The security guard is young, maybe mid-twenties, tall and wiry and armed. He’s friendly. He asks, “Good morning, sir. Recreational? Can I see your ID?” and directs me to the left, into a zig-zag of retractable belt barriers terminating at a long, well-lit counter.
Along the wall behind the counter is a grid of peg hooks not unlike those suspending Combos and corn nuts in roadside convenience stores. The bags here are transparent, their contents green, white-labelled with names like Blue Dream, Ghost Train Haze, Kool-Aid Kush and Industrial Paint. There are orange and red liquids in clear bottles; there are foodstuffs with names like Fantastic Brownie and Cheeba Chews; there are t-shirts, hoodies and beanies monogrammed with the Medicine Man logo.
When it’s my turn, Armando Rios greets me. He has short hair, a plaid button-down and a close-cropped goatee. Until recently he was a trimmer, he tells me, but is now a budtender. I grasp this slowly, only after mentally switching out the word “bud” for “bar.”
I tell him I’d like buy maybe a sixteenth of something with a crazy name, for a story I’m writing. He asks what it’s about. I tell him it’s about a drug test developed by CU Toxicology, a University of Colorado non-profit spinoff just down the road. I tell him that, in a single nine-minute run, it can test for 112 compounds in 500 drugs with ridiculous accuracy, and that it’s cheap. I tell him that the test can detect metabolites of THC, the active ingredient in all these bags of dope, with 20 times the sensitivity of the typical pee cup, turning negatives into positives for days post-toke.
I don’t tell him the test could reveal to your employer much more than a smoke sesh, that it produces an itemized grocery list of every chemical that entered your body, or that it could change the doctor-patient relationship for good.

Ogre Indica, 1g, bought at Medicine Man, Denver, Colo.
Image: Todd Neff
Rios nods. “How about ‘Ogre’? A few of the budtenders highly recommend it. It’s more of a body high.”
Ogre. Perfect.
It’s $16 for the gram, $21.79 with tax. I pay in cash, and Rios seals the marijuana into a white, plastic mailing envelope, in a capped container that might otherwise hold takeout salad dressing. Behind me, a line of eight people has formed. It’s 11 a.m. on a Wednesday.
America is a nation of druggies. Comprising about 5% of the planet’s population, Americans smoke, pop, shoot and otherwise absorb 80% of the global opioid supply, 99% of the world’s hydrocodone (think Vicodin, Percocet, Percodan and Oxycontin) and two-thirds of the world’s illegal drugs. The personal costs can be immeasurable (see Hoffman, Seymour Philip); the economic ones borderline unfathomable, approaching $200 billion a year, according to a 2011 U.S. Justice Department report. That’s about the Czech Republic’s GDP, up in smoke, down the hatch.
Among the major types of drug testing, the first and most familiar relates to pre-employment and employment-related testing. The second has to do with addiction and recovery, and the third with helping patients manage chronic pain. The fourth category, whose profile vastly exceeds its prevalence, has to do with sports-related drug testing of the sort happening before, during and after the 4,906 drug tests were projected for the Sochi games, according to organizers. Compare that to the estimated 486 million tests in the other three categories done in the United States alone in 2011, according to the Venture Planning Group – a number the firm expects to climb to 592 million by 2016.
And depending on a drug test's invasiveness and the person or company requesting it, medical privacy hangs in the balance.

