First, Do No Harm

Carolina Alumni Review – May_June 2018

The troubling web of opioid abuse in the U.S. was woven by just about everyone from drug manufacturers to the FDA. At UNC, a physician and a pharmacist have been at the forefront of solutions.

Dr. Paul Chelminski ‘95 (MD, ’03 MPH), a collector of stories, tells three that arc across his career. The first is about a young Peace Corps volunteer in Mali in the mid-1980s who befriended a local mother. One day he asked her how many children she had at home.

“I just have these two,” she replied, pointing to a boy and a girl.

“How many pregnancies?” he asked.

“Nine. You don’t have children?”

“No,” he replied, still making sense of her bad luck. “No, I’m not married. No children.”

“Well, then,” she said, “I wish someday that you have a good marriage. And I wish that you have lots of children, and I wish that they’re all healthy children.”

Two of nine. And she’s wishing me a happy life, Chelminski thought. That was the day he decided to become a doctor.

Many years later, in 1999, he was six months removed from residency and was staffing a Siler City emergency room when another young, impoverished mother came in with her husband and 7-year-old son. It was 2 a.m., and she thought she was dying. Her family was terrified. But the new doctor was certain she was suffering from meningitis – the less-menacing viral kind. He ran the tests and delivered the news with confidence:

“You are not going to die.”

“Sir,” the husband said, “we were told that Chapel Hill had trained a really good doctor to come down here and work in our community with us. Tonight, you have proved that.”

That was the day, Chelminski says, that he became a doctor.

The third story is different. That story is about neither the promise of medicine, as it was in Mali, nor the power of medicine, as it was for that young family. It’s a story about the limits of medicine in the face of a relentless epidemic.

It’s the story of a man so clouded by his drug addiction that he used his 13-year-old son as an ethical pawn, bringing a minor to an appointment in which Chelminski planned to confront the man about his substance abuse and pharmacy shopping.

It’s the story of a patient who sat in a Staples in Durham, cutting and pasting the safety features of one of Chelminski’s prescriptions with the intent to make copies.

“We decided we didn’t know what an abuser looked like.” 

 — Dr. Paul Chelminski ’95 (MD, ’03 MPH)

And it’s the story of a 50-year-old woman who came to Chelminski’s clinic with an extraordinary infrastructure of deception to continue her opioid habit, collecting multiple doctors for pain treatment and filling the prescriptions they wrote for her at an array of pharmacies — using Medicaid at one location and paying with cash that her church had collected for her treatment at another.

Like so many narratives of the American opioid tragedy that has taken more than 600,000 lives since 2000, the third story is frequently a tale of broken trust.

“In medicine, we train people to be empathetic to the extreme,” Chelminski said. “It’s inherently nonadversarial. With my own patients, I try to have very open and trusting relationships and, as much as I am able, to be empathetic. But when I talk to learners about this, I say that opioid prescribing is the one area of medicine where the trust between a doctor and a patient can be systematically subverted. And then it can harm public health.”

The new frontier of pain

In 2001, not long after Chelminski joined UNC’s internal medicine clinic, many of the same patients who had come through his old job at the Siler City emergency room began showing up in Chapel Hill, demanding opioids for their pain. Chelminski’s alarm bells went off, as did those of UNC clinical pharmacist Tim Ives ’91 (MPH), who also treated patients.

Five years had passed since OxyContin, a prescription tablet containing the opium derivative oxycodone, began its gallop across the pain management landscape, setting fire to long-held treatment protocols that included physical therapy and rehabilitation and lassoing primary care physicians with its promise to banish chronic pain with one pill, every 12 hours. Until OxyContin’s debut in 1996, doctors used opioids with caution in chronic pain, with a focus on addressing surgical patients’ short-term acute pain and comforting cancer and palliative care patients, for whom deadly addiction was not a concern.

But in this brave new frontier, Purdue Pharma, the makers of OxyContin, claimed that its time-release formula reduced the chance of addiction. That was good enough for the Food and Drug Administration, which allowed the company to include the statement in OxyContin’s packaging insert without providing clinical evidence.

