Alcohol Use Disorder: Diagnosis, Severity, and the Gap Between Identification and Care

A comprehensive clinical and research reference synthesizing current evidence on DSM-5 criteria, screening instruments, severity stratification, and diagnostic equity.

Overview — Why Diagnostic Precision Matters

Alcohol use disorder (AUD) is among the most prevalent and undertreated conditions in medicine. Approximately 10.9% of U.S. adults meet criteria for AUD, yet in a cohort of 114,511 individuals who screened positive for unhealthy alcohol use, only 10.1% had a formal AUD diagnosis documented in their electronic health record (EHR) ^[1]. That gap — between the person who drinks in ways that harm their health and the person who receives a diagnosis and a treatment plan — is not a minor measurement artifact. It is a systemic failure with documented consequences.

Why does a formal diagnosis matter? Because it is a gateway. In the same large cohort, receiving an AUD diagnosis increased the adjusted odds of being prescribed medication (disulfiram, acamprosate, or naltrexone) by more than tenfold (adjusted odds ratio [aOR] = 10.68; 95% CI: 9.68–11.79) and the odds of receiving psychotherapy by over 50% (aOR = 1.57; 95% CI: 1.46–1.69) ^[1]. The act of diagnosis is itself a clinical intervention.

Understanding what that diagnosis means — how it is made, what criteria it rests on, how it is graded by severity, and how it differs from a positive screening result — is therefore not an academic exercise. It is the foundation of equitable, effective care.

The DSM-IV to DSM-5 Transition

Before 2013, the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) divided alcohol-related pathology into two separate categories: alcohol abuse (a pattern of harmful use without physical dependence) and alcohol dependence (physiological and behavioral dependence). These were treated as distinct diagnoses, not points on a continuum.

The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) collapsed this binary into a single dimensional disorder — AUD — assessed across 11 criteria and graded by severity ^{[2] [3]}. One criterion was added that had no DSM-IV equivalent: craving. One criterion was removed: recurrent legal problems related to alcohol. The remaining criteria were reorganized from the former abuse and dependence clusters.

This restructuring has direct implications for prevalence research. A patient who met DSM-IV criteria for alcohol abuse (but not dependence) may now meet criteria for mild AUD under DSM-5 — or may not, depending on which criteria they endorse. Conversely, some DSM-IV dependence cases map cleanly to DSM-5 severe AUD, while others do not. When citing prevalence figures from studies conducted before 2013, readers should note that the diagnostic framework changed; DSM-IV and DSM-5 estimates are not directly comparable ^[4].

The 11 DSM-5 Criteria

The criteria are grouped here by conceptual domain, though the DSM-5 does not formally weight them — a point that has itself been criticized (see Evidence Gaps, below).

For each criterion, the source-language framing is presented first, followed by a plain-English translation and a brief clinical illustration.

Domain 1: Impaired Control

Criterion 1 — Larger/Longer Use Than Intended
"Alcohol is often taken in larger amounts or over a longer period than was intended."
The person sets out to have two drinks and regularly has six, or plans to drink only on weekends and finds the pattern expanding. This criterion captures the loss of the internal "stop" signal. In a prospective cohort using SCID-IV-RV assessment, this was among the more commonly endorsed criteria (9.0%) in a young adult sample ^[4].

Criterion 2 — Persistent Desire or Unsuccessful Efforts to Cut Down
"There is a persistent desire or unsuccessful efforts to cut back or control alcohol use."
The person has tried to stop or reduce — perhaps many times — and has not been able to sustain the change. This is distinct from simply not wanting to stop; it requires that the person has made attempts that failed.

Criterion 3 — Great Deal of Time Spent
"A great deal of time is spent in activities necessary to obtain alcohol, use alcohol, or recover from its effects."
When obtaining, using, and recovering from alcohol begins to consume a significant portion of the person's day, this criterion is met. In severe presentations, this can mean that most waking hours are organized around alcohol.

Criterion 4 — Craving (New in DSM-5; not present in DSM-IV)
"Craving, or a strong desire or urge to use alcohol."
This criterion was added to DSM-5 based on neuroscientific evidence that craving reflects a distinct neurobiological process — incentive salience — separate from physical dependence. It has also been the subject of criticism: craving is subjective, difficult to operationalize reliably, and may overlap with normal desire in social drinkers. The addition of craving as a criterion has been questioned in the literature, and its contribution to diagnostic validity remains an open question (see Evidence Gaps).

