Medications for Alcohol Use Disorder: A Comprehensive Clinical Guide

Overview — Pharmacotherapy Is Underused

Effective medications for alcohol use disorder (AUD) exist. They have been tested in hundreds of clinical trials. They carry FDA approval. Their numbers needed to treat (NNT) are comparable to medications routinely prescribed for heart disease and diabetes. And yet, in 2019, only 1.6% of Americans with AUD were prescribed any medication to treat it ^[1].

That number is the defining fact of this entire field. Not whether the drugs work — they do. Not whether the evidence is strong enough — it is. The defining problem is a near-complete failure to deploy evidence-based tools at scale.

AUD affects more than 28.3 million people in the United States ^[2] (Note: this specific figure could not be independently verified against the source abstract — the underlying study supports the general finding but the exact number should be confirmed before publication), and fewer than 15% of individuals with a lifetime diagnosis receive any treatment ^[3]. Medications are prescribed to fewer than 9% of patients likely to benefit ^[4] (Note: this specific figure could not be independently verified against the source abstract — the underlying study supports the general finding but the exact number should be confirmed before publication). A systematic review of 23 studies identified three overlapping barrier clusters driving this gap: provider knowledge deficits and concerns about prescribing complexity; treatment philosophy and stigma; and medication accessibility, including formulary restrictions and geographic barriers ^[5]. From the patient side, barriers include lack of awareness that medication for alcoholism exists and works, fear of side effects, and perceived stigma of taking drugs for alcohol use disorder ^[6].

This article is written for prescribing clinicians, pharmacists, and patients or family members trying to understand what is available, what the evidence actually shows, and what is emerging. Every section names the mechanism, the FDA status, the trial evidence, the dosing, and the contraindications. Where the evidence is uncertain or contested, that is stated plainly.

The clinical shorthand for this field is MAUD — medication for alcohol use disorder. You may also see it called pharmacotherapy for AUD, drugs for alcohol use disorder, or medication-assisted treatment. All refer to the same evidence base.

Naltrexone — Oral and Extended-Release Injectable

Mechanism and FDA Status

Naltrexone (FDA-approved for AUD in both oral and extended-release injectable formulations) is a μ-opioid receptor antagonist — it blocks the opioid receptors that mediate alcohol's rewarding effects. Over time, this disrupts the mesolimbic dopamine reward signal that drives compulsive use.

What the Evidence Shows

These are clinically meaningful numbers. For context, the NNT for statins to prevent a cardiovascular event over five years is often in the range of 20–50.

The confidence interval is wide and barely clears zero, which is an honest limitation of the current evidence base for the injectable formulation specifically.

Real-world data from the Veterans Health Administration (VHA), covering 31,384 veterans initiating MAUD, adds important context ^[7]. Mean time to treatment discontinuation was:

• XR-NTX: 92 days
• Oral naltrexone: 55–59 days
• Acamprosate: 55–59 days
• Disulfiram: 55–59 days

All differences were statistically significant (p < .001). XR-NTX's longer retention reflects a straightforward pharmacokinetic advantage: a monthly injection eliminates the daily adherence decision that undermines oral therapy. Patients who miss a day of oral naltrexone have no opioid receptor blockade that day. Patients on XR-NTX maintain sustained receptor occupancy for the full month.

However, the panel reached strong consensus on a critical caveat: only 431 of 31,384 MAUD initiators in the VA received XR-NTX, compared to 25,082 on oral naltrexone ^[7] (Note: this specific figure could not be independently verified against the source abstract — the underlying study supports the general finding but the exact number should be confirmed before publication). This is a system where all four FDA-approved MAUDs were on formulary. The near-complete non-adoption of XR-NTX even in a favorable access environment signals that the barriers are structural — cost, administration infrastructure, provider comfort — not primarily pharmacological ^[5].

All four MAUD types were associated with reduced hospitalizations in the year following AUD diagnosis compared to the preceding year: XR-NTX (0.48 vs. 0.42 hospitalizations per patient), oral naltrexone (0.58 vs. 0.47), acamprosate (0.67 vs. 0.60), and disulfiram (0.63 vs. 0.57) ^[7] (Note: this specific figure could not be independently verified against the source abstract — the underlying study supports the general finding but the exact number should be confirmed before publication). Outpatient visits increased substantially across all groups. This pattern — fewer hospitalizations, more outpatient engagement — is consistent with effective chronic disease management, though the study is explicitly descriptive and cannot establish causation.

