I have been listening to The Fighter Pilot Podcast because my fantasy career would have been to fly a jet fighter plane (not even remotely possible, given my constitution). I learned that when an aircraft accident occurs, a mishap board is convened, not to assign blame but to try to learn what went wrong and avoid another mishap.
You Might Also Like
Explore This IssueNovember 2022
We should apply the same process to medical practices that were once considered good medical practice, but later were deemed undesirable. The liberal prescribing of high-dose chronic opioids is a good example.
When I began medical school in 1976, many things we now take for granted did not yet exist: computerized tomography and magnetic resonance imaging scans, effective medications for peptic ulcer disease and viral hepatitis, laparoscopic surgery, monoclonal antibody therapies, third-generation cephalosporins, the diagnosis of human immunodeficiency virus and its therapies, and on and on.
The chronic administration of opioid pain relievers would also be on that list. In the fullness of time, we have all realized that prescribing opioids over the long term in high doses for chronic pain produced some serious problems for some patients. I have given a lot of thought to the question of how We—the medical profession and society in general—could have gotten this so wrong.
Blaming pharmaceutical companies for what happened is very much in vogue, and no doubt they deserve some of that opprobrium. However, in my view, that is a vast oversimplification of what happened. Although I recognize that physicians are unconsciously influenced by the pharmaceutical industry, I believe those of us who prescribed chronic opioids did so because we thought we were acting in the best interests of our patients.
A Complicated Situation
What follows is my personal recollection of how this complicated situation evolved over several decades. I think it is critical that those trained in the past 10 years or so understand how this story unfolded: “What were they thinking?”
Before and during medical school, I understood that administration of opioids for short courses was appropriate for post-surgical pain or trauma. As a teen, I had several painful kidney surgeries and learned that a shot of meperidine could dramatically relieve severe pain within a few minutes.
My first encounter with chronic use by a patient was in the fall of 1980, during my medical internship. A slight, elderly lady with advanced multiple myeloma was hospitalized under my care; her oncologist had prescribed methadone for severe bone pain. One morning when we could not awaken her, we realized that her methadone levels had accumulated to toxic levels. After we administered naloxone, she woke up, and this sweet little old lady briefly became a raving maniac.
That was around the time that it had become common practice to prescribe long-acting opioids for patients in the final stages of cancer to relieve the pain of expanding lesions in bone and soft tissue. Because life expectancy was limited, physiologic dependency or addiction was not of concern. A letter in the New England Journal of Medicine in 1980 carried the headline, “Addiction rare in patients treated with narcotics,” and this concept was widely accepted, despite the fact that the assertion was not supported by evidence.1
In 1982, I began the clinical year of my rheumatology fellowship and was shocked to find that our revered senior clinician frequently prescribed propoxyphene napsylate with acetaminophen (Darvocet-N) to severe chronic rheumatoid arthritis (RA) patients.
Propoxyphene is a synthetic compound chemically related to methadone approved by U.S. Food & Drug Administration (FDA) in 1957—five years before evidence of efficacy was required in 1962. Of course, this was premethotrexate/pre-early administration of disease-modifying anti-rheumatic drugs (DMARDs), and these were patients with extensive damage and deformities.
I observed that his patients did not abuse this analgesic and that it did not have the upper gastrointestinal tract toxicity of aspirin. The similar toxicity of non-steroidal anti-inflammatory drugs (NSAIDs) was not recognized for a few more years. So in the early 1980s, I became comfortable regularly prescribing propoxyphene for RA patients with significant chronic pain, which was almost everyone at the time. Questions about its efficacy existed, but I don’t recall problems with tolerance with increasing doses or abuse.
Chronic Non-Malignant Pain
Over time, more attention was given to patients with chronic non-malignant pain. Shortly after I began practice in Wisconsin in 1986, I vividly recall the cover of Newsweek calling attention to this problem: “Why does someone need to be dying to have pain relief? Why are doctors not doing more to relieve severe chronic pain, which is so detrimental to the quality of life of these people?”
A huge leap was taken in believing that the treatment of chronic non-malignant pain would follow the model of treating cancer pain.
The issue of the suffering of those with chronic pain became prominent not only in the lay press, but in the medical literature as well. In 1986, the National Institute on Drug Abuse published an article advocating chronic opioid therapy for intractable, non-malignant pain.2 During my 10 years at Marshfield Clinic in Wisconsin, I managed many chronic pain patients with guidance from pain specialists in the anesthesia department.