Generalized detection periods for drugs in various bodily specimens. Adapted from NHTSA graph, based on NIDA research.
Image: Mashable illustration
Drug use can be tested in a few ways, most directly via the blood. But besides the issue of having to stab someone with a needle, drugs tend to leave the blood quickly – within a few hours. Hair testing is the opposite: A 1.5-inch-long snip of about 100 hairs can spot drugs taken in the last 90 days, but it doesn’t capture anything from the past week to 10 days. Spit testing is gaining traction, but like hair, oral fluids capture a limited number of compounds at detectable levels, and they move out of the mouth in 24 to 48 hours.
That leaves urine, the industry standard now and for the foreseeable future.
Most urine tests involve the pee cup or, more technically, a urine immunoassay using thin-layer chromatography, a technology that dates from the 1970s — the same used in pregnancy tests. They’re cheap — $5 wholesale and $20-$40 at retail drugstores (look for names like “TeenSaver”). They’re also relatively non-invasive, with no hair-snipping or blood-tapping needed, and they provide immediate feedback. They generally test for between 3 and 14 drugs, in other words, for smoking pot or blowing lines of cocaine.
But pee cups have their limits. Drug detection depends on the antibodies being the right shape to lock in a narrow set of antigens – in this case, drugs. Often, they sweep up a wider variety of molecules than intended, resulting in false positives and cross-reactions. The antidepressant trazodone or even vapor sprays like Vick’s can trigger a positive for amphetamines; naproxen, ibuprofen and riboflavin can masquerade as marijuana’s cannabinoids.
The technical term here is “specificity,” which refers to the ability of a drug test to discern what’s what in a sample. For pee cups, specificity isn’t great: a 14-panel cup has just one stripe for opiates, but there are dozens of very different opiates, from heroin to oxycodone. The stripe catches some and misses others.
Another pee cup weakness has to do with its sensitivity, or ability to detect tiny amounts of drug. Often, the cutoffs are high enough that drug users can take a couple of days off and pass a pee cup test.
On top of the shaky science, pee cup drug tests lack legal punch. If the little line marked “BZO” (for benzodiazepines) disappears, it’s merely a presumptive positive. In order to fire you, bust you or throw you out of a pain management program, testers will send your pee to a laboratory, where it’s run through a machine costing several hundred thousand dollars. The machine puts the urine through two gauntlets. The first uses either liquid or gas chromatography to separate compounds based on the drug molecules’ shape and reactivity. Then the compounds are ushered into a mass spectrometer, which blows them apart with ions and samples the colors of light emerging from their death throes. The colors divulge their precise nature – and with far more sensitivity than that of the initial pee cup.
On the other hand, pee cups won’t tell your employer about the prescription medications in your medicine cabinet.

Blair Whitaker (R) and lab manager Kim Plath, at CU Toxicology in Aurora, Colo., on Feb. 3, 2014.
Image: Mashable
CU Toxicology is technically a one-person operation, “a virtual company,” as Blair Whitaker describes it. But Whitaker, 52, operates as more than one person. He talks excitedly and grasps the essence of questions before I'm completely clear on them myself. He has been a computer engineer, software engineer and venture capitalist. Whitaker has multiple titles even now: director of business operations for the non-profit CU Toxicology and CEO of the for-profit Claro Scientific Laboratories.
His successes bought a big house near the Denver Country Club, where he and his wife moved from California a decade ago to raise the three kids. In 2009, he started looking for his next thing. He met people, went to conferences, considered pitches and projects: How about an ultrasound system designed to look at glycogen levels in athletes? Presurgical nutritional supplements? Cat food infused with mouse pheromones? And so on.
Whitaker, a cyclist, had been thinking about developing a way to test nutritional supplements for a large number of potential adulterants – hidden steroids, say, or hormones or narcotics – that could get an elite athlete inadvertently busted.
Then he came across Uwe Christians.

Uwe Christians in his office at CU Toxicology in Aurora, Colo., on Feb. 3, 2014.
Image: Mashable
Christians, 51, an M.D. as well as a Ph.D. pharmacologist/toxicologist, has been doing mass spectrometry since the mid-1980s, when in his native Germany he used a version of the technique for his M.D. thesis on the metabolism of cyclosporine in liver transplant patients. He runs a lab called iC42 in a building called Biosciences East, a former Air Force optics factory.
The lab down the hall from his office hosts 14 high-end mass-spec machines and a team heavy on hand-picked Germans.
Christians had the knowledge and equipment to develop Whitaker’s sports supplement test. Using the mass-spec machines they created a single test spanning 150 compounds in a single run. It worked.
But the sports market didn’t pan out. Whitaker began to realize that World Anti-Doping Agency-type testing is a small niche. Instead, he changed course and focused on toxicology screening (drug testing), a much larger market, which happens in every pediatric and adult pain clinic, methadone clinic and addiction center. Not to mention at thousands of employment-related drug testing locations across the country.
Whitaker and Christians removed steroids and other compounds from the list and, after consulting with addiction recovery and pain management experts, added other drugs. They arrived at 112 compounds in 16 classes spanning 500 prescription, illicit and over-the-counter drugs. It was as sensitive as typical confirmatory tests but as wide as the sea. For example, if a pee cup’s marijuana stripe triggered a presumptive positive at 50 nanograms per milliliter of THC metabolite, this test would detect down to 2.5 ng/ml.
It amounted to what looked like an unbeatable drug screen. But it didn’t quite work.

