How accurate is the AUDIT-C at detecting unhealthy drinking?
Answered November 2, 202611 sources
EvidenceBModerateGood evidence in people with one real weakness. New studies could change the answer.
The AUDIT-C picks up most unhealthy drinking in primary care, but it misses some cases, especially in women. It works about as well as the full 10-question test.
Who it's for
Adults screened for risky drinking in doctors' offices; less certain for severe problems.
What to expect
Good at flagging risky drinking; some people are missed or wrongly flagged.
Watch out for
People under-report drinking, so a low score can miss real problems.
AUDIT-C caught more men than women with misuse
Bars show how often the test was right at its best cutoff, where 1 would be perfect.
Figure 2. Prevalence of three categories of alcohol consumption as indicated with PEth values (µmol/L) and AUDIT (points), in relation to blood pressure control groups, AUDIT (P = 0.865),…
Figure 3. Prevalence of hazardous alcohol usePEth and hazardous alcohol useAUDIT-C in blood pressure control groups and in the entire study population, including men and women separately.
A review that pooled 14 studies, mostly in primary care patients, found no clear difference between the AUDIT-C and the full AUDIT for risky drinking or alcohol use disorders. The full test may be better for severe problems such as dependence, but this was not proven. Studies used different standards and gave mixed results1Meta-analysis · 2008 · 438 peopleMeta-analysis: are 3 questions enough to detect unhealthy alcohol use?Available evidence is inconclusive but suggests that the full AUDIT may be superior to the AUDIT-C for identifying unhealthy alcohol use in adults in some settings.Ann Intern Med.
A study of 1,319 patients at a family practice clinic found the AUDIT-C worked well in both sexes. It was somewhat better in men than in women. The best cutoff was a score of 4 or more in men (86% of misusers caught, 89% of non-misusers correctly cleared) and 3 or more in women (73% caught, 91% cleared)6Study · 2007AUDIT-C as a brief screen for alcohol misuse in primary care.The AUDIT-C was an effective screening test for alcohol misuse in this primary care sample. Optimal screening thresholds for alcohol misuse among men (> or =4) and women (> or =3) were the same as in previously published VA studies.Alcohol Clin Exp Res.
A study of male veterans (243 completed the questionnaire and interview) found the AUDIT-C was slightly better than the full AUDIT for heavy drinking and slightly worse for alcohol abuse or dependence. The two were about equal for either problem5Study · 1998 · 447 peopleThe AUDIT alcohol consumption questions (AUDIT-C): an effective brief screening test for problem drinking. Ambulatory Care Quality Improvement Project (ACQUIP). Alcohol Use Disorders Identification Test.Three questions about alcohol consumption (AUDIT-C) appear to be a practical, valid primary care screening test for heavy drinking and/or active alcohol abuse or dependence.Arch Intern Med.
In 7,141 pregnancies in Finland, the AUDIT-C separated hazardous drinking (defined by the full AUDIT) very well, with a best cutoff of 4 or more. That test compared the short form with the long one, not with a clinical interview9Observational study · 2026Predictive value of alcohol use questionnaires for adverse pregnancy outcomes: Evidence from a Finnish birth cohort.Abbreviated versions of the AUDIT, particularly AUDIT-4, appear comparable to the full AUDIT for identifying risky alcohol use and PTB risk. Women reporting recurrent heavy episodic drinking may benefit from targeted counseling before and during pregnancy.Acta Obstet Gynecol Scand.
Self-reports can be wrong. In a study of 2,581 trauma patients who scored zero on the AUDIT, about 5% had alcohol in their blood10Observational study · 2021Underreporting of alcohol use in trauma patients: A retrospective analysis.Inaccurate self-reporting of alcohol drinking behavior does exist in trauma patients. A composite of objective alcohol screening modalities, in addition to AUDIT, is needed to screen for alcohol use in this population. Healthcare providers should remain highly suspicious of alcohol-related injuries in individuals with…Subst Abus. A review of 13 studies in people treated for alcohol use disorder found under-reporting in many of them, though these were not AUDIT-C studies2Systematic review · 2021 · 73 peopleConsistency between self-reported alcohol consumption and biological markers among patients with alcohol use disorder - A systematic review.The findings indicate that inconsistency between self-reported alcohol consumption and biomarkers may occur in a considerable proportion of patients with AUD. However, further studies applying more sensitive, specific, and easily assessable biological markers are warranted to confirm this preliminary synthesis.…Neurosci Biobehav Rev.
Cutoffs vary. Among older adults, one study suggested 4 or more3Systematic review · 2021A systematic review of self-report measures used in epidemiological studies to assess alcohol consumption among older adults.No consensus was found regarding methods used to assess, define, and measure alcohol consumption in older adults. Identical assessments and definitions must be developed to make valid comparisons of alcohol consumption in older adults. We recommend that alcohol surveys for older adults define the following drinking…PLoS One. A study of people with high blood pressure used 5 or more for men and 4 or more for women7Observational study · 2025Hypertension and alcohol: a cross-sectional study comparing PEth with AUDIT and AUDIT-C in primary care.PEth revealed significantly higher alcohol use in the aTRH group, undetected by AUDIT. AUDIT-C and PEth identified more hazardous use than AUDIT, suggesting their potential to prompt alcohol-related discussions and support evidence-based hypertension care. PEth correlated more strongly with AUDIT-C than with AUDIT.Fam Pract.
Key studies
Study
Who
What it found
Review of 14 studies, 20081Meta-analysis · 2008 · 438 peopleMeta-analysis: are 3 questions enough to detect unhealthy alcohol use?Available evidence is inconclusive but suggests that the full AUDIT may be superior to the AUDIT-C for identifying unhealthy alcohol use in adults in some settings.Ann Intern Med
Mostly primary care adults
AUDIT-C about as accurate as full AUDIT; mixed results
Study, 20076Study · 2007AUDIT-C as a brief screen for alcohol misuse in primary care.The AUDIT-C was an effective screening test for alcohol misuse in this primary care sample. Optimal screening thresholds for alcohol misuse among men (> or =4) and women (> or =3) were the same as in previously published VA studies.Alcohol Clin Exp Res
1,319 clinic patients
Better in men than women; cutoffs 4 and 3
Study, 19985Study · 1998 · 447 peopleThe AUDIT alcohol consumption questions (AUDIT-C): an effective brief screening test for problem drinking. Ambulatory Care Quality Improvement Project (ACQUIP). Alcohol Use Disorders Identification Test.Three questions about alcohol consumption (AUDIT-C) appear to be a practical, valid primary care screening test for heavy drinking and/or active alcohol abuse or dependence.Arch Intern Med
243 male veterans
Matched full AUDIT for heavy drinking or abuse
Study of births, 20269Observational study · 2026Predictive value of alcohol use questionnaires for adverse pregnancy outcomes: Evidence from a Finnish birth cohort.Abbreviated versions of the AUDIT, particularly AUDIT-4, appear comparable to the full AUDIT for identifying risky alcohol use and PTB risk. Women reporting recurrent heavy episodic drinking may benefit from targeted counseling before and during pregnancy.Acta Obstet Gynecol Scand
7,141 pregnancies
Excellent match with full AUDIT; cutoff 4 or more
What we don't know
The best cutoff for each group (women, older adults, pregnant women) is not settled.
How well the AUDIT-C finds severe alcohol dependence is unclear.
Written by AI from the studies listed, without review by a clinician, and graded with the GRADE method. It can be wrong, so check the studies it cites. Education, not medical advice.