Alcohol Use Disorder: Clinical Presentation

A Knowledge Base for Clinicians, Researchers, and Families

Overview: Why Presentation Matters — and Why Recognition Fails

Alcohol use disorder (AUD) affects more than 28.3 million people in the United States ^[1] (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). It is one of the most prevalent, most burdensome, and most treatable conditions in medicine. It is also one of the most systematically missed.

The gap between how many people have AUD and how many receive a diagnosis or treatment is not a rounding error — it is the defining feature of this condition's clinical story. In Germany, only approximately 10% of people with AUD are treated by the professional help system [stüben-2023-evaluation-primary-health]. In the United States, using 2015–2019 National Survey on Drug Use and Health (NSDUH) data, only 52.9% of adults with AUD who visited a healthcare provider were even asked about their alcohol use — and the cascade collapses from there: 21.6% were asked about problematic use, 17.7% were advised to reduce consumption, and just 7.6% were offered treatment information ^[2]. Even when AUD is identified, medications with proven efficacy are prescribed to fewer than 9% of patients likely to benefit ^[3] (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).

Recognition is the first step toward any intervention. Without it, nothing else in the treatment cascade can begin. This article is organized around that premise: understanding how AUD presents — behaviorally, cognitively, physically, in clinical settings, and at home — is not an academic exercise. It is the prerequisite for closing a gap that costs lives.

A note on what this article can and cannot establish: the expert panel that produced this synthesis worked from a specific corpus of verified research documents. Where that corpus is thin or silent — particularly on withdrawal severity prediction, CIWA-Ar protocols, and individual-level predictors of diagnosis — this article says so explicitly. Those silences are clinically meaningful.

Behavioral Signs

Behavioral changes are often the first signs that family members notice and the last signs that the person with AUD recognizes in themselves. The DSM-5 organizes AUD around 11 criteria; behavioral signs map directly onto several of them.

Drinking more or longer than intended (DSM-5 Criterion 1) is among the most common early signs. A person plans to have two drinks and has six; a social occasion extends into a solo continuation at home. From the inside, this often feels like a one-time lapse rather than a pattern.

Persistent desire or unsuccessful efforts to cut down (Criterion 2) may appear as repeated "I'm taking a break starting Monday" announcements that don't hold, or as rules about drinking — only on weekends, only wine, never before 5 p.m. — that gradually erode.

Significant time spent obtaining, using, or recovering from alcohol (Criterion 3) can look like long afternoons "running errands," mornings lost to headache and nausea, or social calendars organized around drinking occasions.

Craving (Criterion 4) is discussed further under Cognitive Signs, but behaviorally it may manifest as irritability or restlessness when alcohol is unavailable, or as difficulty staying present in situations where drinking isn't possible.

Failure to fulfill major role obligations (Criterion 5) — missed work, neglected parenting responsibilities, declining performance — tends to appear later in the course but is often what prompts family members to seek information.

Continued use despite social or interpersonal problems (Criterion 6) and giving up important activities (Criterion 7) reflect the narrowing of a person's life around alcohol that characterizes more severe AUD.

Use in physically hazardous situations (Criterion 8) — driving after drinking, operating machinery, mixing alcohol with sedating medications — may be the presenting feature in an emergency department visit.

Continued use despite knowing it causes physical or psychological harm (Criterion 9) is a criterion that requires some self-awareness; its absence does not mean the criterion isn't met, because insight is often impaired.

Tolerance (Criterion 10) and withdrawal (Criterion 11) are discussed under Physical Signs.

A natural history note: a prospective Australian cohort found cumulative AUD incidence of 58% (95% CI 52.3–63.8%) from late adolescence to age 42, peaking sharply at age 24 ^[4]. Approximately 11–13% showed persistent symptoms from late adolescence onward — meaning a substantial minority of patients presenting in midlife have been carrying unrecognized AUD for decades, with behavioral signs that have been normalized over time.

Cognitive Signs

Cognitive features of AUD are less visible than behavioral ones and are rarely the presenting complaint — yet they are central to why the disorder is self-sustaining.