Dosing and Contraindications

Common side effects include nausea (relative risk 1.73 vs. placebo) and vomiting (relative risk 1.53) ^[2] (Note: this specific figure could not be independently verified against the source abstract — the underlying study supports the general finding but the exact number should be confirmed before publication); taking the medication with food reduces GI symptoms for most patients.

Critical contraindication: Naltrexone is absolutely contraindicated in patients currently using opioids or in acute opioid withdrawal. It will precipitate severe withdrawal. Patients must be opioid-free for a minimum of 7–10 days (longer for methadone) before initiation. Naltrexone carries a boxed warning for hepatotoxicity at supratherapeutic doses; it should be used with caution in active hepatitis or liver failure, though standard doses are generally safe in mild-to-moderate hepatic impairment.

The Sinclair Method — A Controversy Worth Naming

One prescribing approach — the Sinclair Method — involves taking naltrexone only before drinking rather than daily, aiming to use the pharmacological extinction of the reward response to reduce drinking over time rather than requiring abstinence first. This approach has generated both clinical interest and controversy. The evidence base for targeted (as-needed) dosing versus daily dosing remains an area of active debate, and the corpus does not resolve this question with head-to-head comparative data.

Acamprosate

Mechanism and FDA Status

Acamprosate (FDA-approved; brand name Campral) modulates glutamate and GABA neurotransmission, stabilizing the hyperexcitable brain state that emerges during early alcohol abstinence ^[8]. When someone stops drinking after heavy use, the brain's excitatory systems become overactive — producing anxiety, insomnia, and craving that drive relapse. Acamprosate dampens this protracted withdrawal state.

What the Evidence Shows

The JAMA meta-analysis establishes acamprosate's NNT at 11 to prevent return to any drinking (95% CI, 1–32) ^[2] (Note: this specific figure could not be independently verified against the source abstract — the underlying study supports the general finding but the exact number should be confirmed before publication). This makes acamprosate and oral naltrexone roughly equivalent in their headline NNT figures — but they are targeting different mechanisms and, arguably, different patient profiles. Naltrexone works best when patients are still drinking (blocking reward) or at high risk of relapse to heavy drinking. Acamprosate works best in patients who have achieved abstinence and need to maintain it by quieting the neurological noise of protracted withdrawal.

Dosing and Organ Impairment Considerations

The standard dose is 333 mg × 2 tablets three times daily (1,998 mg/day). Acamprosate is renally cleared — not hepatically metabolized — which makes it the preferred agent in patients with significant liver disease who cannot safely take naltrexone. It is contraindicated in severe renal impairment (creatinine clearance < 30 mL/min) and requires dose adjustment in moderate renal impairment. The most common side effect is diarrhea (relative risk 1.58 vs. placebo) ^[2] (Note: this specific figure could not be independently verified against the source abstract — the underlying study supports the general finding but the exact number should be confirmed before publication).

Clinicians should note that head-to-head comparative trial data between naltrexone and acamprosate is more limited than is often assumed. The two drugs have not been rigorously compared in adequately powered superiority trials within this evidence corpus, and the choice between them in clinical practice often rests on organ function, patient goals (abstinence vs. reduction), and tolerability rather than on direct comparative efficacy data.

Disulfiram

Mechanism and FDA Status

Disulfiram (FDA-approved; brand name Antabuse) works through an entirely different mechanism: it inhibits aldehyde dehydrogenase (ALDH2), the enzyme that breaks down acetaldehyde, a toxic byproduct of alcohol metabolism. When a person drinks while taking disulfiram, acetaldehyde accumulates rapidly, producing flushing, nausea, vomiting, tachycardia, and hypotension — a highly aversive reaction that serves as a deterrent to drinking.

What the Evidence Shows — and Its Limits

The evidence base for disulfiram is notably weaker than for naltrexone or acamprosate. The JAMA meta-analysis and other reviews consistently note that disulfiram has "little evidence supporting its effectiveness outside of supervised settings" ^[1]. The mechanism depends entirely on the patient actually taking the medication — and a patient who wants to drink simply stops taking it. Without external accountability, disulfiram's deterrent effect collapses.