Domain 2: Social Impairment

Criterion 5 — Failure to Fulfill Major Role Obligations
"Recurrent alcohol use resulting in a failure to fulfill major role obligations at work, school, or home."
Missing work due to hangovers, neglecting childcare responsibilities, or failing coursework because of drinking patterns. The impairment must be recurrent, not a single incident.

Criterion 6 — Continued Use Despite Social or Interpersonal Problems
"Continued alcohol use despite having persistent or recurrent social or interpersonal problems caused or exacerbated by the effects of alcohol."
The person continues drinking even after arguments with a partner about their drinking, estrangement from family, or loss of friendships — and recognizes the connection between alcohol and these problems. In the SCID-IV-RV cohort, social problems were the second most commonly endorsed criterion at 10.4% ^[4].

Criterion 7 — Important Activities Given Up or Reduced
"Important social, occupational, or recreational activities are given up or reduced because of alcohol use."
Hobbies abandoned, social events avoided, career opportunities declined — because drinking has become the organizing priority.

Domain 3: Risky Use

Criterion 8 — Use in Physically Hazardous Situations
"Recurrent alcohol use in situations in which it is physically hazardous."
Driving while impaired, operating machinery, or drinking in contexts where impairment creates direct physical risk.

Criterion 9 — Continued Use Despite Knowledge of Physical or Psychological Problems
"Alcohol use is continued despite knowledge of having a persistent or recurrent physical or psychological problem that is likely to have been caused or exacerbated by alcohol."
The person has been told — or knows — that their liver disease, depression, or gastritis is worsened by alcohol, and continues drinking. This criterion requires awareness, not just the presence of a medical problem.

Domain 4: Pharmacological Criteria

Criterion 10 — Tolerance
"Tolerance, as defined by either (a) a need for markedly increased amounts of alcohol to achieve intoxication or desired effect, or (b) a markedly diminished effect with continued use of the same amount of alcohol."
The person who once felt intoxicated after three drinks now requires eight to achieve the same effect — or notices that their usual amount no longer produces the effect it once did. Tolerance requires careful interpretation in younger populations. In a prospective cohort using SCID-IV-RV assessment, tolerance was by far the most commonly endorsed criterion at 50.3% — far exceeding the next most common criterion (social problems at 10.4%). When tolerance was removed from the criterion set, DSM-5 AUD lifetime prevalence dropped from 18.4% to 11.0%, a 40% reduction ^[4]. This raises important questions about whether tolerance, as currently operationalized, is specific enough to distinguish pathological from physiological adaptation in heavy social drinkers.

Criterion 11 — Withdrawal
"Withdrawal, as manifested by either (a) the characteristic withdrawal syndrome for alcohol, or (b) alcohol (or a closely related substance, such as a benzodiazepine) is taken to relieve or avoid withdrawal symptoms."
Withdrawal symptoms include tremor, sweating, anxiety, nausea, and — in severe cases — seizures or delirium tremens. A person who drinks in the morning to stop shaking meets this criterion even if they have never experienced a formal withdrawal syndrome. Withdrawal is clinically the highest-risk criterion. Within mild-to-moderate AUD (2–5 criteria), the presence of withdrawal was associated with an adjusted hazard ratio of 11.62 (95% CI: 7.54–17.92) for progression to severe AUD, compared to aHR 5.64 (95% CI: 3.28–9.70) for mild-to-moderate AUD without high-risk criteria ^[2]. Two patients can share the same DSM-5 severity label but have radically different clinical trajectories depending on whether withdrawal is present.

A Note on Criterion Removal

The DSM-IV criterion of recurrent legal problems related to alcohol (e.g., DUI arrests) was removed in DSM-5. This decision was partly driven by evidence that legal consequences are heavily influenced by socioeconomic and racial factors — a person with resources may avoid arrest for the same behavior that results in criminal charges for someone without them. Removing this criterion reduces, though does not eliminate, socially patterned diagnostic bias.

