CoChat

Disulfiram Efficacy in the Treatment of Alcohol Dependence: A Meta-Analysis

Open access

The DOI matches its Crossref record; 69 of the 71 references it cites were found in a registry or an index; 1 of the 50 sources its claims cite were checked for support.

Abstract

BACKGROUND: Despite its success with compliant or supervised patients, disulfiram has been a controversial medication in the treatment of alcoholism. Often, study designs did not recognize a pivotal factor in disulfiram research, the importance of an open-label design. Our objectives are: (1) to analyze the efficacy and safety of disulfiram in RCTs in supporting abstinence and (2) to compare blind versus open-label studies, hypothesizing that blinded studies would show no difference between disulfiram and control groups because the threat would be evenly spread across all groups. METHODS AND FINDINGS: We searched PubMed, EMBASE and the Cochrane Central Register for RCTs on disulfiram use with alcoholics in comparison to any alcoholic control group. The primary outcome was defined by the authors of each trial. Additional analyses included: blind vs. open-label, with or without supervision, cocaine study or not, and type of control. Overall, the 22 included studies showed a higher success rate of disulfiram compared to controls Hedges'g = .58 (95%CI = .35-.82). When comparing blind and open-label RCTs, only open-label trials showed a significant superiority over controls g = .70 (95%CI = .46-.93). RCTs with blind designs showed no efficacy of disulfiram compared to controls. Disulfiram was also more effective than the control condition when compared to naltrexone g = .77, 95%CI = .52-1.02, to acamprosate g = .76, 95%CI = .04-1.48, and to the no disulfiram groups g = .43, 95%CI = .17-.69. LIMITS INCLUDE: (1) a population of 89% male subjects and (2) a high but unavoidable heterogeneity of the studies with a substantial I-square in most subgroups of studies. CONCLUSIONS: Blinded studies were incapable of distinguishing a difference between treatment groups and thus are incompatible with disulfiram research. Based on results with open-label studies, disulfiram is a safe and efficacious treatment compared to other abstinence supportive pharmacological treatments or to no disulfiram in supervised studies for problems of alcohol abuse or dependence.

Overview

CoChat Summary

TL;DR

This meta-analysis found that disulfiram was superior to controls in open-label randomized trials, but not in blinded trials, suggesting that blinding masks the key expectancy mechanism of the drug. It also found more adverse events with disulfiram, but no difference from controls in deaths or serious adverse events requiring hospitalization.

Key findings

  • Across 22 RCTs and 2,414 subjects, disulfiram had a higher success rate than controls: Hedges’ g = 0.58 (95% CI 0.35–0.82); heterogeneity was I² = 72%. After trim-and-fill correction, the effect remained significant: g = 0.47 (95% CI 0.24–0.71).
  • In the blind/open-label subgroup analysis, open-label trials showed benefit: g = 0.70 (95% CI 0.46–0.93), while blinded trials showed no efficacy: g = 0.01 (95% CI −0.29–0.32).
  • With supervision, disulfiram was significantly superior to controls: g = 0.82 (95% CI 0.59–1.05; I² = 46%); without supervision it was not significant: g = 0.26 (95% CI −0.02–0.53).
  • Compared with specific controls, disulfiram was superior to naltrexone (g = 0.77), acamprosate (g = 0.76), and no disulfiram (g = 0.43). It also showed a higher adverse-events rate ratio of 1.40 (95% CI 1.01–1.94).

Methods

  • Systematic review and meta-analysis of randomized controlled trials comparing oral disulfiram with any control group in alcohol-dependent/abusing patients; databases searched were PubMed, EMBASE, Cochrane Central Register, plus manual search (last searches June–November 2012).
  • Final sample included 22 studies published from 1973 to 2010; total analyzed sample was 2,414 subjects, with some studies involving comorbid cocaine dependence.
  • Main analysis used Hedges’ g under a random-effects model, with subgroup analyses by blinding, supervision, cocaine vs non-cocaine studies, and control condition, plus funnel plots, trim-and-fill, leave-one-out sensitivity analyses, and meta-regression.

Limitations

  • The paper states that the included studies had a high and unavoidable heterogeneity with substantial I² in most subgroups, and that there was a possible publication bias suggested by funnel-plot asymmetry in the overall and open-label analyses.
  • The paper states a sample composition limitation: 89% male subjects, so generalizability to women is limited.
  • Evident from the text: many studies were methodologically weak by Cochrane criteria—among the 22 studies, 6 met none of the six risk-of-bias criteria, and only 2 met four—and many outcomes were based on differing primary endpoints across trials.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1371/journal.pone.0087366