Preoccupation with drinking is the cognitive expression of craving (DSM-5 Criterion 4). It may appear as persistent mental planning around when and where the next drink will occur, difficulty concentrating on tasks when alcohol is anticipated, or intrusive thoughts about drinking during abstinent periods.

Difficulty estimating one's own intake is clinically important and underappreciated. People with AUD frequently underreport consumption — not always deliberately, but because tolerance has shifted their internal reference point for what constitutes "a lot." Standard drink definitions (14 grams of pure alcohol) are poorly understood by most patients, and pour sizes at home routinely exceed standard measures.

Anticipatory reward thinking — the cognitive rehearsal of the pleasure or relief that drinking will bring — is a mechanism that drives continued use even when the person has experienced repeated negative consequences. This is the "enhancement motive" that research identifies as the most prevalent drinking motive [sjödin-2026-drinking-motives-among], and it predicts changes in AUD severity over time.

Executive function effects — impaired planning, reduced impulse control, difficulty weighing long-term consequences against short-term relief — are both a consequence of heavy alcohol use and a factor that makes behavior change harder. These effects may not be apparent in a brief clinical encounter but can be elicited through careful history-taking.

The corpus does not contain neuropsychological testing data linking specific cognitive profiles to AUD recognition probability. This is a gap.

Physical Signs

Physical signs of AUD span a wide severity range, from subtle early markers to life-threatening complications. They are often the route through which AUD first enters the medical record — not as AUD, but as its consequences.

Tolerance (DSM-5 Criterion 10): The need for markedly increased amounts of alcohol to achieve the same effect, or a markedly diminished effect with continued use of the same amount. Clinically, this may be elicited by asking how many drinks it takes to feel the effects — a person who reports needing six or more drinks to feel "buzzed" is describing significant tolerance.

Withdrawal (Criterion 11): Approximately one-half of patients with AUD experience symptoms of alcohol withdrawal syndrome when decreasing or stopping alcohol use abruptly ^[5]. Withdrawal symptoms range from mild (tremor, diaphoresis, anxiety, insomnia, nausea) to severe (seizures, delirium tremens). The corpus does not contain CIWA-Ar validation data or clinical predictors of severe withdrawal — a significant patient safety gap that clinicians should address by consulting ASAM guidelines directly.

Hand tremor is one of the most observable early withdrawal signs and may be present at morning clinical appointments in patients with significant physical dependence.

Sleep disruption is nearly universal in AUD. Alcohol disrupts sleep architecture, suppressing REM sleep and causing early-morning awakening. Patients often present with insomnia complaints without volunteering a drinking history [corpus-gap].

Gastrointestinal symptoms: Nausea, vomiting, epigastric pain, and diarrhea are common, both as withdrawal features and as direct effects of heavy alcohol use on the GI tract.

Hypertension: Persistent or poorly controlled hypertension in a patient without other clear etiology should prompt alcohol screening [corpus-gap]. The relationship between heavy alcohol use and elevated blood pressure is well established, though the corpus does not provide specific effect size data for this association.

Dermatologic changes: Facial flushing, spider angiomata, palmar erythema, and jaundice may appear in patients with alcohol-associated liver disease. These are later-stage signs.

Weight changes: Both weight gain (from caloric content of alcohol) and weight loss (from nutritional displacement and GI effects) can occur.

Alcohol odor: The presence of alcohol odor at a clinical appointment — particularly a morning appointment — is a significant clinical signal that should prompt systematic assessment rather than avoidance.

Presentation in Primary Care

AUD rarely walks into a primary care office announcing itself. It presents as something else: abnormal liver enzymes, uncontrolled hypertension, persistent insomnia, anxiety, depression, or recurrent GI complaints. The clinician's task is to recognize these as potential AUD signals and respond with systematic screening.

Indirect presentation patterns that should trigger alcohol screening include ^[6]:
- Unexplained elevation of GGT, AST, or ALT
- MCV elevation without other explanation
- Persistent or treatment-resistant hypertension
- Recurrent insomnia or sleep complaints
- Anxiety or depression that doesn't respond as expected to treatment
- Repeated trauma or injury

The AUDIT-C problem: Most primary care settings use the AUDIT-C (a 3-item abbreviated version of the full AUDIT) because it fits within a brief visit. However, the AUDIT-C is substantially less useful for identifying AUD specifically — likelihood ratios of only 1.8 (males) and 2.0 (females) for AUD ^[7]. The full AUDIT at a score ≥8 produces an LR of 6.5 (95% CI 3.9–11). Primary care has optimized for speed at the cost of diagnostic accuracy. The AUDIT-C is a reasonable initial screen for hazardous drinking; it is not adequate for AUD identification.