The sample preparation lab at CU Toxicology in Aurora, Colo., on Feb. 3, 2014.
Image: Mashable
The problem wasn’t with the chemistry; Christians had nailed that. It was the software. In early 2011, Whitaker signed up a first customer who sent in 89 urine samples, but the reams of data that came out of the mass-spec team quickly overwhelmed his lab team.
Whitaker, who has a degree in electrical engineering, saw the mountain of data as a math problem. He spent six months trying to untangle the data using Excel spreadsheets and Visual Basic code. The result was ugly, but it worked. As a bonus, this odd marriage of analytical chemistry and signal processing also automated much of the testing effort, making the whole thing cheap and scalable.
He convinced the University of Colorado School of Medicine to launch CU Toxicology as a non-profit to host the work and provide public face for the effort – that way his for-profit Claro Labs could stay in the background, taking a royalty for every test performed and owning the software Whitaker created. By January 2014, he was running hundreds of samples a week.
Seven of them were mine.

Image: Eric Risberg /Associated Press
Since recreational marijuana became legal in Colorado Jan. 1, 2014, having a bit of weed available for special occasions has been like having a decent pilsner or single-malt scotch on hand. Guests don’t usually bring it up, but are most enthusiastic when you do. On Friday, Jan. 24 at about 10 p.m., I lit up with a friend.
I hadn’t really known what Rios the budtender had meant by Ogre being a “body high.” A friend of mine who runs a medical marijuana dispensary in Boulder, though, called Ogre “a heavy indica” and described its aftermath as “more of a couch-lock, more a foggy sort of high.”
I smoke rarely, but enough to have learned to stop at about six meaty hits from the glass pipe. The body did tingle, but the head seemed above it all – trains of thought merely derailed as opposed to disappearing into dark tunnels, as if they had never existed in the first place.
At about 3:30 p.m. the next day, I filled two of several cups Whitaker had supplied. One was an Insta-Screen 14-panel pee cup immunoassay; the other was destined for a AB Sciex 5500 tandem mass spectrometer at Whitaker and Christians' CU Toxicology lab. I put that one in the garage freezer, advising my wife to avoid using it to chill her sparkling water.

Image: Todd Neff
The Insta-Screen’s lower THC line faded in a couple of minutes. Positive, as expected. Until I ran out of CU Toxicology cups, which would happen on Wednesday, I would dutifully fill those once a day, seal their orange lids tight, and pile them on top of a bag of Ling Ling Mini Spring Rolls until taking them to CU Toxicology in bulk.
The following Monday, 2.5 days post-Ogre, I drove a couple of miles in light snow to a strip mall on Chambers Road in Aurora to a place called Wiz-Quiz, a contract employment-related sample collector for testing giants such as Quest Diagnostics and LabCorp. The franchisee proprietor, Ben Duka, was unlocking again after grabbing takeout for lunch. He had a sparse moustache and thin, gray-speckled hair. He wore a red Jeff Gordon NASCAR jacket and, on the middle finger of his right hand, a large ring with a BMW logo. He told me business is up and down depending on hiring around town.
On Duka’s walls hung 13 diplomas with citations such as “DOT Urine Drug Testing Course.”
I would do a five-panel cup – amphetamine, methamphetamine, cocaine, marijuana, opiates – for $25. In small room with a toilet, a sink, a short counter and no obvious means of drying one’s hands, I warmed the cup. Duka came in, pulled on latex gloves, checked the mood ring-like temperature readings, tore open a metallic wrapper to produce a sort of immunoassay dipstick, and dipped it in my piss. All negative. Back in his office, he swiped my credit card through an iPhone attachment and printed me a certificate saying as much.
Three days later, on Thursday, I delivered the frozen blocks urine for testing at iC42. I’d soon see how much marijuana metabolite was really in my system. I had little doubt.

The sample preparation lab at CU Toxicology in Aurora, Colo., on Feb. 3, 2014.
Image: Mashable
The development of technology like mass-spec drug tests has people both excited about its potential for addition recovery and senior care, for example, but at the same time nervous about the ethical dilemmas that involve this vast trove of personal data.
One of the primary hurdles is cost. CU Toxicology generally charges between $100 and $200 per test, depending on the arrangement and volume. Even costs associated with basic pee cup tests can add up. For example, doctors often outsource mass-spec analysis for urine samples, racking up additional costs of $50 to $250 per drug, says Tony Smith, lead lab technician for Children’s Hospital Colorado.
“So that $50 drug screen became a $700 analysis by the time it’s complete,” Smith says.
At Northstar Transitions, an addiction recovery program in Boulder, they run CU Toxicology’s test with each new arrival and then once each week they’re in the program, says Mike Ferrell, the organization’s executive director. The test has spotted cocaine in clients’ bloodstreams two weeks out, much longer than the typical three to four days the pee cup detects. The test’s breadth means Ferrell and colleagues know if someone’s been drinking, taking Xanax, spice (synthetic marijuana) and so on, without ordering costly separate tests for confirmation. It even spots nicotine.
“It takes the guesswork out,” Ferrell says. “We might see buprenorphine in a heroin addict trying to self-medicate off the street, or cocaine if they’ve been using speedballs. And that sets the tempo for how we’re going to proceed in treatment. It’s been very informative.”
University of Colorado Hospital’s Center for Addiction Recovery and Rehabilitation (CeDAR) does the test on every new patient and then again when they leave. It can make all the difference, says Michael Dinneen, who leads CeDAR’s extended-care treatment. One client, a nurse by profession, combined Benadryl and several other drugs to produce a particular high. The test spotted them. “It would have gone under the radar of all of the other tests that we do,” Dinneen says.