Cites

69 of 69 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 1 Flowchart of selection of studies for inclusion in the meta-analysis.
Figure 1 Flowchart of selection of studies for inclusion in the meta-analysis.CC BY · PubMed Central
Figure 2 Meta-analysis of Hedges' g effect-size of all RCTs comparing the efficacy of disulfiram and controls.
Figure 2 Meta-analysis of Hedges' g effect-size of all RCTs comparing the efficacy of disulfiram and controls.CC BY · PubMed Central
Figure 3 Meta-analysis for blinded versus open-label RCTs. Meta-analysis of Hedges' g effect-size comparing the efficacy of disulfiram and controls in blinded versus open-label RCTs.
Figure 3 Meta-analysis for blinded versus open-label RCTs. Meta-analysis of Hedges' g effect-size comparing the efficacy of disulfiram and controls in blinded versus open-label RCTs.CC BY · PubMed Central
Figure 4 Meta-analysis of RCTs with supervision versus no supervision. Meta-analysis of Hedges' g effect-size comparing the efficacy of disulfiram and controls in RCTs with supervision versus no supervision.
Figure 4 Meta-analysis of RCTs with supervision versus no supervision. Meta-analysis of Hedges' g effect-size comparing the efficacy of disulfiram and controls in RCTs with supervision versus no supervision.CC BY · PubMed Central
Figure 5 Meta-analysis of RCTs that included cocaine versus non cocaine subjects. Meta-analysis of Hedges' g effect-size comparing the efficacy of disulfiram and controls in RCTs that included alcohol dependent cocaine subjects versus those that did not include cocaine subjects.
Figure 5 Meta-analysis of RCTs that included cocaine versus non cocaine subjects. Meta-analysis of Hedges' g effect-size comparing the efficacy of disulfiram and controls in RCTs that included alcohol dependent cocaine subjects versus those that did not include cocaine subjects.CC BY · PubMed Central
Figure 6 Subgroup analysis of Hedges' g effect-size comparing the efficacy of disulfiram and controls by control types.
Figure 6 Subgroup analysis of Hedges' g effect-size comparing the efficacy of disulfiram and controls by control types.CC BY · PubMed Central
Figure 7 Meta-analysis of adverse events rate ratio comparisons in controls and disulfiram treated patients.
Figure 7 Meta-analysis of adverse events rate ratio comparisons in controls and disulfiram treated patients.CC BY · PubMed Central

Verification

Claims the paper makes

1 Valid · 0 Not valid · 21 Not checked · 74 Own results and interpretations

51 more claims

Create a free account to see all 96 claims and to verify the 49 references still open.

See all claims — free account
  • Probably not a substantial claim. 21
  • It may state two separate claims as one. 24

Create a free account to see these claims.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Citations

Cites

Research works printed but not linked: 2

  1. A De Sousa, 2009, An open label trial of naltrexone versus disulfiram in elderly patients with alcohol dependence, JPPS, 6, 85
  2. K Carroll, 2004, Efficacy of disulfiram and cognitive behavior therapy in cocaine-dependent outpatients, Archives of General Psychiatry, 61, 264, 10.1001/archpsyc.61.3.264 · DOI · Paper page

Cited by

See all 347

Metadata

Authors
Skinner, M. D., Lahmek, P., Pham, H., & Aubin, H.-J.
Affiliations
  • Inserm (FR)
  • Assistance Publique – Hôpitaux de Paris (FR)
  • Université Paris-Sud (FR)
Corresponding author
Skinner, Marilyn D.
Venue
PLoS ONE
Published
2014-02-10
Volume
9
Issue
2
Pages
e87366
ISSN
1932-6203 (ISSN-L)
OpenAlex
W2135168571
Cited by
347
References
71
Citations by year
2026: 24 · 2025: 21 · 2024: 43 · 2023: 51 · 2022: 28 · 2021: 31 · 2020: 26 · 2019: 37 · 2018: 23 · 2017: 18 · 2016: 23 · 2015: 20 · 2014: 2
FWCI
2.76
Type
journal article
Language
en
Open access
Locations
  • PLoS ONE (publishedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • Figshare (submittedVersion, cc-by)
Topics
  • Health Sciences › Medicine › Epidemiology › Substance Abuse Treatment and Outcomes
  • Health Sciences › Medicine › Pathology and Forensic Medicine › Alcohol Consumption and Health Effects
  • Health Sciences › Medicine › Clinical Neurology › Alcoholism and Thiamine Deficiency
Keywords
disulfiram, alcoholism, alcoholism treatment, abstinence, naltrexone, acamprosate, supervised treatment, cocaine use, treatment efficacy, drug safety
MeSH terms
Adolescent, Adult, Aged, Alcohol Deterrents, Alcoholism, Cocaine, Disulfiram, Female, Humans, Male, Middle Aged, Case-Control Studies, Randomized Controlled Trials as Topic, Treatment Outcome, Endpoint Determination

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

A language model drew the claims from the paper on 2026-10-11, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-11 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (qwen3.8-omni-flash).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.