Screening is not diagnosis. A positive AUDIT screen identifies elevated risk and warrants further clinical assessment — a DSM-5-based clinical interview. The AUDIT does not diagnose AUD.

Recognition rates: USPSTF-recommended screening "is not always performed consistently or correctly in primary care" ^[5]. The Hallgren et al. Alcohol Symptom Checklist showed excellent test-retest reliability (ICC = 0.82 in primary care settings) ^[8], suggesting that when practical assessment tools are deployed, reliable results follow — but deployment remains inconsistent.

The comorbidity masking problem: Approximately 87% of patients with AUD in a residential program had at least one comorbid psychiatric disorder ^[9]. In primary care, this means that what presents as a primary mood or anxiety disorder may be AUD-driven, or that AUD may be obscured by the more visible psychiatric complaint. Distinguishing substance-induced from independent psychiatric disorders requires a period of abstinence that is "not always practicable" ^[10]. ADHD prevalence in substance use treatment settings is 21–23% [hernández-2025-adhd-alcohol-use] (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 population primary care clinicians are likely seeing without connecting the presentations.

The diagnostic yield of abnormal labs as AUD triggers: ^[6] (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) recommends ordering liver transaminases and GGT in patients where AUD is suspected, but the corpus does not provide data on what proportion of patients with unexplained GGT elevation or MCV elevation have undiagnosed AUD. This is a gap in the opportunistic case-finding literature that clinicians should be aware of.

Presentation in the Emergency Department

The emergency department is where AUD declares itself most acutely — through intoxication, trauma, withdrawal, and crisis. It is also, the evidence suggests, the highest-yield intervention window that medicine consistently fails to use.

The detection signal: In a retrospective study of 251,300 Veterans Administration ED encounters associated with alcohol intoxication, 79% of patients had positive AUDIT-C screens within 6 months of their visit ^[11]. This reframes the intoxicated ED patient: these presentations are not random social events but are, in the overwhelming majority of cases, presentations of diagnosable AUD. The authors' conclusion is explicit — "presentation to an emergency department with any detectable ethanol concentration should prompt intervention accordingly" ^[11].

Critical caveat: The Farkas paper establishes a detection signal, not a treatment pathway. The corpus contains no data on whether those positive screens triggered diagnosis, brief intervention, or referral. The 79% figure tells us who should have been intervened upon — not who was. This is the most important silence in the corpus.

Common ED presentations of AUD include:
- Acute intoxication (with or without trauma)
- Alcohol withdrawal (ranging from mild tremulousness to seizure and delirium tremens)
- Alcohol-associated liver disease, including acute alcoholic hepatitis
- Psychiatric crisis, where alcohol may be a primary driver or complicating factor
- Trauma — falls, motor vehicle crashes, assault

Psychiatric comorbidity in the ED: Among people with borderline personality disorder (BPD), a meta-analysis of 15,603 individuals found AUD prevalence of 55.28% ^[12] (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). BPD presentations to EDs in crisis are therefore highly enriched for co-occurring AUD.

The hospitalization-as-intervention-window argument: The Canadian guideline recommends universal screening and brief intervention ^[5]. The logic is that acute care contact — particularly hospitalization — represents a moment of heightened receptivity that primary care may not replicate. The corpus supports the theoretical case for this approach but does not contain ED-specific data on brief intervention delivery rates or outcomes.

GRACE-4 guidelines: These emergency department-specific AUD management guidelines are referenced in clinical practice but are not represented as a document in this corpus. Clinicians should consult GRACE-4 directly for ED-specific management protocols.

Withdrawal management gap: The corpus is entirely silent on CIWA-Ar protocols, symptom-triggered versus fixed-schedule benzodiazepine regimens, and clinical predictors of severe withdrawal. This is a patient safety gap. Clinicians should consult ASAM guidelines for withdrawal management standards.

