Behavior of Opioid Prescribing: Admitting What We Don't Know

Share this content:
Clinicians should adhere to the principles of patient autonomy, beneficence, and non-maleficence when making prescription decisions about opioids.
Clinicians should adhere to the principles of patient autonomy, beneficence, and non-maleficence when making prescription decisions about opioids.

Patient autonomy is one of the bedrock ethical principles of modern medicine, and rightfully so. We want patients to be well informed, and we want them to guide the course of their health decisions intelligently. Therefore, it it's cause for celebration when it appears that patients are acknowledging the dangers associated with prescription drug use and embracing holistic, or at least nonpharmaceutical, methods of pain control vocally. Patients have grasped the gravity of opioid dependence and believe that physicians' prescribing habits should be adjusted accordingly. This is probably a straightforward reaction to what ordinary people see in their own lives and environment — more than 15,000 people in the United States die each year from a prescription drug overdose. Approximately a quarter of Americans have used opioids in past year, and nearly half of Americans describe opioid use as a very serious problem in their community.1 “Epidemic” is not too strong a word to describe opioid use in the United States. Approximately 4 out of 5 patients prefer to start with nonpharmacologic approaches to treating pain.2

These patient preferences seem to highlight the connection between the opioid abuse epidemic and physician prescribing habits. The connection is not tenuous. Recent data suggest that nearly 1 in 50 people prescribed an opioid will become a long-term user, which is a quite staggering figure.3 It suggests that a typical outpatient clinic that performs, perhaps 10 cholecystectomies, appendectomies, and hernia repairs per day is also, on average, minting a new long-term opioid user every week. However, that relationship is far from iron clad. It turns out that patients are 30% more likely to become long-term narcotic users if they are treated by a physician who tends to prescribe opioids more frequently than his or her peers. Meanwhile, there is no indication that those managed by clinicians who prescribe opioids infrequently were undertreated or had significant ill effects. The bottom line is that our behavior matters, and if we are genuinely interested in acting in accordance with beneficence and nonmaleficence — 2 more fundamental principles of medical ethics — toward our patients, we need to alter our opioid prescribing habits.

It is gratifying when the principles of autonomy, beneficence, and nonmaleficence point in the same direction. That agreement gives us the sense that we can exhale and proceed with confidence in the ethical basis of our decisions. We might even allow ourselves to forget that there are other ethical principles that should also influence our behavior.

One of those other principles is veracity. Veracity is our commitment to, as much as we can, tell the truth to patients and the other stakeholders with whom we interact. Critically, this requires giving the patient accurate information about diagnosis, prognosis, and treatment options so that he or she can exercise autonomy and make an intelligent and informed decision. I like to think of veracity as the physician's version of a courtroom oath — we have the obligation to tell the truth, the whole truth, and nothing but the truth.

Unfortunately, sometimes veracity does not mix well with the pursuit of autonomy, beneficence, and nonmaleficence. Like so much else in medicine, the problem stems from the struggle to manage complexity. We understand that a great many factors can contribute to any given health outcome. For instance, there are a litany of environmental, lifestyle, and genetic factors that contribute to the development of heart disease. We could in theory run a series of trials that would determine the particular contribution of each factor, but that would be prohibitively expensive, extraordinarily time consuming, and devilishly difficult from an experimental design standpoint. In addition, to respect patient autonomy, we know that we will eventually need to develop a simple and straightforward statement to present to patients that summarizes their personal risk factors cogently. That statement, by necessity, will not be able to provide detailed information about each potential causative element, much less information about the second- and third-order interactions that further influence outcomes. As a result, we will not, but also cannot, present patients with the complete story. Instead, we discuss the effects of our actions in broad rather than patient-specific terms, rely on population-level trends to reassure ourselves of our beneficence and nonmaleficence, and try not to worry too much about the likely outcomes for any particular patient. We tell the truth, but not the whole truth.

This can be problematic, even if when we are not in the witness box. Surveys of patient preferences and the experimental data gathered thus far suggest that clinicians should be prescribing fewer narcotics. Great — but what does that mean? Surely, mandating that every physician — or even just every physician who falls into the “frequent prescriber” category — cut opioid prescriptions by a certain amount is too blunt an approach. We know that some patients are going to require more aggressive pain control, and we should not flippantly resolve to undertreat them. Instead, our intuition supports tailoring prescriptions based on risk factors that have been shown to correlate with future addiction rates and pertinent outcome markers, which is valid as far as it goes. The trouble is that that information is not available to us.

The American Pain Society commissioned an interdisciplinary review of the best evidence available for constructing postoperative pain management guidelines.4 The following is a partial list of evidence gaps that prevent the responsible design of prescription protocols:

  • Duration of opioid treatment and tapering strategies after inpatient discharge
  • Differential outcome effects of neuraxial analgesia based on surgical site
  • Clinical harm of nonsteroidal anti-inflammatory drugs and local anesthetics on bone and soft tissue healing
  • Evidence concerning treatments according to type of pain (muscle or joint, neuropathic, nocioceptive, etc). 

The last entry on that list is especially striking — there is not even good quality evidence about how prescription habits should be changed based on what kind of pain a patient is experiencing. If we don't know that, how can we can conscientiously give physician advice as vague as, “we need to decrease our opioid prescriptions?” Sometimes, the first step towards true understanding is admitting that we don't know.

For now, we'll no doubt continue to pursue the principles of patient autonomy, beneficence, and nonmaleficence when making prescription decisions, as we should. But when it comes to veracity? We need to stop pleading the Fifth.

References

  1. Sifferlin A. Dying from an opioid overdose is more common than you think. Time. time.com/4890536/opioid-heroin-overdose-deaths/. Published August 7, 2017. Accessed December 6, 2017.
  2. Sifferlin A. Here's why most Americans prefer to treat pain without drugs. Time.   time.com/4936193/pain-relief-opioids/. Published September 12, 2017. Accessed December 6, 2017.
  3. Physicians' opioid prescribing patterns linked to patients' risk for long-term drug use.  Harvard T.H. Chan School of Public Health. https://www.hsph.harvard.edu/news/press-releases/opioids-addiction-physicians/. Published February 15, 2017. Accessed December 6, 2017.
  4. Chou R, Gordon DB, de Leon-Casasola OA, et al. Management of postoperative pain: a clinical practice guideline from the American Pain Society, the American Society of Regional Anesthesia and Pain Medicine, and the American Society of Anesthesiologists' committee on regional anesthesia, executive committee, and administrative council. J Pain. 2016;17(2):131-157.

Free E-Newsletter