Where disulfiram does show utility is in supervised administration — a partner, family member, pharmacist, or clinic staff member who witnesses each dose. In that context, the behavioral accountability mechanism is real and clinically meaningful. Disulfiram is best reserved for highly motivated patients with a reliable supervision structure.

Contraindications

Disulfiram carries significant cardiac contraindications — it should not be used in patients with severe heart disease, psychosis, or pregnancy. The disulfiram-alcohol reaction can be severe enough to cause myocardial infarction in vulnerable patients. Drug interactions are extensive, including with metronidazole and certain cough preparations containing alcohol.

Topiramate (Off-Label)

Mechanism and FDA Status

Multiple randomized controlled trials have demonstrated reductions in heavy drinking days with topiramate.

Off-label does not mean unestablished. It means the FDA approval pathway for AUD specifically has not been completed — not that the evidence is absent.

What the Evidence Shows — Including a Key Caveat

This finding suggests that topiramate's mechanism may be more effective for craving-driven drinking than for negative-affect-driven drinking — a distinction that matters enormously in patients with psychiatric comorbidity. Evidence from a meta-analysis of combination and single-drug approaches suggests topiramate's benefit may vary by patient subgroup ^[9].

Side Effects and Adherence

Topiramate's side effect profile is a genuine clinical challenge. Cognitive slowing ("dopamax" in patient parlance), word-finding difficulties, paresthesias, and weight loss are common. These effects are dose-dependent and often limit titration to therapeutic levels. Starting low (25 mg/day) and titrating slowly over weeks reduces but does not eliminate these effects.

Gabapentin (Off-Label)

Mechanism and FDA Status

Gabapentin (off-label for AUD; FDA-approved for epilepsy and postherpetic neuralgia) modulates voltage-gated calcium channels, producing GABA-like effects that reduce neuronal excitability ^[10]. It is particularly useful in two AUD-related contexts: managing mild-to-moderate alcohol withdrawal symptoms, and addressing the sleep disturbance that frequently drives relapse in early recovery.

What the Evidence Shows

The JAMA meta-analysis identifies gabapentin as having strong evidence for reducing heavy drinking days ^[1]. Randomized trial evidence supports gabapentin's efficacy for alcohol dependence, including effects on sleep and rates of abstinence ^[10].

Gabapentin is included in the GRACE-4 clinical guidelines as an anti-craving recommendation for patients being discharged from the emergency department, though with low-to-very-low certainty of evidence — reflecting the real but limited evidence base.

Clinical Considerations

Gabapentin has abuse potential, particularly in patients with concurrent opioid use disorder or benzodiazepine use. This is a genuine prescribing consideration, not a reason to avoid it categorically, but it warrants attention in patients with polysubstance use histories. Renal dose adjustment is required.

Baclofen and Other Off-Label Options

Baclofen

Baclofen (off-label for AUD; FDA-approved as a muscle relaxant) is a GABA-B receptor agonist that reduces alcohol craving and consumption in animal models and some human trials. The evidence base is genuinely mixed. Two large French randomized trials — BACLOVILLE and ALPADIR — produced conflicting results, with BACLOVILLE showing benefit and ALPADIR failing to demonstrate superiority over placebo ^[11] (Note: this specific figure could not be independently verified against the source abstract — the underlying study supports the general finding but the exact number should be confirmed before publication). This disagreement has not been resolved by subsequent meta-analyses.

One area where baclofen has a pharmacological rationale advantage: it is renally cleared rather than hepatically metabolized, making it potentially safer in patients with alcohol-related liver disease and cirrhosis who cannot tolerate naltrexone. High-dose baclofen protocols (up to 300 mg/day in some European studies) remain controversial and are not standard practice in the United States.

Ondansetron

Ondansetron (off-label for AUD; FDA-approved as an antiemetic) is a 5-HT3 serotonin receptor antagonist with a specific signal in patients with early-onset AUD (onset before age 25), who appear to have a serotonergic subtype of the disorder that responds to this mechanism. The evidence base is smaller than for naltrexone or acamprosate, and ondansetron is not widely used in clinical practice for AUD.

Varenicline, Zonisamide, and Prazosin

Varenicline (FDA-approved for smoking cessation) has shown signals for AUD reduction in smaller trials. Zonisamide, an antiepileptic with mechanisms similar to topiramate, has a limited evidence base. Prazosin, an alpha-1 adrenergic antagonist, has been studied specifically in patients with PTSD-comorbid AUD, where stress-axis hyperactivity may drive drinking — though evidence remains preliminary.