Severity Stratification

The DSM-5 grades AUD severity by criterion count:

These cutoffs are clinically useful as shorthand, but the evidence base for the specific thresholds is limited. The 2–3/4–5/6+ boundaries were established by expert consensus rather than empirical validation against treatment outcomes. As ^[2] demonstrated in a combined cohort of 15,928 individuals, simple criterion counts obscure meaningful heterogeneity within severity bands. A patient with mild AUD (2–3 criteria) who endorses withdrawal has a dramatically worse prognosis than a patient with mild AUD who endorses only tolerance and craving — yet both receive the same severity label.

This has direct clinical implications. The current DSM-5 count-based framework, as ^[2] explicitly states, "may be improved by emphasizing specific high-risk criteria." Clinicians using severity to guide treatment intensity should assess which criteria are present, not merely how many.

An important caveat on the severity gradient data: The diagnosis rates of 6.8%, 21.5%, and 41.6% cited above come from ^[1], which operationalized severity using AUDIT-C risk tiers — not DSM-5 criterion counts. These are consumption-based categories, not symptom-based categories. The two systems are not interchangeable, and the apparent severity gradient may partly reflect instrument discordance rather than a true clinical severity gradient (see DSM-5 vs. DSM-IV vs. ICD-11 and Differential Diagnosis sections below).

Remission Specifiers

The DSM-5 includes remission specifiers that clinicians should document in the medical record, as they affect insurance coverage, disability determinations, and treatment planning:

• Early remission: No criteria for AUD have been met (except craving) for at least 3 months but less than 12 months.
• Sustained remission: No criteria for AUD have been met (except craving) for 12 months or longer.
• In a controlled environment: The individual is in an environment where access to alcohol is restricted (e.g., incarceration, residential treatment). This specifier acknowledges that abstinence in a controlled setting does not carry the same prognostic weight as community-based remission.
• On maintenance therapy: The individual is taking a prescribed medication (e.g., naltrexone, acamprosate) as part of treatment.

Natural history data from a prospective cohort study (cumulative incidence of AUD symptoms from late adolescence to age 42: 58.0%, 95% CI: 52.3–63.8%) found that most individuals do remit — 67.0% by age 42 — but 11–13% show persistent symptoms from late adolescence onward ^[5]. AUD symptoms peaked at age 24 in this cohort ^[5]. These natural history data matter for calibrating treatment intensity: for many patients, particularly younger adults, the trajectory is toward remission even without intensive intervention — though the 11–13% with persistent symptoms represent a high-need subgroup requiring sustained engagement.

DSM-5 vs. DSM-IV vs. ICD-11

The same patient can receive different diagnoses depending on which framework is applied. This is not merely a theoretical concern — it directly affects cross-study comparisons and prevalence estimates.

DSM-IV used two separate categories: alcohol abuse (harmful use without dependence) and alcohol dependence (physiological and behavioral dependence). A patient meeting only abuse criteria received a less severe diagnosis with different treatment implications.

DSM-5 collapsed these into a single disorder with 11 criteria. The addition of craving and the removal of legal problems means that some DSM-IV abuse cases become DSM-5 mild AUD, while some DSM-IV dependence cases may not map cleanly to DSM-5 severe AUD. Prevalence estimates from DSM-IV-era studies should not be directly compared to DSM-5 estimates without adjustment ^[4].

ICD-11 (the International Classification of Diseases, 11th Revision, used internationally and increasingly in U.S. billing) retains a distinction between harmful use (a pattern causing health damage) and alcohol dependence (a cluster of physiological, behavioral, and cognitive phenomena). The ICD-11 dependence syndrome maps most closely to DSM-5 moderate-to-severe AUD, while ICD-11 harmful use overlaps with DSM-5 mild AUD — but the overlap is imperfect [corpus-gap].

A prospective cohort study using SCID-IV-RV assessment found that DSM-5 AUD lifetime prevalence was 18.4% in their sample, while ICD-11 harmful use/dependence prevalence differed — with the specific discrepancy driven substantially by the tolerance criterion, which inflated DSM-5 estimates ^[4]. Researchers citing prevalence across frameworks should specify which diagnostic system was used and note that the same individual may be classified differently under each.

