AUDIT Scoring and Interpretation: A Clinician's Guide

The AUDIT is a 10-item alcohol screen scored 0 to 40: items 1 to 8 score 0 to 4, items 9 and 10 score 0, 2 or 4. The WHO bands are 0-7 (alcohol education), 8-15 (simple advice), 16-19 (brief counselling) and 20+ (specialist referral). The 3-item AUDIT-C scores 0 to 12 (positive at 4+ for men, 3+ for women).

Those numbers come from the WHO manual, but they are not the only ones in circulation. Australia's national treatment guidelines use a lower threshold for dependence assessment, and give more than one AUDIT-C cut-off. Below: scoring, the item-level rules most score sheets omit, and how to document a result.

How do you score the AUDIT?

The WHO's AUDIT manual (Babor et al., second edition, 2001) sets out the scoring. The ten items cover three domains:

• Consumption (items 1-3): how often the person drinks, how many drinks on a typical drinking day, and how often they have six or more drinks on one occasion.
• Dependence (items 4-6): being unable to stop once started, failing to do what was expected because of drinking, and needing a morning drink after a heavy session.
• Harm (items 7-10): guilt or remorse, blackouts, injury to self or others, and whether someone has been concerned or suggested cutting down.

Items 1 to 8 each have five response options scored 0, 1, 2, 3 and 4. Items 9 and 10 have only three options, scored 0, 2 and 4: "No", "Yes, but not in the last year" and "Yes, during the last year". Add the item scores for a total out of 40.

The most common hand-scoring error is scoring items 9 and 10 as 0, 1 and 2 because they sit next to five-point items. That shaves up to four points off the total, which is enough to move a client down a whole zone.

The manual also describes two skip rules for the interview version: if the person has not drunk in the past year, or scores 0 on both items 2 and 3, the interviewer can jump to items 9 and 10. The manual recommends using these skips only in interview or computerised formats, not on a paper self-report form.

What do AUDIT scores mean?

The WHO manual groups totals into four risk zones, each paired with a level of intervention:

• Zone I (0-7): alcohol education.
• Zone II (8-15): simple advice focused on reducing hazardous drinking.
• Zone III (16-19): simple advice plus brief counselling and continued monitoring.
• Zone IV (20-40): referral to a specialist for diagnostic evaluation and treatment.

A total of 8 or more is the manual's recommended indicator of hazardous and harmful use, as well as possible dependence. It adds that a cut-off of 10 gives greater specificity at the expense of sensitivity, and that dropping the cut-off to 7 for women and for men over 65 increases sensitivity in those groups.

The manual is candid that these bands are tentative. It calls them tentative and subject to clinical judgement, and notes that the cut-off may vary with a country's drinking patterns and the alcohol content of its standard drink. That caveat matters in Australia.

Do Australian guidelines use different AUDIT cut-offs?

Yes. The Guidelines for the Treatment of Alcohol Problems (Haber and Riordan, 4th edition, 2021), prepared for the Australian Government Department of Health, recommend the AUDIT as the most effective screening tool for primary care and hospital populations (Recommendation 4.8, Grade A). An overview of the recommendations was published in the Medical Journal of Australia.

Where they depart from the WHO bands:

• Dependence threshold. Chapter 4 states that 8 or more indicates presumptive hazardous or harmful drinking, and 15 or more suggests the need to assess for dependence. The WHO manual puts that point at 20.
• Older people. Chapter 18 notes that lowering the AUDIT cut-off to 5 or more, and the AUDIT-C to 4 or more, has been recommended to improve sensitivity in older people.
• Mental health settings. Chapter 21 recommends the AUDIT, in full or as the AUDIT-C, to help identify alcohol use disorder in people attending mental health services as part of a comprehensive assessment (Recommendation 21.9, Grade A).

None of these is wrong. They are different thresholds chosen for different purposes. The practical point is that "an AUDIT of 16" carries a different next step depending on which framework you apply, so your note should say which one you used.

How do you score the AUDIT-C?

The AUDIT-C is simply items 1 to 3 of the AUDIT, scored 0 to 4 each, for a total of 0 to 12. It was validated by Bush and colleagues (1998) in US Veterans Affairs clinics, and later tested in a family practice sample by Bradley and colleagues (2007). Bradley found the thresholds that best balanced sensitivity and specificity for alcohol misuse were 4 or more in men (sensitivity 0.86, specificity 0.89) and 3 or more in women (sensitivity 0.73, specificity 0.91).

The Australian guidelines reproduce those sex-specific cut-offs in their appendix scoring guide. But the chapter 4 text gives a different figure: a score of 5 or more indicates that further assessment is required. Add the 4-or-more threshold for older people and there are three AUDIT-C cut-offs in the one national document.

