Drug Overdose Prevention: An Exercise in Optimism

The ability to “stage” severity of illness and to accurately predict the impact of untreated illness on morbidity and life expectancy is key in the selection of available evidence-based treatment interventions likely to sustain health while minimizing adverse effects. Benefit-risk prediction tools support clinicians’ selection of optimal medication(s) and other clinical interventions in a manner that can be clearly articulated to patients and family, facilitating an educated, collaborative decision-making process about treatment. While we continue to face an escalating opioid crisis, with 106,699 U.S. drug overdose deaths in 2021 (the majority attributable to opioids, especially illicitly manufactured fentanyl analogues in drug-trafficking supplies) (1), predictive tools might improve patient engagement and retention in life-saving treatments for opioid use disorder (OUD).

In this issue, Brandt et al. (2) contribute meaningfully to the literature defining predictive risk of substance overdose in OUD populations who have used illicit prescription opioids or illicit street opioids such as heroin and are currently receiving medication for opioid use disorder (MOUD). Predictive risk research with these OUD populations has lagged compared to overdose risk research for those maintained on opioid analgesics (3).

Brandt et al. pooled data from three national multisite randomized controlled trials within the National Drug Abuse Treatment Clinical Trials Network (CTN) that were each designed as open-label studies to examine or compare OUD outcomes following initiation of MOUD. MOUD includes three medications approved by the U.S. Food and Drug Administration: the mu-opioid agonist methadone (liquid formulation), the mu-opioid partial agonist buprenorphine (in this study, the sublingual formulation of 4:1 buprenorphine:naloxone), and the mu-opioid antagonist naltrexone (in this study, extended-release injection naltrexone). An advantage to pooling these CTN studies lies in their prospective design of OUD safety monitoring for adverse and serious adverse events, which facilitated data harmonization to detect fatal and nonfatal overdose events during a 24-week treatment episode. The authors further harmonized baseline participant variables across the studies that would be predicted to be important in contributing risk or protection for a substance overdose event, based on published studies.

A total of 39 of 2,199 participants (1.78% of the pooled sample) had at least one overdose event within 24 weeks of MOUD randomization. Medication assignment was related to the likelihood of an overdose event, with higher probability for antagonist treatment (5.30% of naltrexone participants) compared with agonist treatments (1.51% of methadone and 1.15% of buprenorphine participants), replicating results from previous retrospective cohort studies. Failure to initiate naltrexone accounted for a substantial portion of the overdose risk in that group—double the risk of those who received at least one injection of extended-release naltrexone (8.9% and 3.9%, respectively). For all three MOUD groups, MOUD nonadherence often preceded an overdose event, but this was significantly greater for buprenorphine and naltrexone than for methadone. This may reflect two protective factors of methadone treatment: 1) the daily dosing structure of methadone dispensation during initiation, in contrast to weekly to monthly prescriptions or monthly administration of buprenorphine and naltrexone, respectively, and 2) severity of opioid withdrawal onset for missed methadone dosing (a negative reinforcer of medication adherence) compared with missed buprenorphine dosing (often delayed onset of opioid withdrawal) and the absence of opioid withdrawal with missed extended-release naltrexone. Consistent with previous reports, baseline benzodiazepine use positively predicted greater risk for an overdose event for all three MOUD groups. In summary, failure to successfully initiate MOUD and early discontinuation of MOUD are significant predictors of overdose risk in the first 6 months of treatment, and concurrent use of benzodiazepines (prescribed or illicit) adds to substance overdose risk in OUD populations. These findings underscore the national imperative to appropriately engage treatment-seeking OUD patients in an enduring MOUD treatment, as national retrospective studies provide convergent evidence that longer MOUD maintenance is increasingly associated with reduced risk of overdose during the first year of treatment (4).

Patient-centered risk-prediction approaches to care generally succeed in reducing population morbidity and mortality if patients have equal access to evidence-based, compassionate care. However, patients sometimes reject life-saving treatment options because they think that the treatments will degrade their quality of life. This may be most prominent when patients know that mortality risk is high, and they don’t have confidence that treatment will improve their fate. It also commonly occurs for any illness in which adequate treatment of the illness involves loss of function or identity, enduring pain or disability, or loss of a previously central source of pleasure (5). Effective treatment options only work when they are adopted by patients.

Many evidence-based treatment strategies for OUD are available to improve outcomes and thus reduce overdose deaths, including MOUD; promoting peer support engagement; naloxone education and distribution (to pharmacologically stop an opioid overdose from progressing to death); and harm reduction approaches to engage drug users in lower-risk practices, such as syringe exchange programs, use of fentanyl test strips to check drug supply before using, and self-injecting at supervised injection sites (6, 7). On December 29, 2022, the Consolidated Appropriations Act of 2023 was signed into law and amended the Controlled Substances Act to eliminate the “DATA-Waiver requirement,” previously codified in 21 U.S.C. 823, thereby allowing all Drug Enforcement Agency–licensed prescribers to use buprenorphine in office-based practice to treat unlimited numbers of OUD patients (8).

While access to care is widening, many individuals with OUD do not actively seek care, or engage in care briefly and then drop out. Treatment avoidance is not surprising, given criminalization of drug use, external and internalized stigma about drug use, pessimism about recovery, and fear that sacrificing being high will leave only a yearning to be high again. When the path to treatment initiation is more difficult, as exemplified by the “induction hurdle” associated with beginning naltrexone, there is a greater likelihood that patients will not access the medication at all. Indeed, in the multisite X-BOT trial comparing extended-release naltrexone with sublingual buprenorphine-naloxone, nearly a third of the patients who were randomized to naltrexone never received a single dose of the medication (9). New methods to rapidly induct patients onto naltrexone are currently being tested in the multisite SWIFT trial, conducted as part of the CTN (10), and may help to improve this situation.

Reasons for discontinuing MOUD treatment are often person specific, with multiple contributing factors. For example, a qualitative study of buprenorphine discontinuation showed that the two most commonly reported reasons for discontinuation were that it did not work to stop opioid use or that it worked so well that a person could no longer get high (11). Here, dosing practices may affect treatment retention. Structural factors increasing the probability of MOUD discontinuation include prescriber access, insurance coverage, and MOUD stigma. A recent study of patients on our inpatient withdrawal management unit demonstrated very low optimism scores compared with similar studies in a variety of populations (12). Addressing pessimism about recovery may be an important target for patient engagement and retention. Sometimes increasing optimism about treatment might require flexible decision making, such as the continuation of prescribed benzodiazepines and/or gabapentin for patients with significant co-occurring anxiety and pain conditions, despite knowledge that these agents elevate opioid overdose risk (13, 14).

Finally, specific pharmacological and behavioral treatment approaches may help increase retention in MOUD. The CTN is currently conducting a large multisite trial entitled “Optimizing Retention, Duration, and Discontinuation Strategies for Opioid Use Disorder Pharmacotherapy” (RDD) (15). One of the goals of the RDD trial is to test strategies to increase MOUD retention by testing different dosages (standard vs. high-dose) and formulations of buprenorphine (sublingual vs. extended-release injectable). The RDD trial is also testing a digital app intervention to help improve retention in both buprenorphine and extended-release naltrexone treatment.

Reducing the unacceptably high rate of opioid overdoses will require interventions on multiple fronts that make widened access to OUD care “real” in terms of patients’ willingness to receive the available life-saving treatments. The NIH-funded Healing Communities Study (16) is a multistate project that is doing exactly that, not only by targeting the need to offer evidence-based treatment practices but also by engaging community members to focus on multiple issues, such as stigma reduction, treatment access, and mental health. Like so many difficult and important tasks, this too takes a village.