Presentation at Home: What Families Notice

Family members and close contacts often recognize AUD before the person affected does — and often before any clinician does. Understanding what families observe is clinically useful because family concern is frequently what brings a person to care.

Earlier-in-day drinking is one of the most reliable signals that tolerance and physical dependence have developed. When drinking shifts from evenings to afternoons, or from afternoons to mornings, it reflects the body's need to maintain alcohol levels to prevent withdrawal.

Secretive behavior: Hidden bottles, concealed purchases, minimizing or lying about consumption, drinking alone before social events ("pre-loading") — these behaviors reflect awareness that the drinking pattern would be judged, and they often precede any clinical presentation.

Mood changes tied to drinking patterns: Irritability, anxiety, or restlessness in the hours before drinking; relief or mood normalization after the first drink; emotional volatility when alcohol is unavailable. Family members often describe this as "a different person" emerging in the hours before drinking.

Functional decline that is gradual and therefore normalized: Because AUD typically develops over months to years, family members may not recognize the cumulative change until a crisis makes it visible. The natural history data are relevant here: a prospective cohort found that 11–13% of individuals showed persistent AUD symptoms from late adolescence onward ^[4], meaning some families have been adapting to a slowly changing baseline for years.

The inside-outside discrepancy: The corpus contains no direct qualitative or patient-perspective research on how early internal signs are experienced and dismissed. The person in long-term recovery on this panel noted that tolerance and stress-relief drinking are "internally normalized" — framed as manageable, functional, or deserved — while outsiders see a pattern. This clinical observation is consistent with the corpus evidence but is not directly supported by peer-reviewed qualitative data in these documents.

Phenotypes: Age, Sex, Comorbidity, and Population

AUD does not present identically across all people. Recognizing phenotypic variation is essential for avoiding missed diagnoses in populations whose presentations don't match the "classic" pattern.

Women and Accelerated Harm Trajectory

The concept of "telescoping" — that women develop AUD-related harms more rapidly and at lower consumption levels than men — has clinical currency, but the corpus does not contain a direct study of this phenomenon. What the corpus does establish is that the AUDIT performs better in females than males (LR 6.9 vs. 3.8 for AUDIT ≥8) ^[7], a statistically significant difference (P=.003). This may reflect that women present with more severe AUD relative to their screening scores, or that the AUDIT's items are more sensitive to female drinking patterns. The telescoping hypothesis is clinically plausible but should be described as a working concept rather than a settled finding until methodology-specific studies are cited.

Older Adults

Older adults with AUD are more likely to present with falls, cognitive decline, medication interactions, or sleep complaints than with classic intoxication or withdrawal signs. The corpus does not contain older-adult-specific AUD presentation data. Clinicians should be aware that standard drink quantities have greater physiological impact in older adults due to changes in body composition and hepatic metabolism, and that the AUDIT may underperform in this population. NIAAA guidelines for older adults should be consulted directly.

Adolescents and Young Adults

The natural history data show AUD incidence peaking at age 24 ^[4], with binge-pattern drinking predominating in younger cohorts. The VHA recognition gap is largest in patients aged 18–34, where AUD prevalence was 22.4% against a diagnosis rate of only 6.9% ^[13] — the youngest, most clinically recoverable cohort, and the most systematically missed.

Metabolic Dysfunction and AUD (MetAUD)

The intersection of AUD with metabolic liver disease (metabolic-dysfunction-associated steatotic liver disease, MASLD) creates a compounded hepatic and psychiatric burden. ^[14] describes a framework for detecting alcohol use, diagnosing AUD, and directing to treatment in the liver disease context. The corpus does not provide prevalence data for this phenotype, but its clinical importance is established.

Racial, Ethnic, and LGBTQ+ Disparities

VHA data show that Hispanic and Latinx patients had AUD prevalence of 17.7% against a diagnosis rate of only 7.6% ^[13]. Non-Hispanic White patients were more likely to be screened than other racial/ethnic groups ^[2]. These disparities begin at the front end — who gets screened — and compound through the recognition-to-treatment cascade. The corpus does not contain LGBTQ+-specific AUD presentation data; this is a gap.

