as high as they wanted to, from a big plastic tub full of pot products.
"Good music, good company, good weed. It all goes together," said Eugene Butler, one of the four volunteers.
Butler and the three others had never met before. They had volunteered to get high and then interact with cops to help the officers learn the signs of cannabis impairment.
"We're going to willfully smell like pot around a bunch of cops," said Sharica Clark, laughing.
Inside the hotel, the officers practiced roadside sobriety tests on the four volunteers — determining each time if, in real life, they would have arrested these people for a DUI.
All the volunteers had smoked a lot of pot inside the RV. But in the sobriety tests, they performed differently.
A volunteer named Christine, for example, did well on math, quickly calculating how many quarters are in $1.75. But she didn't do well on other things, like balancing, remembering instructions and estimating time. (She was concerned about recrimination at work, and NPR agreed to use only her first name).
Christine, the officers all decided, would be a danger behind the wheel. In real life, they would have arrested her.
"Yeah, she'd be going to jail," said Rich Armstrong, an officer with Colorado State Patrol.
But things weren't so clear with the other volunteers. A lot of the officers had decided they wouldn't arrest Eugene Butler or a volunteer named John (who also asked that we not use his last name); both men aced the same roadside tests Christine flunked, even though they, too, had just smoked a lot in the RV.
And when it came to Sharica Clark, the officers decided it was essentially a toss-up as to whether they would have arrested her, based on her performance on the roadside tests. Yes, her pupils were huge, and she had a tough time touching her finger to the tip of her nose while her eyes were closed. But her balance, counting and recitation of the alphabet were, as Colorado State Patrol Officer Philip Gurley put it, "spot on."
"It was a tough one," said Tom Davis, another officer with Colorado State Patrol.
Right now, these officer's opinions loom large. If they decide you're driving high, you're going to jail. But at the end of the day, they're just making educated guesses. Two different officers could watch the same person doing the same sobriety test and make different decisions on whether to arrest. In previous courses, officers had decided that a volunteer was impaired when in fact the volunteer hadn't smoked at all.
So, just like the THC blood test, the judgments officers make can also yield false positives and negatives.
"This is one of those subjective areas," said Armstrong.
"It's too subjective," said Lovestead.
She recently published a paper in the journal Forensic Chemistry where she found the vapor pressure of THC — one of its fundamental physical properties. Lovestead believes finding and standardizing that measurement is a small but significant step toward a more objective route for evaluating intoxicated drivers.
In the meantime, courses like Halsor's are the best resource for officers. And at least now the class participants know what pot strains like Skunk Dawg, Hippie Chicken and Chunky Diesel actually smell like.
"Yeah," said Gurley. "It smells like the bottom side of a rock."