Opioid prescribing began to increase appreciably in the 1990s. From 1990 to 1996, prescriptions for opioids increased from 2 million per year to 8 million per year. Many states passed intractable pain acts intended to protect physicians who, in good faith, prescribed chronic opioids for chronic pain.3
Pain became the fifth vital sign, advocated by the American Pain Society Quality of Care Committee in 1995.4 The Centers for Medicare and Medicaid Services employed a patient satisfaction survey in determining reimbursement for hospital services that asked: “How often did the hospital or provider do everything in their power to control your pain?” Obviously, if you measure a vital sign and it is abnormal, something must be done to address that. The quality of pain relief became a common topic of patient satisfaction surveys overall. Physicians were criticized for poor patient satisfaction because they refused to prescribe medications for pain.
In 1995, sustained-release oxycodone (OxyContin) was approved, and the FDA approved labeling stated that iatrogenic addiction was “very rare.” The OxyContin tablet was purported to be abuse resistant.
In 2001, the Joint Commission on the Accreditation of Healthcare Organizations introduced standards “as part of a national effort to address the widespread problem of underassessment and undertreatment of pain.”5,6
In 2010, the FDA took propoxyphene off the market due to arrhythmia concerns, with the recommendation that it be replaced with codeine, morphine or oxycodone.
Key Hypothesis Never Tested
The first big mistake was to assume that the treatment of chronic non-malignant pain would be like that of treating the chronic pain of cancer patients in the last months of their lives.
This hypothesis was not tested; rather, a huge leap was taken in believing the treatment of chronic non-malignant pain would follow the model of treating cancer pain. This was a logical and intuitive assumption, but with our scientific, evidence-based approach to medical practice, we know that a common-sense assumption often does not pass muster when subjected to a prospective, randomized, double-blind clinical trial.
So why was this hypothesis never tested? I suspect two important reasons. Such trials cost millions of dollars and are financially feasible only for companies seeking FDA approval for an investigational medication. Moreover, such trials are typically done over a relatively short term to gain FDA approval, especially for pain relief, not the many years appropriate for a chronic problem. Who would do such a study over many years, and how would it be funded?
This is not meant to excuse the fact that the key hypothesis was not tested, but rather to offer perspective for how difficult and expensive it would have been to perform a study to adequately test the hypothesis that opioid medications would be safe and effective for the treatment of chronic non-malignant pain. In retrospect, phase 4 post-approval studies should have been done to search for safety signals. Instead, the medical community (and the lay public) relied upon a badly flawed assumption.
The stage was set for the evolution of an enormous problem, which was not anticipated, based on what we now realize were faulty assumptions, especially in a society in which chronic pain was gaining a high profile, and for which a consensus existed that something needed to be done to relieve the suffering of all these people.
A second huge mistake was to recommend administering analgesic medication in the same manner as for cancer pain.
“Stay ahead of your pain,” we advised patients. “Don’t wait until you have pain: Take your pain medicine on a schedule, so you always have a good level of medication on board.” This might make sense for someone with painful bony metastases, but chronically, this is a formula for the development of tolerance to an opioid medication.
Opioid tolerance is “characterized by a reduced responsiveness to an opioid agonist, such as morphine, and is usually manifest by the need to use increasing doses to achieve the desire effect,” and “more than 10-fold escalations of dose in chronic pain management are common,” according to Morgan et al.7 Patients treated this way typically require progressively higher doses of opioid analgesics to achieve pain relief, sometimes reaching doses that suppressed respiration and resulted in fatal overdose.
Standard practice was to advise patients to take their opioid medications in a manner that virtually guaranteed they would develop tolerance, resulting in ever-increasing doses (and toxicity) and physical dependence that produced withdrawal if the dose were decreased or discontinued.
Between 1997 and 2002, morphine, fentanyl and oxycodone prescriptions increased by 73%, 226% and 40%, respectively.8 Around 1999 it was recognized that overdose deaths from prescription opioids were increasing, and of course, this trend has continued for years.
The opioid crisis is a very complicated phenomenon. The Centers for Disease Control and Prevention (CDC) divides it into three general phases: the first from 1990–99, dominated by prescription opioids; the second from 2000–13, dominated by heroin; and the third from 2013 on, dominated by fentanyl.9 From the standpoint of physician prescribing, the first phase dominated by prescription opioids is of the most interest to me.
A third major problem has been the lack of understanding of addiction and which persons have the greatest risk of developing addiction.
