Post-op Opioid Use Affected by Age, Anxiety and Education

Author: Michael Vlessides

Anesthesiology News

A study of factors that influence the amount of postoperative opioid consumption concluded that significantly fewer opioids were used in the first month after surgery by patients who were older, college graduates, more active and less anxious compared with their counterparts. The study also confirmed what many anesthesiologists already suspect: Surgical patients are being discharged with far too many opioids.

“There’s a significant iatrogenic component to the opioid epidemic,” said Daniel B. Larach, MD, MTR, MA, who was a resident at the University of Michigan Medical Center, in Ann Arbor, when the research was conducted.

“Most current heroin users started their opioid use with prescription drugs. This is not a problem confined to the pain clinic or primary care office,” he said. “Nearly one-quarter of initial prescriptions to opioid-naive patients are postsurgical. In fact, the most common complication can be considered new persistent opioid use. So clearly, there is a pressing need for individualized postoperative prescribing. But this is difficult for us to do, because we don’t know what patient factors are associated with postoperative opioid use.”

The researchers included opioid-naive patients undergoing hysterectomy, thoracic surgery, total knee arthroplasty or total hip arthroplasty. Preoperative phenotyping was done using validated questionnaires to assess pain, centralized pain, catastrophizing, depression, anxiety, functional status, fatigue and sleep disturbance.

The study’s primary outcome was total opioid use in oral morphine equivalents (OMEs), which was assessed by patient self-reports at approximately two weeks and one month postoperatively. Values for OMEs were standardized across all surgery types; multivariate linear regression models were constructed to predict total standardized OMEs consumed in the first 30 days after surgery.

As Dr. Larach reported at the 2018 annual meeting of the American Society of Anesthesiologists (abstract BOC04), complete outcome data were available from 688 patients (66.9% women; mean age, 59.2 years; 91.8% white) of the initial cohort. “We had complete phenotype data available for 628 of these subjects,” Dr. Larach said. “From these patients, we constructed multivariable linear regression models and identified four different factors associated with increased opioid consumption.” The factors (Table) were :

• younger age (beta coefficient for age, –0.02 standardized OME units; P<0.001);
• not having graduated college (beta coefficient for college graduate status, –0.16 standardized OME units; P=0.044);
• lower functional status (beta coefficient for functional status, –0.03 standardized OME units; P=0.008); and
• higher anxiety scores (beta coefficient for anxiety, 0.05 standardized OME units; P=0.002).

“When we broke the data down by surgical group, we found that, as with the overall cohort, hysterectomy patients who were younger, more anxious and lacked college degrees all consumed more opioids,” Dr. Larach said. “Similarly, [total hip arthroplasty] patients who were younger and more anxious consumed more opioids, as well as those who were not [white] and who were male.”

Thoracic surgical patients who were current or former alcohol abusers or heavy users were also found to consume more opioids, but no significant associations were found within the group of total knee arthroplasty patients. “Nor did we see any significant differences in opioid consumption by surgical approach within the hysterectomy or thoracic surgical groups, although the number of patients having open surgeries was relatively low in those groups,” he added.

Figure 1. Median opioids prescribed and consumed, in OMEs.

IQR, interquartile range; OMEs, oral morphine equivalents

“As other investigators have shown, opioid overprescribing was common in our surgical cohort,” Dr. Larach commented. “We had an overall excess of more than 178,000 oral morphine equivalents; this is equivalent to 35,768 individual 5-mg hydrocodone pills, or 52 pills per patient” (Figure 2). To alleviate such excess, the researchers recommended the use of evidence-based postoperative opioid prescribing recommendations, such as those found at opioidprescribing.info.

Figure 2. Total opioids prescribed and consumed, in OMEs.

OMEs, oral morphine equivalents

“There are a few ways in which these data might be able to inform clinical practice,” Dr. Larach concluded. “One of them is leading us toward more individualized postoperative prescribing, such as giving fewer opioids to older patients.

“Perhaps more importantly, we may be able to use these data to identify which patients would benefit from more careful postoperative monitoring for adverse events, such as opioid misuse or new persistent use,” he said. “We may also be able to implement perioperative interventions focusing on patients who have anxiety disorders.”

Such initiatives, he added, might take the form of educational interventions, such as setting realistic expectations for normal postoperative pain, or establishing the acceptability of alternatives to opioids for postoperative pain. “There may also be a role for behavioral techniques like breathing exercises or guided imagery,” Dr. Larach said. “There are some data that these can reduce perioperative anxiety and pain, as well as length of stay and cost.”

As Christopher L. Wu, MD, discussed, this investigation offers many intriguing insights, particularly the extent of opioid overprescribing. “I think this is very important work,” he said, an assistant attending anesthesiologist at the Hospital for Special Surgery, in New York City.

“The thing I find most astounding is the sheer number of opioids that are overprescribed.

“But the researchers don’t stop with simply saying that opioids are being overprescribed,” Dr. Wu said in an interview with Anesthesiology News. “They put it in a more relatable context by breaking it down into a per-patient level.”

Dr. Wu also appreciated the researchers’ efforts to pinpoint the factors associated with opioid consumption. “Phenotyping in this context is not well studied, so I applaud them for doing that as well. It’s terrific that they’re looking at the opioid epidemic in a different way. Now that we know we overprescribe, the next question is who gets addicted. And those are very important questions that this group is trying to answer.”