Lab manager Kim Plath prepares the control samples at CU Toxicology.
Image: Mashable
In addition to addiction and recovery, the test has found adherents in UCH’s pain management and chronic pain ranks. Jason Krutsch, M.D., a CU School of Medicine anesthesiologist who directs the university hospital’s interventional pain clinic, says the test helps “confirm and quantify not only metabolites of drugs of abuse, but also those prescribed in pain management. It’s useful to us to see if patients are taking the meds we prescribe for them.”
Consider also that the same capabilities enabling multiplexing drug assays can – and are – used to assess the chemical markers of all variety of activity in the human body. Christians has a 50-compound panel that tests for the overall health of the kidneys and blood vessel linings. It includes amino acid profiles, oxidative stress markers, lipids, blood sugars and more.
The CU Toxicology test is ultimately a new sort of medical tool, and so its impact could reach into realms as diverse as primary care to obstetrics. A broad-spectrum test could establish what medications a given elderly patient is actually taking – as opposed to what they say they’re taking, says Kennon Heard, M.D., a CU School of Medicine toxicology specialist and emergency physician.
“My goal in 10 years is that we’re not going to bother asking what they’re taking,” he says. “We’re going to take a urine sample.”

Image: Matt Rourke/Associated Press
Those ordering urine tests can now affordably plunge into the private habits and inner bodily workings of their targets like never before. It's akin to NSA spying, but for tapping into bloodstreams rather than data center lines.
Within the HIPAA-protected physician-patient relationship, that could well be a good thing, providing a snapshot of bodily function and pharmacological influence, nipping adverse drug interactions in the bud.
But even so, the technology raises ethical and legal questions, says Ben Rich, chairman of the Department of Bioethics at the University of California Davis. Will physicians face malpractice suits for missing some tangential finding of a broad multiplexing assay? Will the payer (an insurance company, say) insist on access to the results, too? Will there be complete disclosure to the patient of the full scope of these tests and informed consent by the patient before the drug test is given? Will patients on long-term opiate therapy really have an option to decline the test?
“What are the implications for me, vis-à-vis our relationship, if I say no?” Rich asks.
The questions become even more poignant – and far-reaching – with respect to employment-related drug testing. Should employers know if you’re taking blood pressure or anti-seizure meds? If you’re on antidepressants? If you’re taking low doses of prescribed painkillers because you pulled a muscle in your back? If you have prediabetes? Should they know if you need a little blue pill to get it up? The technology is there now, and the hundred or so bucks is a drop in the bucket compared to the $4,000 it costs, on average, to replace a U.S. worker.
Fortunately, some safeguards are in place. The Division of Workplace Programs for the federal Substance Abuse and Mental Health Services Administration (SAMHSA) officially sets the standards for employment-related drug testing across executive branch entities. Unofficially, it sets the tone for broad swath of federal, state and private sector drug testing.
Ron Flegel, the division’s director, says the mandatory guidelines are very clear both on how one can test (the standard pee cup is still the required first step, with mass-spec confirmation) and what one can test for. Though private sector employers are not required by law to follow the SAMHSA guidelines, they may find themselves on shaky legal ground if they don’t.
“It’s very strict in the sense that it’s only tested for the drugs of abuse we have in the guidelines,” Flegel says. “You can’t do DNA testing, pregnancy testing, clinical testing, genome testing – that sample is specifically authorized only for the drugs of abuse testing within our program.”