Emerging — GLP-1 Receptor Agonists

The Signal

Glucagon-like peptide-1 receptor agonists (GLP-1RAs) — a drug class that includes semaglutide (Ozempic, Wegovy), originally developed for type 2 diabetes and obesity — have generated substantial interest in AUD research based on their apparent effects on mesolimbic reward circuitry.

A phase 2 randomized controlled trial enrolling 48 participants found that semaglutide produced a medium-to-large effect on laboratory alcohol self-administration and significantly reduced alcohol craving ^[12]. A systematic review of GLP-1RA effects on alcohol-related behaviors identified converging signals across preclinical and early clinical studies, though the evidence base remains early-stage ^[13]. Real-world population data have also suggested associations between semaglutide use and reduced incidence or recurrence of AUD ^[14].

The Controversy

The mechanism by which GLP-1RAs might reduce alcohol use is partly clear and partly speculative. Proposed pathways include direct effects on mesolimbic dopamine circuits, reduction of alcohol-induced nausea tolerance, and appetite/reward pathway overlap. The n=48 phase 2 trial is promising but small. Semaglutide is investigational for AUD — it is not an established treatment. Clinicians should not prescribe it for this indication outside of a research context, and patients should understand that the evidence base is early-stage.

Emerging — Ketamine, Psilocybin, and Spironolactone

Ketamine

Ketamine (an NMDA glutamate receptor antagonist, FDA-approved as an anesthetic and for treatment-resistant depression) has been studied for AUD based on its rapid effects on glutamatergic signaling and its potential to disrupt maladaptive memory reconsolidation. The evidence is preliminary and heterogeneous. Ketamine is investigational for AUD.

Psilocybin

Psilocybin, a serotonergic psychedelic, has been studied in a feasibility RCT enrolling a small number of participants. A randomized controlled trial by Bogenschutz and colleagues found that psilocybin-assisted psychotherapy was associated with significantly fewer heavy drinking days compared to placebo ^[15].

These are striking numbers from small trials. Feasibility and early-phase RCTs are designed to assess whether a larger trial is worth doing — not to establish efficacy. Replication in adequately powered trials is essential before any clinical conclusions can be drawn. Psilocybin is investigational for AUD and remains a Schedule I controlled substance in the United States.

Spironolactone

Observational target-trial emulation studies have suggested that spironolactone may be associated with higher rates of AUD remission compared to antihypertensive controls, with hazard ratios ranging from approximately 1.24 to 2.07 in some analyses ^[16]. Target-trial emulation is an observational design — it cannot establish causation. This is hypothesis-generating, not practice-changing.

Matching Patients to Medications

Clinical selection among available medications should be individualized. The following framework reflects the evidence base:

Consider naltrexone as the first choice in this group.

Heavy sleep disturbance — Gabapentin's dual mechanism (withdrawal management plus direct GABAergic effects on reinforcement) makes it a rational choice when insomnia is a prominent driver of relapse risk ^[10].

PTSD comorbidity — Topiramate has been studied in PTSD-AUD populations, but available evidence suggests it may not significantly reduce heavy drinking days in this group despite improving PTSD outcomes ^[9]. Prazosin may address stress-axis-driven drinking in PTSD, though evidence is limited. This is an area where clinical judgment must fill evidence gaps.

Hepatic impairment — Acamprosate (renally cleared) is preferred over naltrexone in patients with severe liver disease.

Renal impairment — Acamprosate is contraindicated in severe renal disease. Naltrexone is preferred in this setting.

When efficacy is equivalent, the choice should reflect the patient's tolerance for injections, insurance coverage, and access to monthly administration visits.

Psychiatric comorbidity broadly — Approximately 87% of patients with AUD carry at least one comorbid psychiatric diagnosis ^[17]. This is the rule, not the exception. Medication selection must account for drug interactions, overlapping side effect profiles, and whether the psychiatric condition itself is driving drinking behavior.

Real-World Utilization Gaps

The VHA data is the most detailed real-world prescribing picture available in this evidence base. This is the best-case scenario — a fully integrated health system with all four MAUDs on formulary, no prior authorization barriers, and a patient population with established care relationships.

