A mathematical modeling study by Stanford researchers, published yesterday, concludes that while reducing access to prescription opioids may initially spur some users to migrate to heroin use, it will save thousands of lives over a 10-year time horizon.
It also finds that certain other approaches to fighting the epidemic—like enhancing access to naloxone, needle exchanges, medication-assisted treatment, and psychosocial treatment—”have immediate benefits and no down-side costs,” according to study co-author Keith Humphreys, Ph.D. “Expanding them will save lives.”
The lead author of the study, published in the American Journal of Public Health, is Alison Pitt, a graduate student at Stanford University’s School of Management and Engineering. She was working under the supervision of the study’s senior author, Margaret L. Brandeau, Ph.D., the Coleman F. Fung professor at Stanford’s engineering school, and a professor of medicine as well. The third author, Humphreys, is a professor of psychiatry and behavioral sciences at the university. He assisted with literature review and parameter valuation.
Brandeau’s research group focuses on mathematical modeling to inform public health decisions.
“The opioid crisis is the biggest public health crisis the US faces right now,” she writes in an email, “so it was natural that we would want to turn our attention to this critical problem.”
The study’s most important finding, Brandeau says, is this: “In order to curb the opioid epidemic, we’re going to need to reduce the number of opioid prescriptions.
“However,” she continues, “in the short term this has the unintended consequence of inducing some people to shift to heroin, and thus will likely increase the total number of opioid deaths in the short term. This means that we not only need to reduce the number of opioid prescriptions; we also at the same time need to scale up drug treatment.”
“It is going to take a long time before the number of deaths from opioids in the U.S. decreases appreciably.”
The study, mostly performed in 2015, looks at 11 policy interventions to combat the opioid crisis, and projects their impact over time on total prescription opioid-related fatalities. The study performed projections over both a five-year span (2016-2010) and a ten-year span (2016-2025).
In an attempt to take into account the effect of these interventions on pain patients with a legitimate need for prescription opioids, the study also computes impacts on the “quality-adjusted life years” of all affected. To this end, researchers assumed that opioids reduce acute pain, for instance. They made no such assumption for chronic pain patients, however, writing: “no clear evidence supports an average utility benefit for the treatment of chronic pain.”
The study focuses on opioid-addiction related fatalities of three types: deaths caused by (a) prescription opioids prescribed by a doctor; (b) prescription opioids obtained through diversion (e.g., filched from a parent or illicitly purchased on the street); and (c) illicit opioids, like heroin, to which people escalate after having started on prescription opioids, obtained from either source (a) or (b). The study does not examine deaths among users who start on illicit opioids, like heroin.
The researchers anticipated that some interventions would have both positive and negative impacts. For instance, restricting prescription opioid availability in various ways was expected to avert some new cases of opioid addiction, while also spurring some existing users to switch to heroin, resulting in greater risk of death for them.
The basis for the researchers’ mathematical assumptions, the authors write, were “published literature, expert opinion and model calibration.” Model calibration means “adjusting values for model parameters (within ranges that are plausible) so that the model projections match reality,” Brandeau explains in an email.
“Over time the benefit of not addicting people by overprescribing mount up to the point that such policies are extremely beneficial to population health.”
One surprising outcome, Brandeau tells Opioid Watch, is that even with all the interventions, “it is going to take a long time before the number of deaths from opioids in the U.S. decreases appreciably.”
As a baseline, the study estimated that, without interventions, we could expect another 510,000 opioid-addiction related deaths (as defined above) from 2016 to 2025, including 170,000 deaths from prescription opioids and 340,000 after users escalate to heroin.
Under the five-year analysis, the researchers found that none of the 11 interventions could realistically achieve even a five percent reduction in opioid-related deaths. Even a combination of techniques, at the time horizon, was unlikely to do much better, since several policies, in the short-term, would spur migration to heroin.
Under the ten-year analysis, however, the prospects were better. “Some policies,” the researchers write, “avert significantly more deaths over 10 years than be proportionally expected over five years.”
“Over time,” Humphreys explains to Opioid Watch, “the benefit of not addicting people by overprescribing mount up to the point that such policies are extremely beneficial to population health.”
By employing a “portfolio” of eight of the 11 interventions over a 10-year time span—including reducing prescription opioid use in three ways while increasing disposal of excess prescriptions, and enhancing access to naloxone, needle exchange, MAT, and psychosocial treatments—about 59,000 addiction deaths could be averted—an 11 percent improvement over the baseline.
Stefan Kertesz, MD, a primary care doctor and addiction specialist at the University of Alabama School of Medicine and an advocate for the rights of chronic pain patients, takes exception to some of the study’s assumptions.
“I think the models assume a greater benefit from prescription opioid control than is supported by the literature,” he writes in an email, “and they impose an assertion of no benefit from prescribed opioids for any person with chronic pain. I don’t think that is supported by the literature.”
Nevertheless, he writes, “I wound up agreeing with the very sobering conclusions offered by this paper,” referring to its prediction that progress would be slow. “Papers like this remind us that this is a marathon, not a sprint—one that demands the best of us, working together with respect and care.”
“In the history of the world, no epidemic has been reversed by just treating people who already have the disease.”
Some might ask why, if some interventions save lives with no negative effects (like improving access to naloxone, needle exchanges, and MAT), while others (like reducing prescription opioid access) foreseeably drive some users into heroin use, why not just focus on the former policies?
“The answer,” co-author Humphreys tells Opioid Watch, “is that in the history of the world, no epidemic has been reversed by just treating people who already have the disease. Without preventing new people from getting it, you can’t stop an epidemic from getting worse.”