The Marijuana Breathalyzer’s Uncertain Future

The test could help reduce the number of intoxicated drivers on the road—or it could turn up too many false positives to be useful.

AP Photo/Eduardo Verdugo
Add to Briefcase
Thor Benson, The Atlantic
Dec. 28, 2015, 12:12 p.m.

Mike Lynn is cau­tiously ex­cited. A 49-year-old fam­ily man who lives in the Bay Area, Lynn has been an emer­gency-room doc­tor at High­land Hos­pit­al for most of two dec­ades. His com­pany, Hound Labs Inc., has just de­veloped the tech­no­logy to pro­duce a breath­alyz­er that can de­tect marijuana, but its fu­ture is un­cer­tain.

Lynn, who has some ex­per­i­ence in­vest­ing in bi­otech com­pan­ies, is also a re­serve deputy sher­iff for Alameda County. To­geth­er these ex­per­i­ences led him to won­der if tech­no­logy might play a role in pre­vent­ing ac­ci­dents caused by drivers un­der the in­flu­ence of marijuana. No pur­it­an, Lynn also wants to make sure law-abid­ing weed smokers won’t get in trouble for the wrong reas­on.

“It doesn’t make any sense at all to ar­rest some­body that smoked last night and is no longer im­paired, on the road,” Lynn told me. Cur­rent tests used by law-en­force­ment agen­cies de­tect THC through blood, saliva, and ur­ine, but THC can stay in the body for days, weeks, or even a month, so it’s ex­tremely dif­fi­cult to tell if someone is high based on the THC found in those tests. States like Wash­ing­ton and Montana lim­it THC levels to five nano­grams per mil­li­liter of blood, which many ar­gue does not rep­res­ent someone who is still im­paired by their marijuana use. Pennsylvania has it lim­ited to only one nano­gram.

“Typ­ic­ally, if you have five nano­grams in a reg­u­lar smoker, you prob­ably won’t see any be­ha­vi­or­al ef­fects,” said Carl Hart, a pro­fess­or of psy­cho­logy and psy­chi­atry at Columbia Uni­versity. “Where­as, with five nano­grams in someone who’s nev­er smoked, you might see a lot of ef­fects.”

With breath, Lynn ex­plains, the THC sig­nal only sticks around for a few hours. Some stud­ies have found that THC stays on a per­son’s breath for as long as four hours. The com­pany doesn’t yet know if the breath­alyz­er can de­tect if someone ate a marijuana product re­cently.

We have no idea what’s in the brain based on some measure in your mouth or in your lungs. 

The path to cre­at­ing a breath test has not been simple or easy. Lynn said THC is found in quant­it­ies “some­where between a mil­lion and a bil­lion times less con­cen­trated” than al­co­hol in breath. To cre­ate a test that could isol­ate THC in breath and meas­ure it ac­cur­ately, Lynn en­lis­ted the help of a team of sci­ent­ists led by Dan Fletch­er, a pro­fess­or at the Uni­versity of Cali­for­nia, Berke­ley who spe­cial­izes in bioen­gin­eer­ing. They looked at everything from large devices used to de­tect small quant­it­ies of sub­stances to scan­ners that look for residues found in ex­plos­ives to try and find a solu­tion. Even­tu­ally, they figured out how to fil­ter breath through the right chem­ic­als to cre­ate a re­ac­tion so a device can tag and count THC mo­lecules.

Lynn wouldn’t dis­cuss the ac­tu­al chem­istry, say­ing that it was “pro­pri­et­ary.” He said the pro­cess has been suc­cess­fully tested in the lab, and that his com­pany will start tri­als at the Uni­versity of Cali­for­nia, San Fran­cisco, and San Fran­cisco Gen­er­al Hos­pit­al early next year. They will also be test­ing ver­sions of the breath­alyz­er with law en­force­ment around that time, and they hope to have the product ready for de­ploy­ment by the end of 2016. Ini­tial law-en­force­ment tests will be car­ried out in Oak­land and Berke­ley.

Hart ex­pressed skep­ti­cism about the idea be­hind the tech­nique, call­ing it a “dumb idea.” “The chem­istry of al­co­hol is such that there is es­sen­tially no blood-brain bar­ri­er for al­co­hol … so what is in your lungs or blood is the same amount, typ­ic­ally, as what’s in your brain,” he said. “With marijuana or any oth­er drug, we can’t do that. We have no idea what’s in the brain based on some meas­ure in your mouth or in your lungs.” He said the THC doesn’t trans­port it­self along the blood-brain bar­ri­er like al­co­hol, so what’s in the breath prob­ably doesn’t rep­res­ent what’s in the brain.

Many out­side of Hound Labs are con­cerned people who may have smoked re­cently, for med­ic­al pur­poses or oth­er­wise, might get labeled as in­tox­ic­ated when they aren’t. “If you were to take some­body that is a heavy can­nabis user, they’re cer­tainly go­ing to have a much high­er test res­ult than someone who oc­ca­sion­ally uses it … How do you quanti­fy that?” said Jason Thomas, a former Den­ver de­ten­tion of­ficer and Mar­shal’s Deputy. Tol­er­ance will also be dif­fi­cult to as­sess.

Spike Helmick, a former high­way-patrol com­mand­er who has been com­mis­sioned by three gov­ernors to help de­vel­op Cali­for­nia’s DUI stand­ards, told me Cali­for­nia of­ficers have been want­ing a way to test if someone is driv­ing high for at least 15 years. He said he be­lieves we can get to a point where we es­tab­lish a prop­er level of THC that rep­res­ents a state of in­tox­ic­a­tion.

A fed­er­al study done by the Na­tion­al High­way Traffic Safety Ad­min­is­tra­tion this year found driv­ing high from marijuana is sig­ni­fic­antly safer than driv­ing drunk, but both are worse than driv­ing sober. Hound Labs also in­tends to cre­ate a ver­sion of the product that can be used by ci­vil­ians for self-as­sess­ing their THC levels. Lynn said many people don’t know how po­tent the marijuana they’re buy­ing or shar­ing is; they will want to know if they’re at a safe level.

“The real­ity is that over 100 mil­lion people in this coun­try have leg­al ac­cess to marijuana, so let’s ac­know­ledge that,” Lynn said. Wheth­er his product will make things any bet­ter for them re­mains to be seen.


Welcome to National Journal!

You are currently accessing National Journal from IP access. Please login to access this feature. If you have any questions, please contact your Dedicated Advisor.