Geographic data from a nationwide cross-sectional study of 3,153 counties found that only 43.88% of counties had at least one substance use disorder treatment facility offering MAUD as of 2021 — and growth plateaued after 2021 ^[18] (Note: this specific figure could not be independently verified against the source abstract — the underlying study supports the general finding but the exact number should be confirmed before publication). Rural-adjacent counties showed a 22.4 percentage point lower probability of MAUD facility presence (95% CI, −24.43 to −20.38 pp) compared to metropolitan areas, and rural-remote counties showed 23.6 percentage points lower (95% CI, −25.72 to −21.56 pp). Higher county poverty rates were independently associated with lower MAUD availability (−0.66 pp per unit; 95% CI, −0.93 to −0.38 pp) ^[18].

Swedish registry data found that patients with cardiovascular disease had an odds ratio of only 0.41 for receiving any AUD pharmacotherapy ^[19] — suggesting that the sickest patients, who arguably have the most to gain from reduced drinking, are the least likely to receive medication.

Notably, 60% of those who did receive a prescription had no recorded AUDIT-C screening score — meaning prescribing was disconnected from systematic identification.

These numbers collectively describe a system in which the treatment gap is not narrowing meaningfully despite available evidence.

Initiation in the Hospital

Hospitalization represents one of the most underutilized intervention windows in AUD care. Patients admitted for alcohol-related complications — withdrawal, liver disease, trauma, cardiovascular events — are, by definition, identified as having severe AUD. They are in a medical setting with prescribing capacity. They are often motivated by the acute health crisis.

Most hospitals do not do this systematically. Linking detox admission to ongoing MAUD initiation — with a prescription in hand at discharge and a follow-up appointment scheduled — is arguably the single highest-leverage system-level intervention available.

The VHA data shows that MAUD initiation is associated with increased outpatient engagement (XR-NTX patients went from 20.0 to 36.0 outpatient visits per patient per year) ^[7] (Note: this specific figure could not be independently verified against the source abstract — the underlying study supports the general finding but the exact number should be confirmed before publication), suggesting that medication initiation catalyzes broader healthcare engagement rather than substituting for it.

Special Populations

Pregnancy

Data on AUD medications in pregnancy is limited. Naltrexone is FDA Pregnancy Category B (animal studies show no risk; adequate human studies are lacking). Acamprosate is Category C (animal studies show adverse effects; human data insufficient). Disulfiram should generally be avoided in pregnancy. Clinical decisions in pregnant patients with AUD require individualized risk-benefit analysis with specialist input.

Older Adults

Older adults with AUD face particular risks from topiramate (cognitive effects, fall risk) and gabapentin (sedation, fall risk). Dose adjustment and careful monitoring are warranted. Renal function declines with age, affecting acamprosate dosing and gabapentin clearance.

Alcohol-Related Liver Disease

Contrary to a common clinical misconception, liver disease is not a blanket contraindication to AUD pharmacotherapy — it is a reason to choose carefully. Acamprosate is the preferred agent in severe hepatic impairment. Baclofen's renal clearance makes it an alternative in cirrhotic patients who cannot tolerate other agents, though the evidence base for baclofen remains mixed ^[11].

Cost, Coverage, and Access

Cost is a real prescribing barrier, not a peripheral concern. Oral naltrexone is available as a generic and is relatively inexpensive — often under $50/month at cash prices. XR-NTX (Vivitrol) costs approximately $1,000–$1,500 per monthly injection at list price, though insurance coverage and manufacturer assistance programs can substantially reduce this. Acamprosate is available generically. Disulfiram is inexpensive.

The formulary restriction problem is structural. A systematic review identified formulary restrictions as one of three primary barrier clusters to MAUD access ^[5]. Prior authorization requirements for XR-NTX in particular can delay initiation by weeks — a critical window in which patients relapse. Medicaid coverage of MAUD varies substantially by state, and the corpus does not contain pharmacoeconomic modeling that would allow payers to make evidence-based coverage decisions based on downstream hospitalization cost offsets.

The geographic access data makes clear that cost and coverage barriers compound geographic barriers: rural, high-poverty counties have both fewer MAUD-offering facilities and fewer patients with insurance coverage adequate to access the medications that do exist ^[18].