Screening Tools — AUDIT

Important distinction: Screening tools identify individuals who may have AUD or hazardous drinking patterns. They do not diagnose AUD. A positive screen should prompt formal diagnostic assessment using DSM-5 criteria — it does not replace it.

The Alcohol Use Disorders Identification Test (AUDIT) is a 10-item questionnaire developed by the World Health Organization (WHO) and validated across primary care, emergency department, and community settings. Items cover consumption (frequency, quantity, binge frequency), dependence symptoms (impaired control, morning drinking, guilt), and alcohol-related problems (blackouts, injury, concern from others). Each item is scored 0–4, yielding a total of 0–40.

Scoring thresholds (general adult primary care populations):
- ≥8: Hazardous or harmful drinking; warrants brief intervention
- ≥15: Likely alcohol dependence in most populations; warrants referral for further assessment

A systematic review of 35 studies (N = 79,633) found that AUDIT ≥8 yields a likelihood ratio (LR) of 6.5 (95% CI: 3.9–11) for DSM-5 AUD ^[6]. Notably, AUDIT performs better for identifying AUD in females (LR 6.9) than males (LR 3.8), a statistically significant difference (p = .003) ^[6]. This sex difference in diagnostic performance is clinically important: the same cutoff does not carry the same diagnostic weight across sexes.

Population discipline note: These likelihood ratios were derived from studies predominantly in primary care and general adult populations. Performance may differ in emergency departments, specialty addiction settings, or populations with high baseline AUD prevalence.

Screening Tools — AUDIT-C

The AUDIT-C is a three-item abbreviated version of the AUDIT, retaining only the consumption questions (frequency of drinking, typical quantity, and frequency of heavy episodic drinking). It is widely embedded in EHR systems — including the Veterans Health Administration (VHA) — because of its brevity and ease of administration.

Scoring thresholds (general adult populations):
- ≥3 (women) or ≥4 (men): Hazardous drinking; warrants further assessment
- ≥9: Possible alcohol dependence in some health system validation studies

In a head-to-head comparison against a structured diagnostic reference standard, the AUDIT-C demonstrated an area under the curve (AUC) of 0.90 for identifying both problem use and AUD in both males and females, with sensitivity of 0.83–0.89 and specificity of 0.78–0.84 depending on sex and outcome ^[7]. These are strong screening performance figures.

However, the AUDIT-C is considerably less useful than the full AUDIT for identifying AUD specifically. The likelihood ratio for AUD at standard cutoffs is only 1.8 for males and 2.0 for females ^[6] — a modest diagnostic signal compared to the full AUDIT's LR of 6.5. The AUDIT-C was designed to detect excessive drinking, not DSM-5 AUD. These are different constructs, and conflating them is a documented source of the apparent "diagnosis gap" in the literature [corpus-gap].

The practical implication: a positive AUDIT-C is a flag that should trigger further assessment, not a diagnosis.

Screening Tools — CAGE, T-ACE, TWEAK, and Others

CAGE is a four-item questionnaire asking about: attempts to Cut down, Annoyance at criticism of drinking, Guilt about drinking, and Eye-opener (morning drinking). It is rapid and widely recognized, but less sensitive than the AUDIT for identifying hazardous drinking without dependence. CAGE was developed before the DSM-5 and does not map onto the 11-criterion framework. It performs best for identifying established dependence and is less useful for detecting mild-to-moderate AUD or hazardous drinking in the absence of dependence symptoms.

T-ACE and TWEAK were specifically validated for use in pregnancy, a population in which any alcohol use carries fetal risk and standard AUDIT cutoffs may not apply. T-ACE asks about Tolerance, Annoyance, Cut down attempts, and Eye-opener. TWEAK asks about Worry, Eye-opener, Amnesia, Kut down attempts, and Pass out. Both tools have been validated for detecting risky drinking in pregnant women, where the clinical threshold for intervention is lower than in the general population.

MAST (Michigan Alcoholism Screening Test) and its shortened versions (SMAST, b-MAST) have been used in outpatient settings and with older adults. The MAST's 25 items cover lifetime consequences of drinking and may be more sensitive for detecting long-standing AUD in older patients, though its length limits routine use.