The FDA, prompted by data that indicated some people were indeed abusing OxyContin, removed the claim from the label in 2001. But by then, Purdue’s sales force was on-message in doctors’ offices and on golf junkets nationwide, telling prescribers that the risk of addiction— even in long-term use for chronic, noncancer pain patients — was less than 1 percent. The drug’s profits ballooned from $65 million in 1996 to $1.5 billion in 2002, from 670,000 prescriptions to 6.2 million.

“Purdue was very clever in appealing to a doctor’s sense of acceptable risk,” said Chelminski, who had thoroughly reviewed the medical literature and found no clinical proof for Purdue’s sales pitch. “One in 10,000 is a rule of thumb — the same odds you learned in medical school for the risk of anaphylaxis with amoxicillin or penicillin.

“We realized that there was a certain recklessness and inattentiveness with regard to opioid prescribing and that it was a precarious situation for individual patients but also precarious, potentially, for public health.”

In response, Chelminski and Ives chartered a pain management program in 2001 to monitor and safely treat patients with chronic pain. They established protocols that were not common at the time and, for some in the medical community, controversial. The first rule was this: Anyone, including a well-coiffed 55-year-old wearing pearls and church clothing, could be abusing or misusing medication.

“We decided we didn’t know what an abuser looked like,” Chelminski said.

Everyone who came to the internal medicine clinic for pain went through procedures that included mandatory urine screens at each visit, criminal background checks for DUIs or drug-dealing and, in the beginning, frequent calls to multiple pharmacies to track a potential history of prescription forgery. If anything was amiss, those patients did not receive opioids.

Urine screens could be particularly revealing. They could indicate when a patient was taking opioids or other controlled substances that the clinic had not approved, of course. But if a patient on prescription opioids had a clean toxicology screen, that result suggested they were diverting their medication — giving or selling it to others.

“It’s pharmaceutical Whac-A-Mole. Nothing’s changed. It’s just the products have changed. We’re getting better at measuring [drug abuse]. Now we’ve got to start getting better at treatment — and treatment is not a drug. It’s mental health. People are still out there who have a medical condition, and the condition is substance dependence.”

— Tim Ives ’91 (MPH), clinical pharmacist

“Honestly, we were giving people the answers to the test,” Chelminski said. “We were saying, ‘When was the last time you took this medicine?’ If it wasn’t in their urine, they were not taking it, and a lot of people say, ‘Well, I swear I took it.’ ”

A questionably high dose of opioids for pain is about 90 morphine milligram equivalents, or MME, in a day. One patient came to the clinic with existing opioid prescriptions totaling 570 MME a day — “Enough to kill all five of the people in the room with us at the time,” Tim Ives said. “He was the new world champion.”

Through his sleuthing, Ives confirmed the man had prior convictions for drug-dealing.

Over the course of the program’s first year, from 2002 to 2003, 62 of the 196 people in UNC’s pain management program — 32 percent of its patients — committed opioid misuse in the form of abuse, forgery or diversion.

“These were patients who we saw on a regular basis, and we thought that the risk that they might be addicted or abusing the medication would be really quite rare,” said Dr. Tim Carey ’85 (MPH), former director of UNC’s Sheps Center for Health Services Research; Carey mentored Chelminski and Ives through their research. “What they showed was that it was a lot more common than we thought. It really reinforced improving our treatment protocols so these patients could get the treatment they needed and we not expose them to drugs that might make them sicker or kill them.”

Physicians had been asking the wrong key questions before prescribing opioids, Chelminski said. As often as not, the concern wasn’t just whether a patient asking for opioids would become an addict; it was whether he or she already was misusing substances or had a history of doing so, he said. Forty percent of patients who were misusing in UNC’s program had cocaine or amphetamines in their system, and misusers also were more likely to have past alcohol abuse or drug or DUI convictions.