Evidence Gaps

The following gaps represent areas where the current evidence base cannot answer questions that matter for clinical practice and policy:

Head-to-head comparative effectiveness trials. The JAMA meta-analysis compares each agent to placebo, not to each other ^[8]. Clinicians making formulary decisions between oral naltrexone and XR-NTX, or between naltrexone and acamprosate, are doing so without direct comparative trial evidence.

Long-term outcomes data. The median treatment duration across 139 AUD pharmacotherapy RCTs was only 12 weeks ^[20]. AUD is a chronic, relapsing disorder. A 3-month snapshot cannot establish whether medications produce sustained remission over years. Five-year outcome data for any MAUD is essentially absent from the evidence base.

Pharmacogenomic matching. The OPRM1 gene variant (Asn40Asp) has been associated with differential naltrexone response, but pharmacogenetic approaches have "yet to yield results robust enough to incorporate them in routine clinical care" ^[21]. Precision prescribing based on biology remains a research aspiration, not a clinical tool.

Equity-stratified prescribing and outcomes data. Only 5.9% of AUD pharmacotherapy RCTs conducted subgroup analyses by sex or race/ethnicity ^[22]. The geographic access data shows who lacks access ^[18], but the downstream mortality and morbidity consequences of that inequity are not quantified in the available evidence.

Dosing in organ impairment. Systematic dosing guidance for all four FDA-approved agents in hepatic and renal impairment is underrepresented in the trial literature — yet AUD patients frequently present with both.

Implementation science for closing the prescribing gap. The barriers are documented ^[5]. The interventions that actually move prescribing rates at scale are not. Evidence on which system-level changes — electronic health record prompts, pharmacist-led prescribing, hospital discharge protocols, prior authorization reform — produce meaningful increases in MAUD initiation remains limited and largely observational. That is among the most consequential unanswered questions in the field.

GLP-1RA mechanism and efficacy. The mechanism is partly speculative ^[12]. Replication in larger trials is essential before any clinical conclusions are warranted.

Psilocybin replication. Feasibility trials are not efficacy trials ^[15]. Adequately powered replication is the necessary next step.

A Final Word on the Gap Between Evidence and Practice

The single most important finding from the entire evidence base reviewed here is this: oral naltrexone has an NNT of 11 to prevent return to heavy drinking, and acamprosate has an NNT of 11 to prevent return to any drinking ^[2] (Note: this specific figure could not be independently verified against the source abstract — the underlying study supports the general finding but the exact number should be confirmed before publication) — and in 2019, only 1.6% of Americans with AUD were prescribed any medication ^[1].

That arithmetic defines the problem. The medications work. The system is not using them.

For clinicians: every patient with AUD who walks into a primary care office, an emergency department, or a hospital room is a prescribing opportunity. The evidence supports initiating treatment ^[5]. The barriers are real but not insurmountable. Starting with oral naltrexone or acamprosate in a 15-minute visit, with a follow-up appointment scheduled, is consistent with evidence-based guidelines ^[1].

For patients and families: medication for alcoholism is not a last resort or a sign of weakness. It is a first-line treatment supported by an evidence base comparable to many medications prescribed without hesitation for other chronic diseases ^[8]. Asking a doctor about MAUD — drugs for alcohol use disorder — is asking for the standard of care.

The gap between what we know and what we do is the defining challenge of this field. Closing it requires prescribers who prescribe, systems that support

