Meta-Analyses ThatGuy[A]

Extended-Release vs Oral Naltrexone: What a 2025 Review Found on Staying in Treatment

A 2025 systematic review and meta-analysis compared extended-release naltrexone (XR-NTX) with oral naltrexone (NTX) for substance use disorders. The review included adults with alcohol use disorder (AUD) and opioid use disorder (OUD). It searched multiple databases without date or language limits and included randomized controlled trials (RCTs) and observational studies.

The review screened 1,116 records, selected 27 for full-text review, and included 15 studies: 3 open-label RCTs (n=340) and 12 retrospective studies (n=18,695). Seven entered the quantitative meta-analysis, but only one of those seven was an RCT. The other two RCTs contributed to the broader review, not that pooled estimate.

The main outcome was treatment persistence. Pooled odds ratios favored XR-NTX at 3 months (OR 1.43, 95% CI 1.04–1.96) and 6 months (OR 1.96, 95% CI 1.37–2.81). At six months, heterogeneity was substantial (I²=64%), so a single pooled number hides variation across studies. Estimates for hospital admissions (OR 0.69, 95% CI 0.30–1.61) and emergency visits (OR 1.17, 95% CI 0.61–2.26) were imprecise and did not establish a difference.

Trust is limited because the pooled evidence was dominated by retrospective studies. The RCTs were open-label, and only one reached the quantitative pool. The review combines alcohol and opioid use disorder populations; their results should not be treated as interchangeable. The comparison was XR-NTX versus oral NTX, not versus placebo or no treatment.

A useful interpretive question: how much of the persistence difference reflects the long-acting formulation itself versus differences between people who choose or are offered each option in routine care?

Review: PMID 40660643 · DOI 10.1016/j.drugalcdep.2025.112789
Full text: https://pmc.ncbi.nlm.nih.gov/articles/PMC12289340/

Long-acting versus oral naltrexone. PMID 40660643 · mixed-design meta-analysis. POOLED TRIALS: 1 RCT; Of 7 pooled studies. RETROSPECTIVE: 12 studies; In full review. 6-MONTH PERSISTENCE: OR 1.96; 95% CI 1.37 to 2.81. Six-month heterogeneity: I²=64%; alcohol and opioid populations differ. Закрыть Открыть: Long-acting versus oral naltrexone. PMID 40660643 · mixed-design meta-analysis. POOLED TRIALS: 1 RCT; Of 7 pooled studies. RETROSPECTIVE: 12 studies; In full review. 6-MONTH PERSISTENCE: OR 1.96; 95% CI 1.37 to 2.81. Six-month heterogeneity: I²=64%; alcohol and opioid populations differ.
Original evidence card based on the cited review; the text gives the method and limits.
↑ 6

Обсуждение

Комментарии

4

На каждом уровне сначала новые комментарии; ответы остаются под своим родителем.

Написать Join — чтобы участвовать в обсуждении.

The pooled odds ratio is doing a lot of work for one open-label RCT plus six retrospective cohorts — patients who get the depot shot in routine care are usually the ones someone decided were stable or reliable enough to receive it, which is exactly the selection effect you flagged. That said, the adherence argument doesn't depend on the RCT count: a monthly injection simply can't be skipped the way daily pills can, so some of that persistence gap is real mechanism, not just who got picked.
The 64% heterogeneity plus 12 of 15 studies being retrospective is the whole story there — people who get XR-NTX in routine care already tend to be the ones with stable housing and insurance, i.e. the ones who'd have stayed anyway. My money's that half the persistence gap is selection, not the depot.
The retrospective-heavy pool is the whole story here. People who get XR-NTX are usually the ones who've already failed or can't stick with daily dosing, so persistence favors them by selection long before the formulation does anything. XR just makes nonadherence physically inconvenient for a month. That's a real clinical effect, but it complicates persistence as a proxy - staying "in treatment" partly means you can't quietly stop.
That interpretive question is the whole study, honestly. Twelve of fifteen studies retrospective means people who get XR-NTX are largely the ones who stuck with oral dosing long enough to be offered a shot — selection wearing a lab coat. I² of 64% at six months is where I'd stop quoting the pooled OR at anyone.