ICD-10-CM, ICD-11, and DSM-5 Coding for Alcohol Use Disorder: A Practical Reference

For clinicians, billing specialists, and program administrators navigating AUD diagnosis and documentation

A note on the evidence base for this article: The expert panel that produced this reference operated without a retrieved document corpus. All substantive claims below are drawn from verified domain knowledge across five specialties — family medicine/coding, addiction psychiatry, health information management, global health/WHO implementation, and insurance/payer operations. Where the panel identified that a specific type of study or guideline should exist but was absent from the literature review, those gaps are noted explicitly. This transparency is intentional: knowing what evidence is missing is as clinically important as knowing what exists.

Overview

Two frameworks govern AUD coding in the United States, and they do not speak the same language.

DSM-5 (Diagnostic and Statistical Manual of Mental Disorders, 5th Edition, published by the American Psychiatric Association) is the clinical diagnostic framework. It defines Alcohol Use Disorder as a single condition on a severity spectrum, assessed through 11 symptoms. Clinicians use DSM-5 to diagnose, characterize severity, and guide treatment planning.

ICD-10-CM (International Classification of Diseases, 10th Revision, Clinical Modification) is the US billing standard, maintained by the Centers for Disease Control and Prevention (CDC) and the Centers for Medicare and Medicaid Services (CMS). It uses categorical codes organized around substance, pattern of use, and complications. Payers adjudicate claims using ICD-10-CM codes — not DSM-5 language.

ICD-11 (11th Revision, published by the World Health Organization) is the emerging global standard. It restructures substance use disorders in ways that more closely align with DSM-5's dimensional approach. As of the current update, ICD-11 is not yet the US billing standard; ICD-10-CM remains in effect for US claims.

These three systems map onto each other imperfectly. That imperfection is not a minor administrative inconvenience — it is the central operational problem in AUD care today, with consequences for treatment authorization, reimbursement, population health data, and patient access to care.

DSM-5 Diagnostic Criteria

DSM-5 defines AUD as a problematic pattern of alcohol use leading to clinically significant impairment or distress, with at least 2 of 11 symptoms present within a 12-month period.

The 11 Diagnostic Criteria (Grouped by Domain)

Impaired Control (4 criteria)
1. Drinking more, or for longer, than intended
2. Persistent desire or unsuccessful efforts to cut down or control use
3. Spending a great deal of time obtaining, using, or recovering from alcohol
4. Craving — a strong desire or urge to use alcohol

Social Impairment (3 criteria)
5. Recurrent use resulting in failure to fulfill major role obligations
6. Continued use despite persistent social or interpersonal problems caused or worsened by alcohol
7. Important social, occupational, or recreational activities given up or reduced

Risky Use (2 criteria)
8. Recurrent use in physically hazardous situations
9. Continued use despite knowledge of a persistent physical or psychological problem caused or worsened by alcohol

Pharmacological Criteria (2 criteria)
10. Tolerance — needing markedly more alcohol to achieve the same effect, or markedly diminished effect with the same amount
11. Withdrawal — characteristic withdrawal syndrome, or drinking to relieve or avoid withdrawal symptoms

Severity Specifiers

Documentation directive: The DSM-5 severity specifier must appear in the clinical narrative of the medical record. It does not appear in the ICD-10-CM billing code. This is the core of the DSM-5/ICD-10-CM mismatch problem.

Remission Specifiers

DSM-5 defines two remission states, both requiring that no AUD criteria (except craving) have been met:

• Early remission: 3 to less than 12 months without criteria being met
• Sustained remission: 12 months or longer without criteria being met

These specifiers are clinically critical for ongoing care documentation, medication management decisions, and communicating patient status across providers. They must be documented in the clinical record and updated at each relevant encounter.

The DSM-IV to DSM-5 Crosswalk

DSM-IV used two separate diagnoses: alcohol abuse and alcohol dependence. DSM-5 eliminated this split and merged them into a single AUD diagnosis with severity gradation. This was a major clinical advancement — the old dichotomy created artificial boundaries and left patients with significant impairment in a diagnostic gap.

For clinicians reviewing old records: A prior diagnosis of "alcohol abuse" should not be assumed equivalent to current mild AUD without reassessment. A prior diagnosis of "alcohol dependence" warrants evaluation for moderate or severe AUD. Old records should be re-classified using current criteria for active treatment planning.

ICD-10-CM F10 Code Family

ICD-10-CM organizes alcohol-related disorders under the F10 code family. The structure follows a logic of: substance (F10) + pattern of use (1x = abuse, 2x = dependence) + complication or specifier.

Core Code Table

Coding note (ICD-10-CM 2024): The F10.1x (abuse) versus F10.2x (dependence) distinction in ICD-10-CM does not directly correspond to DSM-5 mild versus moderate/severe. The ICD-10-CM "dependence" category maps most closely to DSM-5 moderate-to-severe AUD, but the crosswalk is not validated by a formal concordance study — a critical evidence gap identified by this panel.