When to use which tool: The AUDIT remains the best-validated general-purpose screening instrument for primary care adults ^[6]. The AUDIT-C is appropriate for high-volume EHR-based screening where brevity is essential, with the understanding that positive results require follow-up. T-ACE or TWEAK should be used in pregnancy. CAGE is a reasonable rapid screen in settings where AUDIT is unavailable, but clinicians should recognize its limitations for mild AUD. MAST variants may be appropriate for older adults or forensic settings.

SBIRT — Screening, Brief Intervention, Referral to Treatment

SBIRT (Screening, Brief Intervention, and Referral to Treatment) is a public health framework — not a single tool — that links alcohol screening to a graduated clinical response. The three components are:

1. Screening: Using a validated instrument (AUDIT, AUDIT-C, or equivalent) to identify individuals with hazardous drinking or possible AUD.
2. Brief Intervention (BI): For individuals with hazardous or harmful drinking who do not meet AUD criteria, a structured 5–15 minute motivational conversation aimed at reducing consumption. BI is not a treatment for AUD; it is an intervention for risky drinking.
3. Referral to Treatment (RT): For individuals who screen positive and are subsequently diagnosed with AUD, referral to appropriate treatment — which may include medications for AUD (MAUD), behavioral therapies, or specialty addiction care.

The U.S. Preventive Services Task Force (USPSTF) recommends screening for unhealthy alcohol use in primary care settings in adults 18 and older, including pregnant women, and providing brief counseling interventions to those who screen positive ^[8]. However, screening is "not always performed consistently or correctly in primary care" ^[8], and the pathway from positive screen to formal diagnosis to treatment referral breaks down at multiple points.

A critical implementation gap: the SBIRT framework assumes that a positive screen will trigger diagnostic assessment and, if AUD is confirmed, treatment referral. In practice, many clinical systems stop at the screening step — recording an AUDIT-C score in the EHR without completing the diagnostic evaluation needed to generate an ICD code or initiate treatment [corpus-gap]. The result is a documented positive screen with no clinical follow-through.

Screening in Special Populations

Pregnancy: Any alcohol use during pregnancy carries risk of fetal alcohol spectrum disorders. Standard AUDIT cutoffs are not appropriate; T-ACE and TWEAK are the validated tools for this population. The clinical threshold for intervention is lower than in non-pregnant adults.

Adolescents: The CRAFFT (Car, Relax, Alone, Forget, Friends, Trouble) is the most widely validated screening tool for alcohol and drug use in adolescents aged 12–21. Standard adult AUDIT cutoffs should not be applied to this population.

Older adults: Alcohol pharmacokinetics change with age — decreased lean body mass, reduced hepatic metabolism, and increased sensitivity to central nervous system effects mean that older adults experience greater impairment at lower blood alcohol concentrations. Standard AUDIT cutoffs may underestimate risk in this population; modified lower thresholds (e.g., ≥5 rather than ≥8) have been proposed for adults over 65, though the evidence base for specific cutoffs in older adults is less robust than for general adult populations.

College students: High rates of heavy episodic drinking in this population mean that AUDIT-C may flag a large proportion of students without AUD. The full AUDIT or structured DSM-5 assessment is needed to distinguish hazardous drinking from disorder.

Military and veterans: The VHA has implemented AUDIT-C as a universal annual screen for all enrolled patients. VHA data show that survey-based AUD prevalence (10.1%) substantially exceeds EHR-documented diagnosis rates (6.0%), with the largest gaps among patients aged 18–34 (prevalence 22.4% vs. diagnosis rate 6.9%) and Hispanic/Latinx patients (17.7% vs. 7.6%) ^[2]. Even in a system with mandatory universal screening, the screening-to-diagnosis gap persists.

Racial and ethnic considerations: Screening tool performance has been validated primarily in majority-White populations. The available evidence documents that diagnostic gaps fall along racial and ethnic lines [corpus-gap], but the corpus does not provide criterion-level data on whether specific DSM-5 criteria perform differently across racial/ethnic groups. This is a documented gap in the evidence base.