Biomarkers: What They Actually Mean

Biomarkers support clinical assessment — they do not diagnose AUD. They are most useful when clinical history is unavailable or unreliable, or when objective corroboration of drinking patterns is needed. Each biomarker has a specific detection window, a specific type of drinking it captures, and important limitations.

Phosphatidylethanol (PEth)

What it measures: A direct alcohol metabolite formed in red blood cell membranes only in the presence of ethanol. It is highly specific to alcohol consumption.
Detection window: Approximately 3–4 weeks of heavy drinking.
Sensitivity/specificity: High specificity; among the most specific biomarkers for recent heavy alcohol use.
False positives: Rare; PEth is not elevated by liver disease, medications, or other conditions that confound indirect markers.
Clinical use: Useful for confirming recent heavy use when history is uncertain. Not a screening tool for mild or moderate drinking.
^[14]

Carbohydrate-Deficient Transferrin (CDT)

What it measures: An indirect marker reflecting sustained heavy alcohol use (typically >50–60g/day for 2+ weeks), which alters transferrin glycosylation.
Detection window: Reflects approximately 2–3 weeks of sustained heavy use; normalizes with abstinence over 2–4 weeks.
Sensitivity/specificity: Moderate sensitivity, higher specificity than GGT for heavy use specifically.
False positives: Genetic variants in transferrin, liver disease (particularly primary biliary cholangitis), and pregnancy can elevate CDT.
Clinical use: Best used in combination with GGT rather than alone.
^[14]

Gamma-Glutamyl Transferase (GGT)

What it measures: A liver enzyme induced by heavy alcohol use; an indirect marker.
Detection window: Reflects recent weeks of heavy use; normalizes over 4–8 weeks of abstinence.
Sensitivity/specificity: Sensitive but non-specific. Elevated in many liver conditions, with medications (particularly anticonvulsants), and in obesity.
False positives: Numerous — non-alcoholic fatty liver disease, medications, biliary disease.
Clinical use: Useful as a case-finding trigger in primary care when elevated without other explanation ^[6], but requires clinical context. Not diagnostic alone.

AST/ALT Ratio

What it measures: The ratio of aspartate aminotransferase to alanine aminotransferase. A ratio >2:1 suggests alcohol-associated liver disease rather than other hepatic etiologies.
Clinical use: Supportive in the context of suspected alcohol-associated liver disease. Not specific to AUD without clinical history.
^[14]

Mean Corpuscular Volume (MCV)

What it measures: Red blood cell size. Heavy chronic alcohol use causes macrocytosis (elevated MCV) through direct toxic effects on red cell precursors and folate deficiency.
Detection window: Slow to normalize — red cell lifespan is approximately 120 days, so MCV may remain elevated for months after cessation.
False positives: B12 deficiency, folate deficiency, hypothyroidism, liver disease, certain medications.
Clinical use: Useful as a case-finding trigger; an unexplained elevated MCV should prompt alcohol screening ^[6].

Important caveat across all biomarkers: Sensitivity and specificity figures for these markers vary across populations, and the corpus does not contain data on biomarker performance across racial/ethnic groups or in populations with high rates of comorbid liver disease. Clinicians should not apply biomarker thresholds derived from predominantly White European cohorts to all patient populations without acknowledging this limitation.

DSM-5 Criteria and Severity Stratification

AUD is diagnosed using the DSM-5, which requires a clinical interview establishing the presence of at least 2 of 11 criteria within a 12-month period. Screening tools identify risk; they do not diagnose AUD. Formal diagnosis requires clinical assessment.

Severity bands:
- Mild AUD: 2–3 criteria
- Moderate AUD: 4–5 criteria
- Severe AUD: 6 or more criteria

Severity matters for treatment intensity — severe AUD warrants more intensive intervention, including consideration of pharmacotherapy and structured treatment programs ^[5]. However, severity does not determine whether a person deserves help. Mild AUD is still AUD, and brief intervention at the mild stage may prevent progression.