A sensible approach is to pick one threshold for your practice, write it into your assessment protocol, and apply it consistently. If a client screens positive on the AUDIT-C, administering the full AUDIT gives you the dependence and harm items that the short form cannot.

What do individual AUDIT items tell you?

The total score is only half the information. The WHO manual gives three item-level rules that are easy to miss:

1. A score of 1 or more on item 2 or item 3 indicates consumption at a hazardous level.
2. Any points on items 4 to 6, especially weekly or daily responses, imply the presence or early development of dependence.
3. Points on items 7 to 10 indicate that alcohol-related harm is already occurring.

The manual also suggests stepping up one level of intervention when a client scores 2 or more on items 4, 5 or 6, or 4 on items 9 or 10, regardless of the total. A total of 9 that includes a monthly morning drink is not the same as 9 made up of consumption items alone.

Items 9 and 10 carry a second message. A response of "Yes, but not in the last year" flags a past problem even when current drinking is low. The manual advises using those responses to discuss the need for vigilance, even in the absence of current hazardous drinking.

Why do standard drinks matter when scoring?

The manual states that items 2 and 3 assume a standard drink contains 10 grams of alcohol, and item 3's "six or more drinks" is intended to capture about 60 grams. Australia's standard drink is also 10 grams, according to the NHMRC, so the questionnaire does not need adjusting here the way it does in countries with larger standard drinks.

The catch is that clients count glasses, not standard drinks. The Department of Health's standard drinks guide lists an average restaurant serve of red wine (150 ml at 13.5%) as 1.6 standard drinks, and a 375 ml can of full-strength beer (4.8%) as 1.4. A client who reports "three glasses of wine" may be having close to five standard drinks, which moves item 2 from 1 point to 2.

Take a worked example. A client drinks two or three times a week, usually four restaurant-sized glasses of wine, and says they never have "six drinks". Scored on their own count, the AUDIT-C is 3 + 1 + 0 = 4. Converted to standard drinks (about 6.4 per occasion), item 2 rises to 2 points and item 3 to 3 (weekly), giving 3 + 2 + 3 = 8. Both totals screen positive for men under Bradley's threshold, but 8 out of 12 tells a very different clinical story from 4.

It is also worth comparing answers against the NHMRC guideline itself: healthy adults should drink no more than 10 standard drinks a week and no more than 4 on any one day. The AUDIT's six-drink item sits above that daily limit, so a client who regularly has five standard drinks can exceed the guideline while scoring 0 on item 3. The Australian treatment guidelines recommend a quantity and frequency history as the way to detect drinking above the NHMRC levels (Recommendation 4.7), which is a good reason to record actual drinks alongside the score.

How should you document an AUDIT result?

A useful entry records:

• Version and format: full AUDIT or AUDIT-C, self-report or interview, and whether skip rules were used.
• Total and item scores, including which domain the points came from.
• The cut-off and framework applied, for example "WHO zone II" or "positive on AUDIT-C, 4 or more (male threshold)".
• Standard drink conversion if you clarified glass sizes.
• Clinical context: mood, sleep, medications and any risk factors, since the treatment guidelines note that intoxication and early abstinence can both raise suicide risk. See our guide on documenting a suicide risk assessment.
• The plan: advice given, brief intervention, referral or review date.

The AUDIT asks about the past year, so it is better suited to screening and periodic review than to week-to-week outcome tracking. A repeat after a month of treatment will still include drinking from before treatment began. For session-by-session change, pair it with a measure built for shorter windows, such as the K10 or DASS-21 for distress, and a drinking diary for consumption. Screening results also belong in your intake assessment summary.

PractaLuma is AI-native practice management software for Australian mental-health practices. Both the AUDIT and AUDIT-C are in its standardised assessments library, so clients can complete them as self-report before a session and the scores sit in the client file with the rest of the record. See pricing for plan details.

FAQ

Is the AUDIT a diagnostic tool?

No. The AUDIT is a screening instrument. A high score indicates the need for further assessment, and a diagnosis of alcohol use disorder requires a clinical interview against diagnostic criteria.

What is the difference between the AUDIT and AUDIT-C?

The AUDIT-C uses only the three consumption items and scores 0 to 12. The full AUDIT adds dependence and harm items and scores 0 to 40, which gives more information for treatment planning.

What AUDIT score indicates alcohol dependence?

No single score confirms dependence. The WHO manual says scores of 20 or above clearly warrant further diagnostic evaluation, while Australia's treatment guidelines suggest assessing for dependence from 15 or more.

Where can I read the WHO's own scoring notes?

The WHO manual is free to download, and the AUDIT developers keep a summary at auditscreen.org.

This article is general information for clinicians and is not a substitute for clinical judgement or the full published guidelines.