Differential Diagnosis — Heavy Drinking vs. AUD

The diagnostic threshold for AUD is 2 of 11 criteria within a 12-month period. Below that threshold, a person may drink heavily — even at levels that carry significant health risk — without meeting criteria for the disorder.

The National Institute on Alcohol Abuse and Alcoholism (NIAAA) defines low-risk drinking as no more than 4 drinks on any single day and no more than 14 drinks per week for men; no more than 3 drinks on any single day and no more than 7 drinks per week for women. Drinking above these limits is considered hazardous or at-risk drinking — a target for brief intervention — but does not constitute AUD unless at least 2 DSM-5 criteria are met ^[3].

This distinction matters clinically. A person who regularly drinks 5–6 drinks per evening but has not experienced loss of control, craving, withdrawal, or functional impairment may be at significant health risk (liver disease, cardiovascular disease, cancer) without meeting AUD criteria. They need intervention — but a different kind than someone with moderate or severe AUD. Conflating hazardous drinking with AUD leads to both over-treatment (applying addiction-focused interventions to someone who needs a brief consumption-reduction conversation) and under-treatment (failing to recognize that some heavy drinkers do meet AUD criteria and need more than brief advice).

A positive AUDIT-C score identifies hazardous drinking. It does not diagnose AUD. The formal diagnosis requires assessing all 11 DSM-5 criteria — ideally using a validated structured instrument such as the Alcohol Symptom Checklist, which demonstrated excellent test-retest reliability in routine primary care (intraclass correlation coefficient [ICC] = 0.82, 95% CI: 0.77–0.85) and can be completed by patients between visits or in waiting rooms ^[3].

Diagnostic Equity and Recognition Gaps

The underdiagnosis of AUD is not evenly distributed. Multiple data sources document that diagnostic gaps fall systematically along lines of gender, race, ethnicity, age, and socioeconomic status.

In the All of Us cohort, lower odds of AUD diagnosis were documented among females, racial/ethnic minorities, individuals in economically deprived areas, and privately insured patients — despite comparable or higher screening positivity in some of these groups ^[1]. In the VHA, the largest prevalence-to-diagnosis gaps were among patients aged 18–34 (survey prevalence 22.4% vs. EHR diagnosis rate 6.9%) and Hispanic/Latinx patients (17.7% vs. 7.6%) ^[2].

These disparities are not explained by differences in actual AUD prevalence. They reflect differential recognition, differential documentation, and differential follow-through from positive screen to formal diagnosis. The consequence is that the patients most likely to be underdiagnosed are also the patients least likely to receive medication or psychotherapy — compounding inequity at every step of the care pathway ^[1].

The EHR-embedded Alcohol Symptom Checklist — which allows patients to self-report all 11 DSM-5 criteria — represents one structural intervention that may reduce clinician-level recognition bias by standardizing the assessment process ^[3]. Whether this tool actually narrows equity gaps by race, ethnicity, and sex has not been documented in the available corpus.

A note on the COVID-19 period: The corpus references a collapse in systematic screening during the COVID-19 pandemic, though the available evidence does not provide detailed quantitative data on the magnitude or recovery of screening rates post-pandemic. This is a gap in the current evidence base.

Evidence Gaps and Open Questions

The panel's most important methodological contribution is identifying that the "diagnosis gap" literature conflates two fundamentally different severity constructs: AUDIT-C consumption-based risk tiers and DSM-5 symptom-count criteria. These are not interchangeable. Direct concordance statistics between AUDIT-C tiers and DSM-5 criterion counts are not available in the current evidence base, and the mapping between these two systems remains incompletely characterized.

Several specific open questions deserve attention:

1. Craving as a criterion. The addition of craving to DSM-5 was theoretically motivated but has been questioned on grounds of specificity and operationalization reliability. The corpus does not provide criterion-level reliability data for craving specifically, and the clinical literature has not resolved whether craving adds diagnostic validity beyond the other 10 criteria.

2. Mild AUD validity. A patient who endorses only tolerance and craving — the two criteria most susceptible to over-endorsement in heavy social drinkers — meets criteria for mild AUD. Whether this represents a clinically meaningful disorder or a false positive generated by the criterion structure is an unresolved question. The finding that removing tolerance from the criterion set reduces DSM-5 AUD lifetime prevalence by 40% ^[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) suggests that mild AUD prevalence estimates are highly sensitive to how tolerance is operationalized.