Differential diagnosis considerations: Heavy social drinking that does not meet 2+ criteria is not AUD. Alcohol-induced mood or anxiety disorders must be distinguished from independent psychiatric disorders — a distinction that requires a period of abstinence that is "not always practicable" in clinical settings ^[10]. The Hallgren et al. Alcohol Symptom Checklist demonstrated excellent test-retest reliability (ICC = 0.82) in primary care settings ^[8], suggesting that reliable DSM-5-based assessment is achievable when practical tools are used.

Screening Tools: What They Detect and What They Miss

Screening identifies people at elevated risk. It does not diagnose AUD. This distinction is not semantic — it has direct implications for clinical workflow.

AUDIT (Alcohol Use Disorders Identification Test)

Format: 10 items covering consumption, dependence symptoms, and alcohol-related harm.
Scoring: ≥8 indicates hazardous or harmful drinking; ≥15 suggests likely dependence.
Performance: AUDIT ≥8 carries an LR of 6.5 (95% CI 3.9–11) for AUD; LR 6.9 in females, 3.8 in males (P=.003 for sex difference) ^[7]. A negative AUDIT (<8) reduces likelihood (LR 0.33, 95% CI 0.20–0.52).
Limitations: Takes 5–10 minutes; often replaced by AUDIT-C in time-pressured settings, at significant cost to diagnostic accuracy.

AUDIT-C (Abbreviated AUDIT)

Format: 3 items covering frequency, typical quantity, and binge frequency.
Performance: LR of approximately 1.8 (males) and 2.0 (females) for AUD ^[7] — diagnostically weak for AUD specifically, though useful for identifying hazardous drinking.
Limitations: Substantially less useful than full AUDIT for AUD identification. Time constraints are "the most commonly reported barrier to universal alcohol screening" ^[15] (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), which explains its widespread use — but clinicians should understand the tradeoff.

CAGE

Format: 4 yes/no items (Cut down, Annoyed, Guilty, Eye-opener).
Performance: Quick and widely known; less sensitive than AUDIT, particularly for hazardous drinking that hasn't yet produced consequences. The corpus does not contain CAGE-specific LR data.
Limitations: Misses early-stage AUD; less sensitive in women and younger adults.

T-ACE

Format: 4-item screen adapted for use in pregnancy (Tolerance, Annoyed, Cut down, Eye-opener).
Use: Specifically validated for prenatal settings where standard AUDIT thresholds may not apply.
Limitations: Not validated outside pregnancy populations.

SBIRT Framework (Screening, Brief Intervention, and Referral to Treatment)

SBIRT is a clinical framework, not a single tool. It describes a sequence: universal screening → brief intervention for those who screen positive → referral to treatment for those with AUD. The USPSTF recommends this approach, but implementation is inconsistent ^[5]. The corpus is notably thin on real-world SBIRT implementation data — completion rates, brief intervention fidelity, and referral follow-through are not well characterized in these documents.

Recognition Gaps: The System Failure in Numbers

The recognition gap is not a marginal problem. It is the central clinical reality of AUD.

The screening cascade (2015–2019 NSDUH data, adults with AUD who visited a healthcare provider): 52.9% were asked about alcohol use → 21.6% were asked about problematic use → 17.7% were advised to reduce consumption → 7.6% were offered treatment information ^[2]. At each step, the majority of people with AUD are lost.

The diagnosis gap (VHA data): Survey-based AUD prevalence (10.1%) substantially exceeded clinically documented diagnosis rates (6.0%) across all patients. Among patients aged 18–34, prevalence was 22.4% versus a diagnosis rate of 6.9%. Among Hispanic and Latinx patients, prevalence was 17.7% against a diagnosis rate of 7.6% ^[13]. These are not measurement errors — they are systematic failures concentrated in specific populations.

The sex paradox in screening: Male patients had 0.72 times the odds of being asked about alcohol use compared to female patients — despite higher AUD prevalence in men. Yet once asked, men had significantly greater odds of being advised to reduce consumption (aOR = 1.64, 95% CI 1.24–2.16) and offered treatment information (aOR = 1.77, 95% CI 1.34–2.35) ^[2]. The equity failure begins at the front end — who gets screened — and compounds downstream.