An employee trims unneeded leaves from pot plants at Medicine Man marijuana dispensary, in Denver, Colo., on Dec. 27, 2013
Image: Brennan Linsley/Associated Press
Things do change, though. Designer drugs like spice have been added to the U.S. Drug Enforcement Administration list of controlled substances, he says, and an advisory board has suggested adding synthetic opiates (oxymorphone, oxycodone, hydrocodone, hydromorphone) to SAMHSA’s mandatory guidelines list. They’re considering adding oral fluid testing and, perhaps sometime later, hair testing to the menu. Would they consider scrapping the immunoassay and going straight to a broad-based mass spectrometry test?
“You have to allow technology to move you forward,” Flegel says. “You could, possibly, have a screening instrument that could do it all-in-one.”
Existing employment-related law could provide a firmer backstop, says Matt Elliott, a labor law specialist and partner with Beckman Lawson in Fort Wayne, Ind. The Genetic Information and Nondiscrimination Act (GINA) of 2008 prohibits discrimination based on someone’s genes or genetic history, “so employers are not allowed to even obtain that information except under limited circumstances, and if they do obtain it, they’re not allowed to use it to hire, fire or determine benefits,” Elliott says.
Also, the Americans with Disabilities Act (ADA) is broad enough that abnormal cell growth or even past drug addiction are covered under the act, and employers can be liable if they discriminate on such bases, he says. The same goes for prescription drugs.
“How do you tell if it’s illicit?” Elliott asks. “It’s probably one of those things that you as an employer don’t want to know, because you’re opening your doors to liability if you do test for it.”
For his part, Whitaker is willing to let giants like LabCorp and Quest Diagnostics battle for the employment market.
“We can do employment stuff, but you know, philosophically, I’d much rather be in trying to help people than bust ‘em,” he says.
But it’s naive to believe that multiplexing drug-screening assays won’t soon be in the testing giants’ – and their many clients’ – hands. CU Toxicology may be ahead of the pack, but its scientists aren’t the only ones capable of making this goal a reality. ARUP Laboratories already has a mass-spec blood test covering 67 compounds, and a urine test covering 37.

The author Todd Neff (left) with friend Louie one day after smoking Ogre, in Aurora, Colo.
Image: Todd Neff
Nearly a week after delivering my samples to CU Toxicology, Blair Whitaker called.
“You’re not going to believe this,” he said. “You were negative.”
He emailed me the .pdf of my drug test results – my pre- and post-smoking tests on Friday, Jan. 24 and daily follow-ups through the following Wednesday, seven in all. The Friday afternoon pre-test picked up the remnants of the 100 mg naproxen pill I had popped that morning and the single beer I had sipped after my Thursday night indoor soccer game. But there was no hint of THC in my system, despite the lowly pee cup positive.
Whitaker, seemingly much more intrigued and excited than worried, came by to pick up a bud of Ogre for chemical analysis over at CannLabs. He wondered if the Ogre was all non-psychoactive CBD, a.k.a. cannabidiol, a slightly different chemical. The mass-spec, tuned for THC, would have been blind to it. The less-choosy pee cup immunoassay could have just piled CBD in with other cannabinoids of its general shape and come out positive.
But CannLabs found almost no CBD in the bud. There was precious little THC, either – about 5%, as compared to the 15-20% in most strains they test. Heather Despres, who directs the lab, said the Ogre sample was a somewhat smaller than the 180 mgs they usually set as a minimum for analysis, but that the results were probably solid.
Whitaker says 9% of CU Toxicology’s most recent 3,058 tests came back positive for THC carboxy, one of two metabolites (the other being a hydroxy) that his and other marijuana mass-spec tests look for. His test has passed every quarterly federal proficiency screen (the feds send urine) for the drug. The implication could be that casual users everywhere fail their pee cup immunoassays but then pass the legally binding mass-spec tests, he said.
“Are we missing them? I do know we have the most sensitive test for THC carboxy,” Whitaker says. “The question for us is, if we get this many and we don’t get them all, what are we missing?”
It’s going to be tough to find out. Efforts to map out the metabolic pathways of meandering cannabinoids are hamstrung by their banishment to the DEA’s schedule I list, with rogues like LSD and heroin. Winning federal funding – the lifeblood of basic medical research – to study schedule I drugs the DEA describes as having “no currently accepted medical use” is exceedingly difficult, Whitaker says. The irony that medical marijuana is legal in 20 states and the District of Columbia is hard to miss. Crystal meth and cocaine, schedule II drugs, are easier to study.
No doubt an upcoming version of the CU Toxicology screen will include CBD metabolites, prednisone and other drugs. Probably, other companies will follow. How far will this new urine-scouring technology take us? How far will we let it take us?
Regardless, the people who order these tests are going to know an awful lot about us.