3. Does severity classification improve outcomes? The corpus documents that receiving any AUD diagnosis dramatically increases treatment odds ^[1]. Whether finer-grained severity classification — mild vs. moderate vs. severe, or criterion-specific profiling — leads to differential treatment selection and better outcomes has not been established in the available evidence. The current corpus does not link DSM-5 severity specifiers to treatment assignment or remission rates, leaving the outcome utility of diagnostic precision an important unanswered clinical question.

4. Cross-cultural validity. The DSM-5 criteria were developed primarily from research conducted in Western, high-income populations. The corpus does not address whether the 11 criteria perform equivalently across cultural contexts where alcohol use patterns, social consequences, and help-seeking norms differ substantially.

5. Continuous risk measures as an alternative. The WHO's Risk Drinking Levels framework — which categorizes drinking by grams of pure alcohol per day into low, medium, high, and very high risk — offers a continuous alternative to categorical diagnosis for some purposes, particularly population-level monitoring and brief intervention targeting. Whether continuous risk measures should supplement or partially replace categorical AUD diagnosis for specific clinical purposes is not resolved by the current evidence base.

Summary for Clinical Practice

The evidence supports several clear clinical priorities:

1. Make the diagnosis. The single most impactful action a clinician can take is to move from a positive AUDIT-C screen to a formal DSM-5 assessment and, where criteria are met, document the diagnosis. The tenfold increase in medication treatment odds associated with a formal diagnosis ^[1] makes this a high-yield intervention.

2. Distinguish screening from diagnosis. AUDIT-C identifies; SCID-5-CV (or equivalent structured interview) diagnoses. A positive screen is the beginning of the diagnostic process, not the end.

3. Assess which criteria are present, not just how many. Withdrawal in a patient with mild AUD (2–3 criteria) carries an adjusted hazard ratio of 11.62 for progression to severe AUD ^[2]. Criterion identity matters as much as criterion count.

4. Apply population-appropriate tools. AUDIT for general adults; AUDIT-C for EHR-based universal screening with follow-up; T-ACE or TWEAK in pregnancy; CRAFFT in adolescents; modified thresholds in older adults.

5. Document remission specifiers. Early vs. sustained remission, controlled environment, and maintenance therapy specifiers affect insurance coverage, disability determinations, and treatment planning.

6. Attend to equity. Younger patients, women, racial/ethnic minorities, and economically disadvantaged patients are systematically underdiagnosed despite comparable or higher prevalence [y

Verified References

• ^[7] Adam, Angéline, Laska, Eugene, Schwartz, Robert P et al. (2025). "Identifying Alcohol Use Disorder and Problem Use in Adult Primary Care Patients: Comparison of the Tobacco, Alcohol, Prescription Medication and Other Substance (TAPS) Tool With the Alcohol Use Disorders Identification Test Consumption Items (AUDIT-C).". Subst Use Addctn J. DOI: 10.1177/29767342251326678 [abstract-verified: partial]
• ^[8] Grissom, Maureen O, Reed, Brian C, Starks, Steven M et al. (2024). "Addiction Medicine: Alcohol Use Disorder.". FP Essent. [abstract-verified: partial]
• ^[3] 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]
• ^[5] 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]
• ^[2] Miller, Alex P, Kuo, Sally I-Chun, Johnson, Emma C et al. (2023). "Diagnostic Criteria for Identifying Individuals at High Risk of Progression From Mild or Moderate to Severe Alcohol Use Disorder.". JAMA Netw Open. DOI: 10.1001/jamanetworkopen.2023.37192 [abstract-verified: partial]
• ^[4] Slade, Tim, Mewton, Louise, O'Dean, Siobhan et al. (2021). "DSM-5 and ICD-11 alcohol use disorder criteria in young adult regular drinkers: Lifetime prevalence and age of onset.". Drug Alcohol Depend. DOI: 10.1016/j.drugalcdep.2021.109184 [abstract-verified: partial]
• ^[3] 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]
• ^[2] 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]
• ^[6] 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]
• ^[1] Yue, Yihua, Rothberg, Michael B, Back, Sudie E et al. (2026). "Rates of Diagnosis and Treatment for Alcohol Use Disorder Among All of Us Participants with Unhealthy Alcohol Use.". J Gen Intern Med. DOI: 10.1007/s11606-025-10089-5 [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.