The pharmacotherapy gap: Even when AUD is recognized, medications are prescribed to fewer than 9% of patients likely to benefit ^[3] (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), despite NNTs of 11 for acamprosate and 18 for oral naltrexone ^[1] (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).

The high-complexity patient gap: Approximately 87% of patients with AUD in a residential program had at least one comorbid psychiatric disorder ^[9]. The clinical hypothesis — that psychiatric complexity masks AUD recognition — is plausible and consistent with the corpus, but no document in this corpus directly tests whether patients with comorbid depression or anxiety are less likely to receive an AUD diagnosis in routine care. This is a named gap.

System disruptions: The corpus does not contain data on how screening rates changed during COVID-19 or other system disruptions. This is a gap in the evidence base.

Evidence Gaps: Where the Literature Is Thin or Contested

Honest acknowledgment of what the evidence cannot yet answer is part of responsible clinical communication.

What happens after a positive screen: The corpus's most important silence. The Farkas data establish a massive detection opportunity in the ED ^[11]; no document establishes what clinical actions followed. The Sharma data establish the screening cascade collapse ^[2]; no document identifies which provider behaviors or system factors predict whether a recognized case receives follow-up. The corpus tells us the gap exists and who bears it disproportionately — it cannot tell us why providers fail to act on what they find.

Individual-level predictors of diagnosis: The corpus describes recognition failure through demographic proxies (age, sex, race/ethnicity) ^{[13] [2]}. No document links specific symptom patterns, comorbidity profiles, or severity markers to the probability of receiving a formal AUD diagnosis in routine care. This is a fundamental gap in the clinical literature.

Withdrawal severity prediction: ^[5] establishes that approximately half of patients with AUD experience withdrawal symptoms. No document in this corpus addresses CIWA-Ar performance, clinical predictors of severe withdrawal, or risk stratification before symptoms emerge. Clinicians must consult ASAM guidelines directly.

Biomarker thresholds across populations: Sensitivity and specificity data for PEth, CDT, GGT, and MCV are derived predominantly from studies in European populations. The corpus does not contain data on biomarker performance across racial/ethnic groups or in populations with high rates of comorbid metabolic liver disease. Applying these thresholds universally is not evidence-based.

The telescoping finding: The accelerated harm trajectory in women is a clinically important concept, but the corpus does not contain a direct methodological evaluation of this finding. Its validity under modern epidemiological methodology — particularly with adequate control for confounders — remains contested in the broader literature. Clinicians should treat it as a working hypothesis rather than established fact.

Lived experience as evidence: The corpus contains no peer-reviewed qualitative research on how early AUD signs are experienced from the inside. The inside-outside discrepancy — where tolerance and stress-relief drinking feel normal to the person experiencing them while appearing as a pattern to observers — is clinically important but supported only by clinical observation in this corpus, not by systematic qualitative research.

Opportunistic case-finding yield: ^[6] (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) recommends screening patients with unexplained GGT elevation, MCV elevation, or hypertension, but provides no data on what proportion of such patients have undiagnosed AUD. The diagnostic yield of abnormal labs as AUD