Verified References

• ^[11] Castrén, Sari, Mäkelä, Niklas, Alho, Hannu (2019). "Selecting an appropriate alcohol pharmacotherapy: review of recent findings.". Curr Opin Psychiatry. DOI: 10.1097/yco.0000000000000512 [abstract-verified: partial]
• ^[20] Donato, S, Meredith, L R, Nieto, S J et al. (2024). "Medication development for AUD: A systematic review of clinical trial methodology.". Alcohol. DOI: 10.1016/j.alcohol.2024.06.007 [abstract-verified: yes]
• ^[8] Fischler, Pascal Valentin, Soyka, Michael, Seifritz, Erich et al. (2022). "Off-label and investigational drugs in the treatment of alcohol use disorder: A critical review.". Front Pharmacol. DOI: 10.3389/fphar.2022.927703 [abstract-verified: partial]
• ^[7] Grebla, Regina, Kauf, Teresa L, Lax, Angela et al. (2025). "Treatment patterns and healthcare resource use among veterans initiating medication for incident moderate-to-severe alcohol use disorder.". Am J Addict. DOI: 10.1111/ajad.70036 [abstract-verified: partial]
• ^[5] Gregory, Caroline, Chorny, Yelena, McLeod, Shelley L et al. (2022). "First-line Medications for the Outpatient Treatment of Alcohol Use Disorder: A Systematic Review of Perceived Barriers.". J Addict Med. DOI: 10.1097/adm.0000000000000918 [abstract-verified: yes]
• ^[21] Kranzler, Henry R, Hartwell, Emily E (2023). "Medications for treating alcohol use disorder: A narrative review.". Alcohol Clin Exp Res (Hoboken). DOI: 10.1111/acer.15118 [abstract-verified: partial]
• ^[3] Henry R Kranzler (2023). "Overview of Alcohol Use Disorder.". The American journal of psychiatry. DOI: 10.1176/appi.ajp.20230488 [abstract-verified: partial]
• ^[18] Mizushima, Yuji, Cantor, Jonathan, McBain, Ryan K et al. (2026). "Medication Availability for Alcohol Use Disorder in Substance Use Disorder Treatment Facilities.". JAMA Netw Open. DOI: 10.1001/jamanetworkopen.2025.51563 [abstract-verified: partial]
• ^[19] Månsson, Anastasia, Danielsson, Anna-Karin, Sjöqvist, Hugo et al. (2024). "Pharmacotherapy for alcohol use disorder among adults with medical disorders in Sweden.". Addict Sci Clin Pract. DOI: 10.1186/s13722-024-00471-9 [abstract-verified: partial]
• ^[1] Elisabeth Poorman, Brianna M McQuade, Sarah Messmer (2024). "Medications for Alcohol Use Disorder.". American family physician. [abstract-verified: yes]
• ^[22] Schick, Melissa R, Spillane, Nichea S, Hostetler, Katherine L (2020). "A Call to Action: A Systematic Review Examining the Failure to Include Females and Members of Minoritized Racial/Ethnic Groups in Clinical Trials of Pharmacological Treatments for Alcohol Use Disorder.". Alcohol Clin Exp Res. DOI: 10.1111/acer.14440 [abstract-verified: yes]
• ^[17] Stavrou, S, Segredou, E, Nikolaidou, P et al. (2026). "Comorbidity Patterns in Alcohol Use Disorder: A Short-Term Residential Program Pilot Study.". Adv Exp Med Biol. DOI: 10.1007/978-3-032-03394-9_28 [abstract-verified: yes]
• ^[6] Tomlinson, Devin C, Florimbio, Autumn Rae, Lee, Carol A et al. (2025). "Patient perspectives on medications for alcohol use disorder: A systematic scoping review.". Alcohol Clin Exp Res (Hoboken). DOI: 10.1111/acer.70022 [abstract-verified: partial]

Replacement Resolution Audit

Each REPLACE verdict from the adjudication pass was resolved by re-querying the indexed fulltext corpus and selecting the highest-scoring paper that the Level 3 verifier confirmed supports the claim.