Dependence vs. Addiction
Opioid dependence develops when patients have clinical withdrawal as the dose is reduced or withheld. This is distinct from addiction, which is the compulsive, harmful, sometimes illegal use of a chemical substance.
A common misconception exists that anyone can become addicted by taking opioids. It is true that patients taking opioids for over three months do have an increased risk of developing addiction.10 Observational studies show that “of patients who receive a single opioid prescription in the emergency department, after surgery or at the dentist’s office, 1% to 6% end up using opioids for at least 12 months or being diagnosed with opioid use disorder.”11
The potential to develop addiction is neurochemical. Addiction develops when the limbic system (or the lizard brain) hijacks the brain’s frontal lobe. In the case of opioids (or alcohol, or other substances), the lizard brain derives such intense pleasure from the opioid stimulation that it demands more input, overriding the frontal lobe’s restraint. This is not a conscious act: the primitive brain takes over, and opioid use becomes compulsive and unrestrained.12
A prevailing assumption is that taking prescription opioids leads to heroin addiction. An analysis by the National Institute on Drug Abuse, National Institutes of Health and the FDA notes, “Available data indicate that the nonmedical [emphasis mine] use of prescription opioids is a strong risk factor for heroin use”; however, “heroin use among people who use prescription opioids for medical reasons is rare, and the transition to heroin use appears to occur at a low rate.”13
I do not make this point to excuse those with chemical dependency. We are all ultimately responsible for our actions, and if an addicted individual develops insight into their problem, it requires hard work and making good decisions to recover. The point is that only a minority of people are preprogrammed to develop addiction.20,21 It is often stated that many individuals who become addicted to illicit drugs start with prescription drugs, but it does not logically follow that all persons who take prescription opioids are at risk of becoming addicted.
So how does one know in advance that a patient is prone to develop addiction, that their brains are primed for addiction? Some situations are obvious, but in many cases it is not possible to know in advance, and therein lies the dilemma for the prescribing physician. One solution is to have, at the outset, a pain contract, which stipulates the rules for receiving opioids (or other controlled medications), and the understanding that violating the contract means the end of more prescribing. (In my experience, second chances have a very predictable negative outcome.)
Although many patients have abused opioids, it is wrong to categorically label all patients taking chronic opioids as addicted.
Finally, I think it is a mistake that patients doing well on “acceptable,” stable doses of opioids have, in many cases, had these medications withdrawn entirely.13
The 2016 CDC Guideline for Prescribing Opioids for Chronic Pain was misconstrued by insurance companies and physicians (myself included) as a mandate to aggressively reduce opioid doses.10 The U.S. Department of Health and Human Services issued subsequent clarification: “More judicious opioid analgesic prescribing can benefit individual patients as well as public health when opioid analgesic use is limited to situations where benefits of opioids are likely to outweigh risks.”15
Nonetheless, it has become common practice for doctors assuming care of these patients to withdraw them completely from opioids. “The fact that someone has been taking opioids for years does not mean the person has opioid use disorder, but many people make that stigma-driven assumption.”16
The CDC and the FDA have described legacy patients who were started on chronic opioids years ago when the treatment was considered medically appropriate. How should these patients be treated when pain clinics close, or their physicians relocate or retire?17 Finding fault with patients for taking medications prescribed by their physicians is not appropriate; nor is regarding every patient taking opioids an addict.
Thoughtful opinion pieces have appeared in major journals taking issue with “an all-or-nothing approach to pain management.” A perspective piece in the New England Journal of Medicine opined that “as the pendulum swings from liberal opioid prescribing to a more rational, measured, and safer approach, we can strive to ensure that it doesn’t swing too far, leaving patients suffering as the result of injudicious policies.”18
Many negative outcomes have occurred as a result of tapering and discontinuing opioid therapy for chronic pain.
We have seen such drastic change before, when post-menopausal estrogen therapy was dramatically curtailed after data showed that continuous Prempro therapy increased the risk for breast cancer. Almost overnight, estrogen replacement therapy was virtually abandoned, even regimens for which an increase in breast cancer had not been demonstrated.
Forty-two years after receiving my medical degree, it seems to me that the medical community finds it difficult to recognize Aristotle’s maxim that “the virtue in all things lies in a mean between two extremes.” Very few therapeutic decisions are fundamentally binary.