“We were not paying attention,” Chelminski said of the medical community’s prescribing practices.

Pharma con

Purdue’s persistent marketing engine helped blur medical reason. With help from nonprofits, such as the Federation of State Medical Boards, the drug company deftly exploited for profit a physician’s default programming chip: The desire to help people feel better.

For Chelminski, the most egregious of these campaigns was “pseudoaddiction,” a syndrome in which too-cautious doctors were said to mistake their opioid patients’ “drug-seeking behavior” for addiction when the real problem was that their chronic pain was undermanaged — the solution for which was to prescribe more opioids. The same Purdue executive who coined the concept earlier in his career — in research based on a single patient who was suffering from leukemia pain — helped the state medical boards’ federation write it into its model policy. The medical boards, including North Carolina’s, adopted pseudoaddiction as a legitimate disease, despite its woeful lack of epidemiological evidence, Chelminski said.

In 2004, the Federation of State Medical Boards went a step further, recommending punishment for the “undertreatment of pain.”

Chelminski and Ives chartered a pain management program in 2001 to monitor and safely treat patients with chronic pain. They established protocols that were not common at the time and, for some in the medical community, controversial. The first rule was this: Anyone, including a well-coiffed 55-year-old wearing pearls and church clothing, could be abusing or misusing medication.

To fight a promotion machine so complete is not easy. Neither is battling the medical establishment.

Chelminski’s work earned him a spot in 2003 on then-Gov. Mike Easley’s (’72) task force on opioids; he was the only physician in the 25-person group and was instrumental in the creation of the N.C. Controlled Substances Database, which debuted in 2007. The database, in which pharmacists log prescription drug activity and doctors can check patients’ records, has made tracking abuse and misuse easier.

“We were the Starsky and Hutch on that one,” said Ives, who beta-tested the monitoring program with Chelminski. “The first day they turned it on, we caught eight patients who were doctor-shopping, going right up Highway 86, from practice to practice.”

Yet Chelminski could not convince any of the major medical journals to publish what he and Ives had learned about predicting misuse. One editor rejected their paper because he thought requiring urine screens was “coercive.”

“Coercive? I don’t think so,” Chelminski said. “There are all kinds of quid pro quos in medicines. We don’t prescribe medicines for heart failure without requiring the patients have their kidney checked because of the damage that these medicines can do. The idea that there’s a double harm here, not just to the patient, but to public health, are two reasons to require urine screens.”

And when Chelminski presented their findings at a national meeting in 2004, a physician questioned whether checking a patient’s criminal record was a step too far. Again, Chelminski was incredulous: Treatment guidelines prohibit opioid prescriptions for patients with a history of substance abuse, he said, and those patients are not known for their transparency.

“We don’t feel guilty about this,” he said. “People just lie, they deceive their own family about their drug and alcohol abuse problems. So why would they tell their doctor that they have two DUIs?”

They published their research in a “not-very-prestigious” journal in 2005. Two years later, death rates from prescription opioid overdoses were double what they were in 2002, Purdue executives pleaded guilty to fraud in federal court for misrepresenting OxyContin’s abuse potential — and the respected Annals of Internal Medicine finally extensively reviewed Chelminski and Ives’ study. Since then, hundreds of other papers have cited their work.

As a cascade of sobering news reports about the opioid epidemic flooded the public conscience in 2016, the Centers for Disease Control and Prevention issued nonbinding guidelines for opioid prescribing. They included urine drug-testing and consideration of past substance abuse — elements that Chelminski and Ives had begun using in their program 15 years earlier.

Last fall, after The New Yorker published a damning expose of the family behind Purdue Pharma and the harm their company had created, a reporter asked Chelminski if he had read it.

No, he said — and he would not.

“It’s too heartbreaking,” Chelminski said. “We’ve known about this behavior for over 10 years. What changed was that we finally had the epidemiologic evidence of the damage.”

