Review compares medicines for opioid use disorder across 103 trials
The analysis found differences in treatment retention and opioid abstinence, but limited evidence on deaths and other important outcomes.
High evidenceReviewSome caution advised
Medical disclaimer: This article summarizes research findings and is for informational purposes only. It is not medical advice.
Editorial illustration — not from the study.
This systematic review and network meta-analysis compared 13 interventions for opioid use disorder. Treatment retention meant remaining in follow-up until the end of a study, while abstinence was assessed during follow-up, including through urine drug tests.
Compared with placebo, heroin had the strongest result for treatment retention, followed by heroin combined with methadone and higher-dose methadone. Buprenorphine and higher-dose methadone were associated with improved abstinence measures, but the certainty of this evidence varied. The researchers said evidence for mortality, overdose-related death, and other important outcomes remains limited.
The question reviewed
The researchers searched eight medical and clinical-trial databases through August 10, 2025, for randomized controlled trials involving people with opioid use disorder. They found 103 trials covering 26,624 participants and 13 interventions. Reviewers independently selected studies, extracted information, assessed risk of bias, and used a random-effects network meta-analysis to compare treatments, including comparisons that may have been indirect. The main outcomes were treatment retention, opioid abstinence, criminal activity, all-cause mortality, and overdose-related death.
What the review concluded
Compared with placebo, moderate-certainty evidence indicated that heroin ranked highest for treatment retention, with an odds ratio of 31.47 and a 95% confidence interval of 12.29 to 80.58. Heroin combined with methadone and higher-dose methadone also ranked highly, with odds ratios of 8.00 and 6.94, respectively; both findings had moderate-certainty evidence. High- and low-dose buprenorphine and higher-dose methadone were associated with improved abstinence measured by urine drug tests, although certainty was moderate for buprenorphine and low for higher-dose methadone. The analysis did not find evidence that any intervention reduced criminal activity, all-cause mortality, or overdose-related death; certainty for these outcomes ranged from moderate to low. These results describe findings across the included trials and do not establish that one treatment is appropriate for every person.
Who this is relevant to
These findings may be most relevant to people with opioid use disorder who resemble participants in the included randomized trials and to settings where the studied therapies are available. The heroin-assisted therapy evidence specifically came from people with prior heroin dependence and should not be assumed to apply to people without such prior use. The review also noted that more evidence is needed for people using fentanyl, so applicability to current populations with different patterns of opioid exposure is uncertain. The results are population-level research findings, not individualized medical advice.
Why it matters
The review brings together a large body of randomized-trial evidence and compares several medication-based approaches using common outcomes. It suggests that treatment retention and abstinence may differ among therapies, while also highlighting that evidence about deaths, overdose-related deaths, criminal activity, and other outcomes is incomplete. The findings are especially relevant to research and health-system planning, but they do not by themselves determine individual treatment choices.
Limitations & evidence assessment
The certainty of evidence was not the same for all outcomes: it was moderate for some retention and abstinence findings and low for several other results. Network meta-analysis can rely partly on indirect comparisons, so comparisons between treatments may be less certain than direct head-to-head trials. The abstract does not provide enough detail about follow-up duration, participant characteristics across all trials, or the risk of bias for each outcome. Heroin-assisted therapy trials exclusively enrolled people with heroin dependence, and the researchers noted that larger direct comparisons are needed, particularly among people using fentanyl in the changing unregulated drug supply.
Why this evidence level: Meta-analysis pooling multiple studies sits at the top of common evidence hierarchies.
Evidence levels are editorial estimates derived from study metadata — they are not clinical appraisals.
// Source
The Canadian Journal of Psychiatry · 2026 · DOI: 10.1177/07067437261479071
A prospective, multicenter phase II trial examined whether removing only the treated tumor area through the anus could preserve the rectum in people with locally advanced rectal cancer who responded well to chemoradiotherapy. Among 80 participants, three-year disease-free survival was 84.8%, and local recurrence occurred in 5.0% during the reported follow-up. Because everyone received local excision and there was no comparison group, the findings do not establish that this approach is as effective as more extensive surgery.
This retrospective human study compared a visual CT scoring method with an automated computer analysis for tracking fibrosing interstitial lung disease. Among 92 people followed with CT scans and lung-function tests, both approaches were associated with changes in lung function and signs of disease progression. The study does not show that either method improves treatment or patient outcomes.
A systematic review and meta-analysis examined how silver diamine fluoride affects human dental pulp cells in laboratory studies. Across seven in vitro studies, higher concentrations were associated with lower cell viability, but these findings do not show how the material affects dental pulp in people.