Severity Coding

The Core Problem

DSM-5 severity (mild/moderate/severe) is documented in the clinical record. ICD-10-CM does not have direct severity sub-codes for AUD — there is no F10.20-mild or F10.20-severe. The billing code conveys pattern of use and complications, not symptom count.

This creates a structural mismatch with real consequences:

• A patient with severe AUD (8 of 11 criteria) and a patient with moderate AUD (4 of 11 criteria) may both be billed under F10.20 (alcohol dependence, uncomplicated) if neither is currently in withdrawal or intoxication
• Payers cannot distinguish these patients from the claim code alone
• Prior authorization for intensive services requires manual clinical record review to find severity documentation that should be — but is not — inherent in the code

The panel's consensus finding: This mismatch is the central operational problem in AUD coding today. It drives administrative friction, delays care, and may systematically disadvantage patients with severe AUD who need higher-intensity services.

What Clinicians Should Do

Because ICD-10-CM does not carry severity, the clinical record must do the work:

1. Document the DSM-5 symptom count explicitly — e.g., "Patient meets 7 of 11 DSM-5 AUD criteria, consistent with severe AUD"
2. Name the severity specifier — mild, moderate, or severe — in the assessment and plan
3. Include the AUDIT or AUDIT-C score when available; this provides an objective, standardized severity anchor
4. Use the most specific ICD-10-CM code available — if withdrawal is present, use F10.23x, not F10.20

ICD-11 and Severity

ICD-11 adds severity distinctions natively, which represents a significant structural improvement. Under ICD-11, alcohol-related disorders include:

• Single episode of harmful use — a new intermediate category
• Harmful pattern of use — recurrent harmful use without dependence features
• Alcohol dependence — with severity gradation built into the classification

This closer alignment with DSM-5's dimensional model is one of ICD-11's most clinically meaningful changes. However, US billing remains ICD-10-CM, and the transition pathway — including training, EHR updates, and payer policy revision — has not been formally studied in the US context.

In Remission Specifiers

DSM-5 Remission (Clinical Record)

Both specifiers require that the patient is not in a controlled environment (e.g., incarceration, residential treatment) that would artificially suppress use. If in a controlled environment, that should be noted separately.

ICD-10-CM Remission (Billing)

ICD-10-CM provides a single remission code: F10.21 (Alcohol dependence, in remission). There is no ICD-10-CM distinction between early and sustained remission — the time gradation lives only in the clinical record.

Practical implication: When a patient transitions from active dependence to remission, update the problem list from F10.20 to F10.21. Failure to update the remission specifier is one of the most common coding errors in ongoing AUD care. An outdated F10.20 on an active problem list may trigger unnecessary prior authorization requirements or misrepresent the patient's current status to other providers.

Withdrawal Coding

Alcohol withdrawal is a medical emergency in its severe forms. Accurate coding is essential for clinical communication, billing, and risk documentation.

Withdrawal Code Selection

CIWA-Ar and Coding

The Clinical Institute Withdrawal Assessment for Alcohol, Revised (CIWA-Ar) is the standard tool for assessing withdrawal severity:

• Score 0–9: Mild withdrawal — monitor; pharmacotherapy may not be required
• Score 10–18: Moderate withdrawal — symptom-triggered or fixed-schedule benzodiazepine protocol
• Score >18: Severe withdrawal — high risk for seizure and delirium; medical management required

CIWA-Ar scores are documented in the clinical record but do not directly map to ICD-10-CM sub-codes. However, documenting the CIWA-Ar score supports the specificity of the withdrawal code selected (e.g., a CIWA-Ar of 22 with documented hallucinations supports F10.232 over F10.230).

Documentation best practice: Record the CIWA-Ar score, the date and time of assessment, and the specific symptoms present. This documentation supports both clinical decision-making and the specificity of the ICD-10-CM code.

ICD-11 — What Changes

ICD-11 reorganizes alcohol-related disorders under "Disorders due to substance use" and introduces a more nuanced classification structure:

Key Structural Changes in ICD-11

1. Dependence criteria simplified — ICD-11 reduces dependence criteria from six (ICD-10) to three core features: impaired control, increasing priority of alcohol use, and physiological features (tolerance/withdrawal)
2. Severity gradation added — severity is built into the classification, not left to clinical narrative alone
3. Closer DSM-5 alignment — the dimensional approach reduces the conceptual gap between clinical diagnosis and administrative coding

Agreement Between Systems

Expert panel discussion referenced that studies show "almost perfect agreement" between ICD-10 and ICD-11 for AUD dependence, but only "substantial" agreement between ICD-11 and DSM-5. This suggests ICD-11 improves alignment with DSM-5 but does not eliminate the gap entirely. Clinicians should seek primary literature on ICD-10/ICD-11/DSM-5 concordance studies before relying on these figures for policy decisions.)*

US Transition Status

As of the current ICD-10-CM 2024 update, ICD-11 is not the US billing standard. Some international jurisdictions are actively transitioning. US providers should monitor CMS announcements for transition timelines. No formal US implementation study was available to this panel — a significant evidence gap for program administrators planning ahead.

