Alaska News

How do Alaska police test for marijuana in drivers suspected of DUI?

This week, we'll examine a two-part question from Roger about cannabis intoxication and impaired driving. It's been a hot topic in states where cannabis is now legal, and laws, policies and technologies are constantly evolving across the nation. The situation in Alaska is relatively stable because it apparently hasn't changed very much since voters approved the legalization of recreational pot. The Legislature has taken no action, and the ballot measure itself didn't include DUI details. Roger asks:

If a person is stopped by police, and suspected of impaired driving with cannabis, what test can police do that doesn't simply find traces in the blood or urine that could have already been there for weeks? Would this mean a person would have to wait a month before driving again, legally? With alcohol, there is a percentage limit that can be tested. Will they ever be able to do that with cannabis, to actually treat it "like alcohol?"

As Roger rightly notes, some marijuana tests detect past use, but don't test for intoxication. Without getting technical here, urine tests look for a version of THC which no longer has an active effect, and which the body can store for over a month or more depending on individual factors. Tests seeking that substance are commonly used by employers with zero-tolerance drug policies, and because those tests don't test for intoxication, they're different from the ones used in court on DUI charges. So, no, people won't have to be afraid of waiting to drive until all traces of marijuana consumption have left their bodies to avoid risking a marijuana DUI.

Read more Highly Informed: Seeking answers to Alaska's cannabis questions

The chief intoxicating element of cannabis is delta-9 THC, or for our purposes here, "active THC." Basically it's THC that has not been used by the body yet. The only currently drop-dead accurate test to determine how much of it is in someone's system within the narrow window of time associated with suspicion of impaired driving is to draw blood and test it. Blood tests in that case do test for active THC, not the inactive kind that sticks around for a while. But, for reasons we'll get into later, that level by itself doesn't prove intoxication on a basis as reliable as a similar level for alcohol does.

Charges of DUI strictly involving cannabis are rare in Alaska. Megan Peters, public information officer for the Alaska State Troopers, said in the last two months, troopers have charged only one driver with driving under the influence of marijuana. According to a troopers dispatch, that driver was a 55-year-old Eagle River man who got pulled over because his passenger wasn't wearing a seat belt. That court case is pending, Peters said.

When asked whether Anchorage police had charged anyone with a marijuana DUI in the past two months, Chief Mark Mew didn't mention any specific cases but said, "We suspect marijuana involvement in approximately two dozen cases over the last couple months. This is a loose estimate, and the numbers cannot be confirmed until lab results are back. However, this number (and corresponding lab work) is based on cases in which suspects provide a breath (alcohol test) result below .08. It is not currently possible for us to estimate or know the total number of drivers who are impaired by both alcohol above a .08 and marijuana."

How do police judge marijuana impairment?

Mew wrote in an email that the investigation of a suspected marijuana DUI typically follows a protocol for APD officers:


An officer observes signs of impairment in a driver. The driver fails "standardized field sobriety tests" conducted at the scene. The officer arrests the suspect and takes him/her in for a breath test. The driver blows either a 0.00 (scenario A) or some value under the per se limit for alcohol (scenario B). So we now suspect alcohol's role in the impairment is either non-existent (example A) or amounts to a partial contributor (example B). But the driver is obviously wasted, so some other substance is involved. Is it pot? Some other illicit drug? A legally prescribed drug? Paint fumes? A seizure or diabetic episode? We ask more questions, which the driver may or may not choose to answer. We call a Drug Recognition Expert (DRE) if we have one on duty. The DRE may be able to ferret out the likely answer. Ultimately, we apply for a search warrant to get (and later analyze) the driver's blood. That is the only way we will know for sure what has been ingested. We take the driver to the magistrate, not knowing the results of the blood test. We explain whatever we do know and what we are waiting to learn. The magistrate considers all this in determining what kind of bail and conditions to impose.

Peters in an email described troopers following essentially the same process.

Compared to alcohol breath tests, blood tests for cannabis are expensive and something of a hassle for law enforcement because of special protocols, storage, delay of test results and extra training needs. In Alaska, drawing blood for determining impaired driving requires consent or a warrant unless there are other factors beyond suspicion of cannabis intoxication. Alaska law issues driver's licenses on the condition of implied consent for alcohol breath tests under suspicion of DUI, and for blood or urine tests for any controlled substances in the case of an accident that causes death or serious injury. But the state's driver manual doesn't mention cannabis by name in the section explaining implied consent laws.