I do not claim to have the answer to this conundrum, but all physicians can learn very important lessons from this experience. We need to be cautious about adopting therapies that make sense because, in the fullness of time, we may recognize that such a decision was misguided. And we need to be humble in judging previous practices made in good faith by physicians who thought they were doing the right thing for their patients.
I have been in medicine long enough to see the error of such ideas as “don’t start treatment of rheumatoid arthritis until it has been present for at least a year, because it might go away, and the treatments are very toxic.” In many cases, what we consider standard of care in 2022 may be regarded with derision decades later.
Richard Brasington Jr., MD, FACP, MACR, is emeritus professor of medicine at Washington University School of Medicine, St. Louis, where he was director of the Rheumatology Fellowship Program for 23 years. He is now a part-time community rheumatologist in Creve Coeur, Mo., and Chester, Ill. He is a Master of the ACR, and was elected to the Alpha Omega Alpha Honor Medical Society and the Gold Humanism Honor Society. He served as an associate editor of The Rheumatologist from 2012–17.
- Porter J, Jick H. Addiction rare in patients treated with narcotics. N Engl J Med. 1980 Jan 10;302(2):123.
- Tennant F Jr, Robinson D, Sagherian A, Seecof R. Chronic opioid treatment of intractable, non-malignant pain. NIDA Res Monogr. 1988;81:174–180.
- Rummans TA, Burton MC, Dawson NL. How good intentions contributed to the opioid crisis: The Opioid Crisis. Mayo Clin Proc. 2018 Mar;93(3):344–350.
- Quality improvement guidelines for the treatment of acute pain and cancer pain. American Pain Society Quality of Care Committee. JAMA. 1995 Dec 20;274(23):1874–1880.
- Baker DW. The Joint Commission’s pain standards: Origins and evolution. Oakbrook Terrace, Ill. The Joint Commission. 2017.
- Baker DW. History of the Joint Commission’s pain standards: Lessons for today’s prescription opioid epidemic. JAMA. 2017 Mar 21;317(11):1117–1118.
- Morgan MM, MacDonald JC. Analysis of opioid efficacy, tolerance, addiction and dependence from cell culture to human. Br J Pharmacol. 2011 Oct;164(4):1322–1334.
- Bay Area Addiction Research and Treatment. When did the opioid crisis start? 2 Nov 2020. https://baartprograms.com/when-did-the-opioid-crisis-start.
- Centers for Disease Control and Prevention. Understanding the opioid overdose epidemic. https://www.cdc.gov/opioids/basics/epidemic.html.
- Dowell D. Haegerich TM, Chou R. CDC guideline for prescribing opioids for chronic pain—United States, 2016. MMWR Recomm Rep. 2016: March 18;65(1):1–49.
- Barnett ML. Opioid prescribing in the midst of crisis—Myths and realities. N Engl J Med. 2020 Mar 19;382(12):1086–1088.
- Teresi L. Hijacking the brain: How drug and alcohol addiction hijacks our brains—the science behind twelve-step recovery. AuthorHouse. Bloomington, Ind. 2011.
- Compton WM, Jones CM, Baldwin GT. Relationship between non-medical–prescription opioid use and heroin use. New Engl J Med. 2016 Jan 14;374(2):154–163.
- Mills, L. “Not allowed to be compassionate”: Chronic pain, the overdose crisis, and unintended harms in the U.S. Human Rights Watch. 18 Dec 2018. https://tinyurl.com/24p8d6mp.
- Dowell D, Jones C, Compton WC. The U.S. Department of Health and Human Services Working Group on Patient-Centered Reduction or Discontinuation of Long-term Opioid Analgesics. October 2019. https://www.hhs.gov/opioids/sites/default/files/2019-10/Dosage_Reduction_Discontinuation.pdf.
- Barnett, ML. Ibid.
- Coffin PO, Barreveld AM. Inherited patients taking opioids for chronic pain—Considerations for primary care. N Engl J Med. 2022 Feb 17;386(7):611–613.
- Comerci Jr. G, Katzman J, Duhigg D. Controlling the swing of the opioid pendulum. N Engl J Med. 2018 Feb 22;378(8):691–693.
- Hemel DJ, Ouellette LL. Innovation institutions and the opioid crisis. J Law Biosci. 2020 Jan–Dec;7(1):lsaa001.
- Volkow ND, Koob GF, McClellan AT. Neurobiologic advances from the brain disease model of addiction. N Engl J Med. 2016 Jan;374:363–371.
- Camí J, Farré M. Drug addition. N Engl J Med. 2003;349:975–986.