Clinical Context

The United States has confronted an opioid epidemic for over a decade, and the Centers for Disease Control and Prevention (CDC) provided guidance in an attempt to limit opioid prescribing on March 15, 2016.^[1] There were 12 main recommendations in their document, which was published April 19, 2016, in JAMA.^[2] Some of these recommendations included:

• Avoiding opioids unless a clear benefit:risk ratio favors opioids for pain control and function
• New prescriptions should be for immediate-release, not extended-release, opioids
• Opioid prescriptions ≥ 50 MME/d should be used only when absolutely necessary
• The management of acute pain should usually require ≤ 3 days of treatment with opioids
• Clinicians should assess the benefits and harms of opioid therapy within 4 weeks after the first prescription and ≤ 3 months thereafter
• Urine drug testing should be performed at least annually to detect drug misuse
• Clinicians should avoid prescribing opioids and benzodiazepines together

These recommendations received a lot of attention, but did they make a difference? The current study by Scherrer and colleagues evaluates prescription patterns for opioids in the US prior to and after the presentation of the CDC guidelines.

Study Synopsis and Perspective

Since the publication of the CDC's opioid guideline in March 2016,^[1] there has been a substantial drop in the prescription rate for these medications -- except for patients at highest risk for opioid abuse and overdose, new research shows.

Results of a large cross-sectional study show that opioid prescriptions for patients with depression, anxiety, or a substance abuse disorder have increased since the guideline for noncancer pain was published.

"The opioid prescription rate is going down, but not in the right populations," study investigator Jeffrey F. Scherrer, PhD, professor, department of family and community medicine, St Louis University School of Medicine, St Louis, Missouri, told Medscape Medical News.

"Prescriptions for high-risk opioids are not going down in the populations we're most concerned about in terms of becoming chronic opioid users and developing misuse, and overdoses."

The study was published December 1 in JAMA Network Open.^[3]

Unanswered Questions

As reported by Medscape Medical News, the CDC guideline provided recommendations for opioid treatment of primary care patients with noncancer pain. Its publication was followed by a drop in opioid prescribing rates, high-dose prescribing, and benzodiazepine coprescribing.

It is unclear, however, whether the types of opioids prescribed have changed since the guideline's publication or whether patients with psychiatric disorders who are at risk for opioid misuse are less likely to receive opioids with high potential for abuse.

The investigators used Optum's deidentified integrated claims-clinical dataset. The analysis included 279,435 patients (141,219 before and 138,216 after guideline publication). The mean age of the patients was 53 ± 17 years, 61% were women, and 80% were White.

The researchers compared the odds of prescribing new Schedule II opioids, which include codeine, hydrocodone, and oxycodone, with the odds of prescribing Schedule IV opioids (tramadol) for the 18 months before and after the guideline's release.

Scherrer noted that an important element of the study is that it measured new opioid prescriptions, not the continuation of existing ones.

The researchers limited the analysis to the opioids that were most frequently dispensed in the United States. Tramadol, the comparator drug, has been associated with fewer adverse events and lower abuse potential compared with Schedule II drugs.

Covariates included age, sex, race/ethnicity, clinician type, geographic region, obesity, and pain condition. Researchers also adjusted for time in months from the first possible new opioid prescription.

Good News, Bad News

The investigators categorized diagnostic codes for more than 900 pain conditions for which opioids are indicated. These included arthritis, musculoskeletal pain, back pain, neuropathic pain, and headache pain.

After adjusting for covariates, overall prescriptions for oxycodone decreased 14% vs prescriptions for tramadol (adjusted odds ratio [aOR] = 0.86 [95% CI: 0.82, 0.9]) and a 5% reduction in hydrocodone prescriptions (aOR = 0.95 [95% CI: 0.91, 0.98]).

"That's good news," said Scherrer.

Hydrocodone is by far the most commonly prescribed opioid in the United States. The small (5%) decrease in hydrocodone prescriptions "may partly explain the previously reported overall decline in opioid prescribing between 2015 and 2017," the authors noted.