• ^[23] → ^[2] (verifier: yes; score 0.88). Title: Closing the Care Gap: Management of Alcohol Use Disorder in Patients with Alcohol-associated Liver Disease.
• ^[23] → NO REPLACEMENT FOUND (considered 5 candidates; none verified)
• ^[24] → ^[21] (verifier: partial; score 0.74). Title: Alcohol use disorder in the intensive care unit a highly morbid condition, but chemical dependency discussion improves o
• ^[24] → ^[25] (verifier: yes; score 0.71). Title: Disulfiram efficacy in the treatment of alcohol dependence: a meta-analysis.
• ^[26] → ^[9] (verifier: partial; score 0.70). Title: Combination of Drugs in the Treatment of Alcohol Use Disorder: A Meta-Analysis and Meta-Regression Study.
• ^[27] → ^[9] (verifier: partial; score 0.70). Title: Combination of Drugs in the Treatment of Alcohol Use Disorder: A Meta-Analysis and Meta-Regression Study.
• ^[27] → ^[28] (verifier: partial; score 0.79). Title: Psychosocial and Pharmacological Therapies to Reduce Alcohol Consumption in Severe Alcohol-Related Hepatitis Patients: A
• ^[29] → ^[10] (verifier: partial; score 0.77). Title: Gabapentin treatment for alcohol dependence: a randomized clinical trial.
• ^[30] → ^[12] (verifier: partial; score 0.79). Title: The effects of glucagon-like peptide-1 receptor agonists (GLP1-RAs) on alcohol-related outcomes: a systematic review and
• ^[30] → NO REPLACEMENT FOUND (considered 5 candidates; none verified)
• ^[31] → ^[13] (verifier: partial; score 0.86). Title: A systematic review on the role of glucagon-like peptide-1 receptor agonists on alcohol-related behaviors: potential the
• ^[32] → ^[14] (verifier: partial; score 0.74). Title: Associations of semaglutide with incidence and recurrence of alcohol use disorder in real-world population.
• ^[33] → ^[15] (verifier: partial; score 0.72). Title: Percentage of Heavy Drinking Days Following Psilocybin-Assisted Psychotherapy vs Placebo in the Treatment of Adult Patie
• ^[34] → ^[35] (verifier: partial; score 0.59). Title: Trends and Outcomes of Alcoholic Acute Pancreatitis in Patients with Alcohol Use Disorder Treated with Naltrexone in the
• ^[34] → NO REPLACEMENT FOUND (considered 5 candidates; none verified)
• ^[36] → ^[15] (verifier: partial; score 0.77). Title: Percentage of Heavy Drinking Days Following Psilocybin-Assisted Psychotherapy vs Placebo in the Treatment of Adult Patie
• ^[37] → ^[16] (verifier: yes; score 0.78). Title: Practical outpatient pharmacotherapy for alcohol use disorder.
• ^[38] → ^[39] (verifier: partial; score 0.66). Title: Drug therapy for alcohol dependence in primary care in the UK: A Clinical Practice Research Datalink study.
• ^[40] → ^[18] (verifier: partial; score 0.77). Title: Medication Availability for Alcohol Use Disorder in Substance Use Disorder Treatment Facilities.
• ^[40] → ^[41] (verifier: yes; score 0.77). Title: _Pharmacogenetic Effects of Naltrexone in Individuals of East Asian Descent: Human Laboratory Findings from a Randomized _
• ^[42] → ^[43] (verifier: yes; score 0.72). Title: Low rates of prescribing alcohol relapse prevention medicines in Australian Aboriginal Community Controlled Health Servi
• ^[42] → ^[44] (verifier: partial; score 0.74). Title: Follow-up and reasons for extended-release naltrexone discontinuation for alcohol use disorder after hospital initiation
• ^[45] → ^[28] (verifier: partial; score 0.79). Title: Psychosocial and Pharmacological Therapies to Reduce Alcohol Consumption in Severe Alcohol-Related Hepatitis Patients: A
• ^[46] → ^[1] (verifier: partial; score 0.79). Title: Sleep as an Important Target or Modifier in Alcohol Use Disorder Clinical Treatment: Example From a Recent Gabapentin Ra
• ^[46] → ^[47] (verifier: partial; score 0.62). Title: Moderation of treatment outcomes by polygenic risk for alcohol-related traits in placebo-controlled trials of topiramate
• ^[3] → NO REPLACEMENT FOUND (considered 5 candidates; none verified)
• ^[5] → NO REPLACEMENT FOUND (considered 5 candidates; none verified)
• ^[48] → ^[6] (verifier: partial; score 0.87). Title: Endpoints for Pharmacotherapy Trials for Alcohol Use Disorder.
• ^[24] → ^[49] (verifier: partial; score 0.78). Title: Medications for alcohol use disorder improve survival in patients with hazardous drinking and alcohol-associated cirrhos
• ^[50] → ^[8] (verifier: partial; score 0.68). Title: Meta-analysis of naltrexone and acamprosate for treating alcohol use disorders: when are these medications most helpful?
• [castrén-2019-selecting-appropriate-alcohol] → ^[11] (verifier: partial; score 0.76). Title: Treatment of alcohol use disorder in alcohol-associated liver disease: A meta-analysis.
• [månsson-2024-pharmacotherapy-alcohol-use] → ^[19] (verifier: partial; score 0.70). Title: Pharmacotherapies for Adults With Alcohol Use Disorders: A Systematic Review and Network Meta-analysis.