Opioid reduction service

When Purdue discontinued its OxyContin tablet and began selling an abuse-deterrent version in 2010 — a gel formulation that users cannot crush nor easily manipulate for immediate effect — OxyContin addicts turned to heroin for a cheaper and easier high instead. (The FDA outlawed generic versions of the original tablet on the same day Purdue’s patent for it expired, in 2013.) Now, illicit synthetic versions of the opioid fentanyl, which is 30 to 50 times more potent than heroin, are making their way to American addicts from overseas.

For Tim Ives, who worked in a Utah methadone clinic during his pharmacy residency in the early 1980s, the OxyContin era is the latest in an endless cycle of drug epidemics in the U.S.

“It’s pharmaceutical Whac-A-Mole,” Ives said. “Nothing’s changed. It’s just the products have changed. We’re getting better at measuring [drug abuse]. Now we’ve got to start getting better at treatment — and treatment is not a drug. It’s mental health. People are still out there who have a medical condition, and the condition is substance dependence.”

Chelminski, Ives and colleagues at Duke and Vanderbilt are part of a five-year, $9 million clinical research grant from the Patient-Centered Outcomes Research Institute, announced last fall, that they hope will help prove the value of cognitive behavioral therapy in pain management protocols — something that has never been studied.

“We have such great needs,” said Ives, who treats mild depression and anxiety but looks to therapists and social workers for more serious disorders. “It’s almost abuse of [mental health providers], the way we have just great reliance on them.”

As for the future of opioids for chronic pain, a growing body of medical literature is hinting at just how chemically faulty their use for long-term, noncancer pain always was.

In fact, Chelminski believes he will be able to eliminate some patients’ suffering by taking them off opioids. That’s due to a condition called hyperalgesia, in which long-term opioid treatment causes a person to become more sensitive to pain stimuli; it’s a concept the PCORI study also will measure.

“In many ways,” Chelminski said, “we were fueling an epidemic of opioid-related deaths and simultaneously not improving the quality of life of the people who didn’t have adverse effects from abusing them — possibly making their functional outcomes worse.”

Although opioids still play an invaluable role in relieving suffering for acute conditions such as kidney stones, bone fractures and joint replacement surgery — and remain unquestionable choices in palliative care and battlefield trauma — “no pain medication was ever meant to carry the burden of pain management,” including over-the-counter drugs like ibuprofen, Chelminski said.

That’s why Ives views UNC’s pain management program as an “opioid reduction service,” in which 10-point-scale pain levels take a backseat to more useful information about activity, mood and daily function for patients with back pain or osteoarthritis: Can you do the physical aspects of your job? Can you get back to your gardening? Would losing weight, exercising or therapy help restore function — and decrease depression —  without drugs?

“There’s been a lot of attribution of blame in this crisis,” Ives said. “And don’t worry: There’s plenty of blame to go around, for everybody. But also, the solutions are right in front of us, just like the problems are, if we can marshal the resources to do it right.”

Chelminski, who grew up around eccentrics and storytellers, majored in medieval and Renaissance studies at Duke. He still turns to the humanities to interpret problems, which help him understand the highs and lows of human nature — how a woman could lose seven children and still exude grace; how an addicted father could hide behind his 13-year-old son; how the admonition to “do no harm” could become lost in the shuffle of care.

“People never come to you with data. They come to you with their stories. All the data in the world still doesn’t paint a clear picture of reality. Leonardo da Vinci revolutionized modern painting by blurring the edges of bodies, whereas early Renaissance painters thought they were injecting more physical realism by showing the outlines. In medicine, too, we’re often looking for definitive answers where the edges are really blurred.

“I believe the purpose of science in medicine is not so much to give us those answers as it is to protect us from false hope, to prove that there are many other pathways that are definitively wrong. With opioids, we didn’t protect people from false hope.”

This third story, a narrative that holds so many battered and broken souls damaged by OxyContin’s false promises, ends with the simplest of human truths — the one lesson that all of us too often forget:

“If something sounds too good to be true,” Chelminski said, “it probably is.”

Copyright 2018 UNC General Alumni Association