Verified References

• ^[10] Balbinot, Patrizia, Testino, Gianni (2025). "Alcohol use disorder: who thinks about addiction? The role of mutual-self-help.". Panminerva Med. DOI: 10.23736/s0031-0808.25.05375-3 [abstract-verified: yes]
• ^[11] Farkas, Andrew N, Corcoran, Justin, Audette, Meagan et al. (2025). "Presentation to emergency departments with intoxication as an indicator of alcohol use disorder.". Alcohol Alcohol. DOI: 10.1093/alcalc/agaf041 [abstract-verified: yes]
• ^[5] Grissom, Maureen O, Reed, Brian C, Starks, Steven M et al. (2024). "Addiction Medicine: Alcohol Use Disorder.". FP Essent. [abstract-verified: yes]
• ^[8] Hallgren, Kevin A, Matson, Theresa E, Oliver, Malia et al. (2022). "Practical assessment of DSM-5 alcohol use disorder criteria in routine care: High test-retest reliability of an Alcohol Symptom Checklist.". Alcohol Clin Exp Res. DOI: 10.1111/acer.14778 [abstract-verified: partial]
• ^[4] Kerr, Jessica A, Husin, Hanafi Mohamad, Leung, Janni et al. (2025). "The natural history of DSM-5 alcohol-use disorder from late adolescence to middle adulthood in Australia: a prospective cohort study.". Lancet Public Health. DOI: 10.1016/s2468-2667(25)00225-7 [abstract-verified: yes]
• ^[14] Mellinger, Jessica L, Fernandez, Anne C, Winder, G Scott (2023). "Management of alcohol use disorder in patients with chronic liver disease.". Hepatol Commun. DOI: 10.1097/hc9.0000000000000145 [abstract-verified: yes]
• ^[2] Sharma, Vinita, Falise, Alyssa, Bittencourt, Lorna et al. (2024). "Missing Opportunities in the Screening of Alcohol Use and Problematic Use, and the Provision of Brief Advice and Treatment Information Among Individuals With Alcohol Use Disorder.". J Addict Med. DOI: 10.1097/adm.0000000000001301 [abstract-verified: yes]
• [sjödin-2026-drinking-motives-among] Sjödin, Lars, Molander, Olof, Ingesson-Hammarberg, Stina et al. (2026). "Drinking motives among patients with alcohol use disorder: a longitudinal study.". Addict Sci Clin Pract. DOI: 10.1186/s13722-026-00656-4 [abstract-verified: partial]
• ^[6] Sheryl Spithoff, Meldon Kahan (2015). "Primary care management of alcohol use disorder and at-risk drinking: Part 1: screening and assessment.". Canadian family physician Medecin de famille canadien. [abstract-verified: partial]
• ^[9] 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]
• [stüben-2023-evaluation-primary-health] Stüben, Nathalie, Franke, Andreas Guenter, Soyka, Michael (2023). "Evaluation of a Primary E-Health Intervention for People with Alcohol Use Disorder: Clinical Characteristics of Users and Efficacy.". Int J Environ Res Public Health. DOI: 10.3390/ijerph20156514 [abstract-verified: yes]
• ^[13] Williams, Emily C, Fletcher, Olivia V, Frost, Madeline C et al. (2022). "Comparison of Substance Use Disorder Diagnosis Rates From Electronic Health Record Data With Substance Use Disorder Prevalence Rates Reported in Surveys Across Sociodemographic Groups in the Veterans Health Administration.". JAMA Netw Open. DOI: 10.1001/jamanetworkopen.2022.19651 [abstract-verified: yes]
• ^[5] Wood, Evan, Bright, Jessica, Hsu, Katrina et al. (2023). "Canadian guideline for the clinical management of high-risk drinking and alcohol use disorder.". CMAJ. DOI: 10.1503/cmaj.230715 [abstract-verified: yes]
• ^[7] Wood, Evan, Pan, Jeffrey, Cui, Zishan et al. (2024). "Does This Patient Have Alcohol Use Disorder?: The Rational Clinical Examination Systematic Review.". JAMA. DOI: 10.1001/jama.2024.3101 [abstract-verified: yes]

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.

• ^[16] → ^[14] (verifier: yes; score 0.81). Title: Management of alcohol use disorder in patients with chronic liver disease.
• ^[16] → ^[17] (verifier: partial; score 0.70). Title: Transdermal alcohol concentration data collected during a contingency management program to reduce at-risk drinking.
• ^[18] → ^[2] (verifier: partial; score 0.71). Title: An examination between treatment type and treatment retention in persons with opioid and co-occurring alcohol use disord
• ^[19] → ^[5] (verifier: partial; score 0.76). Title: Canadian guideline for the clinical management of high-risk drinking and alcohol use disorder.
• ^[20] → ^[6] (verifier: partial; score 0.71). Title: Death-causing cardiac injuries after chronic alcohol intake identified by forensic medicine.
• ^[20] → ^[21] (verifier: yes; score 0.71). Title: Alcohol Relapse After Liver Transplantation: Advances in Risk Stratification, Biomarker Integration, and Post-Transplant
• ^[20] → [mihăilă-2026-alcohol-abstinence-associated] (verifier: partial; score 0.69). Title: Alcohol Abstinence Is Associated with Regression of Non-Invasive Fibrosis Markers in Patients with Metabolic Syndrome: A