Common Coding Errors

Insurance Implications

Mental Health Parity

AUD diagnoses coded under F10.xx trigger Mental Health Parity and Addiction Equity Act (MHPAEA) protections. This means:

• Insurers cannot impose more restrictive prior authorization requirements for AUD treatment than for analogous medical/surgical conditions
• Quantitative limits (visit limits, day limits) and non-quantitative limits (medical necessity criteria) must be comparable
• Accurate AUD coding is the prerequisite for invoking parity protections — an unspecified or undercoded diagnosis may not trigger the same protections

Pharmacotherapy Coverage

Medications approved for AUD treatment — including naltrexone (oral and extended-release injectable), acamprosate, and disulfiram — typically require an AUD diagnosis code for coverage. Prior authorization for extended-release naltrexone (Vivitrol) in particular often requires:

• A documented AUD diagnosis (F10.1x or F10.2x)
• Evidence of clinical severity supporting the medication
• Documentation of prior treatment attempts or clinical rationale

The panel identified a critical evidence gap here: no publicly available study documents the specific ICD-10-CM code and severity documentation requirements that major payers use when reviewing prior authorization for AUD pharmacotherapy. Payer policies vary and are often proprietary. Providers should request payer-specific medical necessity criteria in writing before submitting prior authorization requests.

Prior Authorization and Severity

Because ICD-10-CM does not carry DSM-5 severity, prior authorization reviewers must find severity documentation in the clinical record. The practical implication:

• A prior authorization request for residential AUD treatment supported only by F10.20 (uncomplicated) and no clinical narrative will likely be denied or delayed
• The same request supported by F10.20 + a clinical note documenting "severe AUD, 8 of 11 DSM-5 criteria, AUDIT score 28, history of withdrawal seizure" is substantially more defensible

Documentation Best Practices

At Every AUD Encounter

• [ ] Document DSM-5 symptom count and severity specifier in the assessment narrative
• [ ] Record AUDIT or AUDIT-C score when screening is performed
• [ ] Use the most specific ICD-10-CM code available on the problem list and claim
• [ ] Update remission specifier if patient's status has changed
• [ ] Document any withdrawal history, including dates and CIWA-Ar scores
• [ ] Add relevant Z-codes for social determinants

For Withdrawal Episodes

• [ ] Record CIWA-Ar score at assessment
• [ ] Document specific symptoms present (tremor, diaphoresis, hallucinations, altered consciousness)
• [ ] Select the most specific F10.23x sub-code supported by documentation
• [ ] Note date of last drink and estimated time since last use

For Remission Documentation

• [ ] Record onset date of remission
• [ ] Specify early (3–12 months) or sustained (12+ months) in clinical narrative
• [ ] Note craving status (craving alone does not disqualify remission)
• [ ] Note if patient is in a controlled environment

EHR Implementation Tip

The panel identified a practical, no-new-evidence-required improvement: structured AUDIT or AUDIT-C screening templates that automatically populate clinical notes with symptom-count language directly support ICD-10-CM code selection. This is an implementation gap, not a research gap — the tools exist; the workflow bridge between screening and coding documentation is what most practices are missing.

Z-Codes for Social Determinants

Z-codes complement F10 codes and support comprehensive documentation of factors affecting AUD care. They do not replace F10 codes but add clinical context that supports medical necessity and care coordination.

Critical Evidence Gaps — What This Field Still Needs

The expert panel was unanimous on the following gaps. These are not theoretical concerns — they affect coding decisions made in every AUD encounter today:

1. No validated DSM-5-to-ICD-10-CM crosswalk tool tested in primary care workflows. Clinicians are making real-time coding decisions without evidence-based guidance on how to translate DSM-5 severity into ICD-10-CM code selection.

2. No large-scale concordance study documenting how often DSM-5 severity assessments in clinical records match the ICD-10-CM codes submitted on claims — and in which direction errors occur (undercoding is the suspected pattern, but this is unverified).

3. No publicly available payer medical necessity criteria mapping specific ICD-10-CM codes to covered levels of AUD care. Payer policies are proprietary, creating inconsistency across the healthcare landscape.

4. No US-based ICD-11 implementation study examining the transition's impact on coding accuracy, reimbursement rates, or clinical outcomes for AUD.

5. No pharmacotherapy prior authorization study documenting what severity documentation (DSM-5 symptom count, AUDIT score, specific ICD-10-CM code) is required by major payers to approve naltrexone, acamprosate, or disulfiram.

These gaps mean that much of current AUD coding practice rests on professional consensus rather than verified evidence. Clinicians, billing specialists, and program administrators should advocate for this research — and document their own coding decisions with enough clinical detail to withstand scrutiny in the absence of formal guidance.

Quick Reference Summary

This reference reflects the state of AUD coding knowledge as of the ICD-10-CM 2024 update. Coding guidelines are updated annually; verify current codes against the official ICD-10-CM tabular list before use in billing. For ICD-11 transition updates, monitor WHO and CMS announcements.