Still, what's "concerning" is that the ORs for people at risk "are actually going in the opposite direction," said Scherrer.

For example, depression and anxiety disorders were significantly associated with greater odds of receiving a hydrocodone prescription (aOR = 1.09 [95% CI: 1.05, 1.12] for depression and aOR = 1.07 [95% CI: 1.04, 1.11] for anxiety disorder) and an oxycodone prescription (aOR = 1.23 [95% CI: 1.18, 1.27] for depression and aOR = 1.16 [95% CI: 1.11, 1.2] for anxiety disorder) compared with tramadol.

In addition, benzodiazepine coprescriptions were associated with greater odds of receiving a prescription for hydrocodone or oxycodone than for tramadol (aOR = 1.07 [95% CI: 1.04, 1.1] for hydrocodone and aOR = 1.19 [95% CI: 1.15, 1.23] for oxycodone).

"We expected the opposite to what we found here," said Scherrer. "This suggests there's still some education to do regarding safer opioid prescribing for patients at risk."

A Fine Balance

Scherrer added that the study included prescribers from all disciplines, from surgery and pain management to primary care, and that their practices may vary in terms of pain prescriptions.

For example, he said, it is possible that surgeons don't use tramadol as often as primary care practitioners because it may be less effective for surgery patients.

High-risk patients may "doctor shop" until they find a physician who will give them an opioid stronger than tramadol, said Scherrer.

Several states have mandated limits on new opioid prescriptions. In some places, for example, a patient can only get a 3-day supply, said Scherrer.

He noted, however, that he is concerned about "pushing the pendulum too far the other way," resulting in inadequate pain control.

"There's a place for potent opioids; the question becomes, why are people who are at most risk for misuse more likely to get these opioids?" he speculated.

Scherrer suggested a possible "amendment or supplement" to the CDC guideline to address opioid recommendations for high-risk patients.

In addition, there needs to be additional education for prescribers about the importance of screening for risk factors such as depression, substance use disorder, and benzodiazepine comedication, he said.

A Complex Issue

Commenting on the study for Medscape Medical News, Andrew Saxon, MD, professor, department of psychiatry and behavioral science, University of Washington, Seattle, who is a member of the American Psychiatric Association's Council on Addiction Psychiatry, said he took issue with some aspects of it.

He noted that it is unclear whether prescriptions were for acute or chronic pain, which is "a problem." He also disagreed with the assertion that tramadol is a safer option that results in better outcomes.

Although it might be "slightly true" that tramadol is less addictive or has less misuse potential than other opioids, "it still can cause opioid use disorder if people overuse it and certainly can cause physiological dependence," he warned.

Tramadol, he added, "can actually have a really complex withdrawal syndrome" compared with other opioids.

Saxon said he was not surprised that patients with anxiety and depression continue to receive opioid prescriptions. As recently as a decade ago, the message was that "a compassionate, caring physician" would prescribe an opioid for a patient with chronic pain, he said.

"Based on that false belief that we all fell prey to, we have a whole group of people we call legacy patients who were started on an opioid for chronic pain and now it's hard to get them off," he said.

Saxon also pointed out that patients with chronic pain have high levels of anxiety and depression.

"Some people in the field think chronic pain is almost another manifestation of a psychiatric syndrome," he said.

Scherrer has received compensation for services as editor of Family Practice and grants from St Louis University Research Institute and the National Institutes of Health. He has received compensation for lectures at Washington University.

Clinical Implications

• In 2016, the CDC recommended using opioids only when their benefit for pain and function outweighed the known risks for therapy. It also recommended avoiding treatment with opioids for > 3 days in cases of acute pain and dosing < 50 MME/d. The CDC also recommended evaluating the need to continue chronic opioid therapy at least once every 3 months, with annual urine drug testing.
• The current study by Scherrer and colleagues found that the 2016 CDC guidelines on opioid use had little effect on opioid prescribing habits in a large cohort of US adults.
• The results of the current study demonstrated that the healthcare team needs to work harder to provide multimodal pain management programs that minimize the use of opioids.

Earn Credit