All of these reasons mean there is strong incentive for technological innovation. Once the breath test for alcohol became standardized and states set blood alcohol limits, known as "per se" limits, police and citizens alike gained a clear standard to make decisions on. But that doesn't exist yet at a statewide level, and according to Mew, it would have to be a change made at that level.

Can pot have a legal limit like alcohol?

Colorado, along with other states, has set a per se limit of active THC at 5 nanograms per milliliter of blood. But that ratio means different things to different people. Many smokers may be intoxicated at that point, but many heavy or frequent users, including medical patients, may feel perfectly sober there. But everyone, no matter body type or experience level, is a certain, quantifiable amount "drunk" with a blood alcohol content of .08 percent.

The per se law in Colorado has led to complications for law enforcement and citizens, a phenomenon some call "sober DUIs" because the ratio of intoxication to blood THC level depends largely on individual use patterns.

The cannabis critic for the Denver weekly Westword (and a medical card holder in Colorado), for instance, tested at more than three times his state's legal limit after abstaining from all forms of pot overnight plus 15 hours. That bears repeating: He tested at nearly three times Colorado's legal limit for active THC in the blood, but he was completely sober.

There are saliva tests available, but they're not as reliable as the courts would need them to be for evidence of intoxication. A Canadian company says it has developed a prototype pot breath test machine, but details are sparse. If a reliable and accurate breath test were ever developed and put widely in use, like the one in place for water-soluble alcohol, marijuana's fat-soluble nature may mean evidence vulnerable to challenge. But who knows how any of this will play out in the courts.

All of this means, in my opinion, that setting per se limits for pot pose a big challenge. Unless future technology can reliably and accurately test for active THC and correlate it to impairment based on a known and regular formula, society won't be able to treat marijuana strictly "like alcohol" when it comes to DUIs. Still, because of the great need among citizens and law enforcement to determine what amount of THC in the body is safe to drive with, and then find a cheap and reliable way to test for it, science and entrepreneurs are searching for the magic bullet.

Time will tell however, and because of THC's behavior in the body, the most reliable future test of impaired driving may not even end up being one of blood. Because of the expense and hassle of blood tests, if someone could figure out a way to get blood out of the cannabis DUI enforcement equation, it could be worth a mint.

To that end, researchers at Arizona State University have developed a smartphone app they say is a new use of technology already in use to detect Alzheimer's and Parkinson's and one Marines are adapting to train helicopter pilots. The app uses a phone's camera to measure the tiny, involuntary movements of the eyes to determine how impaired someone is. It will soon be studied in the field, according to ASU, in one study with users to see if the app has an effect on their decisions whether to drive or not. Another study will work with police to "detect what drug is in a person's system and the degree of impairment," starting with marijuana.

Are stoned drivers a problem in Alaska?

According to the Christian Science Monitor, in Washington in 2013, the first full year of legalization there, drivers testing positive for marijuana jumped nearly 25 percent over the previous year, but there was no corresponding increase in car accidents or fatalities. Intoxicated driving of any kind is never a good idea, but to complicate things even more, transportation studies have found that including losing some reaction time, stoned drivers tend to recognize their own impairment and compensate for it.

So while the world researches the problem and the Legislature approaches the end of its session, what enforcement tools are needed right now according to authorities?

Mew said that an implied consent law for both blood and breath would help but that it would have to pass muster in the courts and Legislature. He and Peters both said that their agencies' marijuana DUI enforcement efforts would benefit from some sort of easy, accurate and reliable test, either of breath or blood. But Mew added that in the meantime, an expansion of refrigeration capacity would help APD store test samples, as well as an expansion of training programs for patrol and specialist officers who encounter drugged drivers.

Mew said that wish list was part of what APD has learned from other agencies around the country, but another need the department heard of has less to do with enforcement and more to do with gathering data: "Reprogram your reporting systems to capture the marijuana and drugged driving statistics that everyone is soon going to demand."

That data would be of use to policy makers in Alaska, for nothing else to determine the real scope and nature of problems stoned drivers pose on Alaska roadways.

Have a question about marijuana news or culture in Alaska? Send it to with "Highly Informed" in the subject line.