• ^[9] → ^[3] (verifier: partial; score 0.85). Title: _The impact of DSM classification changes on the prevalence of alcohol use disorder and 'diagnostic orphans' in Lebanese _
• ^[9] → ^[7] (verifier: partial; score 0.74). Title: Identifying Alcohol Use Disorder and Problem Use in Adult Primary Care Patients: Comparison of the Tobacco, Alcohol, Pre
• ^[10] → ^[6] (verifier: partial; score 0.97). Title: Cognitive Impairments in Early-Detoxified Alcohol-Dependent Inpatients and Their Associations with Socio-Demographic, Cl
• ^[2] → ^[1] (verifier: partial; score 0.82). Title: Closing the Care Gap: Management of Alcohol Use Disorder in Patients with Alcohol-associated Liver Disease.
• ^[2] → ^[11] (verifier: partial; score 0.82). Title: Equivalence of Alcohol Use Disorder Symptom Assessments in Routine Clinical Care When Completed Remotely via Online Pati
• ^[2] → ^[12] (verifier: partial; score 0.62). Title: Racial/ethnic disparities in alcohol-related problems: differences by gender and level of heavy drinking.
• ^[2] → ^[13] (verifier: partial; score 0.77). Title: Agile implementation of alcohol screening in primary care.
• ^[2] → ^[14] (verifier: partial; score 0.78). Title: _Practical assessment of DSM-5 alcohol use disorder criteria in routine care: High test-retest reliability of an Alcohol _
• ^[2] → ^[15] (verifier: yes; score 0.67). Title: Comparison of Substance Use Disorder Diagnosis Rates From Electronic Health Record Data With Substance Use Disorder Prev
• ^[16] → ^[2] (verifier: yes; score 0.69). Title: Rates of Diagnosis and Treatment for Alcohol Use Disorder Among All of Us Participants with Unhealthy Alcohol Use.
• ^[16] → NO REPLACEMENT FOUND (considered 3 candidates; none verified)
• ^[16] → ^[2] (verifier: partial; score 0.56). Title: Rates of Diagnosis and Treatment for Alcohol Use Disorder Among All of Us Participants with Unhealthy Alcohol Use.
• ^[9] → ^[17] (verifier: partial; score 0.82). Title: Practice facilitation to promote evidence-based screening and management of unhealthy alcohol use in primary care: a pra
• ^[18] → ^[4] (verifier: yes; score 0.74). Title: Differences in the Prevalence and Profile of DSM-IV and DSM-5 Alcohol Use Disorders-Results from the Singapore Mental He
• ^[18] → NO REPLACEMENT FOUND (considered 5 candidates; none verified)
• ^[14] → ^[3] (verifier: partial; score 0.85). Title: _The impact of DSM classification changes on the prevalence of alcohol use disorder and 'diagnostic orphans' in Lebanese _
• ^[14] → ^[19] (verifier: partial; score 0.77). Title: Poor subjective sleep predicts compromised quality of life but not cognitive impairment in abstinent individuals with Al
• ^[14] → ^[20] (verifier: partial; score 0.64). Title: _Implementing Technology-Supported Care for Depression and Alcohol Use Disorder in Primary Care in Colombia: Preliminary _
• ^[10] → ^[21] (verifier: partial; score 0.68). Title: Alcohol use during pregnancy: the impact of social determinants of health on alcohol consumption among pregnant women.
• ^[22] → ^[8] (verifier: partial; score 0.84). Title: Psychosocial treatment options for adolescents and young adults with alcohol use disorder: systematic review and meta-an
• ^[22] → ^[23] (verifier: partial; score 0.71). Title: Demographic differences in the cascade of care for unhealthy alcohol use: A cross-sectional analysis of data from the 20
• ^[15] → ^[2] (verifier: partial; score 0.82). Title: Rates of Diagnosis and Treatment for Alcohol Use Disorder Among All of Us Participants with Unhealthy Alcohol Use.