The CAGE is a four-item, yes/no screen for problem drinking across the patient's lifetime. Its name is an acronym for the four questions: felt the need to Cut down, been Annoyed by criticism of drinking, felt Guilty about drinking, and needed an Eye-opener (a drink first thing in the morning). Each "yes" scores one point, giving a range of 0-4. It takes well under a minute and can be woven into conversation rather than administered as a form, which is much of its enduring appeal.
The CAGE screens for alcohol abuse and dependence - the more severe end of the drinking spectrum. It is not a measure of consumption, and it does not ask about quantity or frequency at all. Because its items are lifetime-framed ("Have you ever..."), a positive screen may reflect a past problem that has since resolved; a positive result is the start of a conversation, not a conclusion.
02 - Origin & purpose
Where it comes from.
The four questions were developed by John Ewing around 1968 at the University of North Carolina, first appeared in a validation study by Mayfield and colleagues in 1974, and were formally described by Ewing in JAMA in 1984. The intent was a screen brief and unthreatening enough for routine medical settings, where longer instruments such as the 25-item MAST were rarely used.
The CAGE became one of the most widely taught screening tools in medicine precisely because it can be delivered from memory. In recent years, guidelines have increasingly favoured the AUDIT and AUDIT-C, which capture current hazardous drinking that the CAGE misses - but the CAGE remains common in psychiatric intake and hospital settings where dependence is the primary concern.
03 - Scoring & cutoffs
How scoring works.
Each item scores 1 for "yes" and 0 for "no"; total range 0-4. The conventional cutoff is ≥2, which is considered a clinically significant positive screen warranting assessment for an alcohol use disorder. Some settings use ≥1 to raise sensitivity at the cost of more false positives - a reasonable trade in populations where dependence is common. The score itself carries no severity gradient in the way the AUDIT's does; a 4 is not "twice as ill" as a 2.
Score
Severity
Interpretation
0–1
Low likelihood
Below the cutoff.
2–4
Positive screen
Two or more is a positive screen and is clinically significant. Assess for an alcohol use disorder.
04 - Validation evidence
How well it performs.
A review of reliability and validity studies by Dhalla and Kopec (2007) concluded the CAGE is a valid screen for alcohol abuse and dependence in medical and surgical inpatients, ambulatory medical patients and psychiatric inpatients, with average sensitivity of about 0.71 and specificity of about 0.90 at the ≥2 cutoff. A diagnostic meta-analysis in general clinical populations (Aertgeerts et al., 2004) found broadly similar pooled performance for abuse and dependence, but markedly weaker results when the target was at-risk or hazardous drinking, where reported sensitivities have ranged widely across studies. Performance is consistently weaker in white women, prenatal women and college students. Internal consistency is adequate, though somewhat lower in non-clinical samples.
~0.71
SENSITIVITY (≥2)
Average at cutoff ≥2 for abuse or dependence across clinical populations (Dhalla & Kopec 2007).
~0.90
SPECIFICITY (≥2)
Average at cutoff ≥2 in the same review.
0-4
SCORE RANGE
One point per "yes"; positive screen at ≥2.
<1 min
ADMINISTRATION
Four spoken questions, no form required.
05 - How it compares
How it compares to the alternatives.
Instrument
Items
Time
When to reach for it
CAGE
4
<1 min
Lifetime yes/no screen for alcohol abuse and dependence. Fast, verbal, no severity gradient.
Drug-use counterpart; pairs with an alcohol screen for full substance coverage.
MAST
25
~8 min
The older, longer lifetime instrument the CAGE was designed to replace at the bedside; now rarely used in routine screening.
06 - When to use it
Right tool, wrong tool.
The CAGE earns its place where speed and conversational delivery matter more than measurement precision, and where dependence rather than consumption is the clinical question.
Reach for it when
-Psychiatric intake or hospital admission where alcohol dependence is the concern.
-Opportunistic verbal screening in a consultation where a form would disrupt rapport.
-Settings where administration time must be near-zero.
-Patients for whom a written questionnaire is impractical.
Reach for something else when
-Screening for hazardous or binge drinking in a broadly healthy population - use the AUDIT or AUDIT-C.
-Measuring drinking severity or monitoring change over time; it has no severity gradient, and lifetime framing cannot register improvement.
-Younger and female populations, where sensitivity is documented to be weaker.
-Distinguishing current from past problems - the lifetime frame cannot.
07 - Confidence & precision
Reading the score with care.
The CAGE is a binary lifetime screen, so SEM and minimal-clinically-important-difference statistics do not apply - there is no continuous severity dimension to measure change on. Precision is better framed through predictive values: with specificity around 0.90, false positives are relatively uncommon in clinical populations, but with sensitivity around 0.71 roughly three in ten patients with an alcohol use disorder will screen negative. A negative CAGE therefore cannot rule out a drinking problem, particularly hazardous drinking, which it was never designed to detect. Never re-administer it to track progress; scores cannot fall, because the questions are lifetime-framed.
08 - Limitations
What it cannot tell you.
Lifetime framing conflates past and current problems, and the screen is insensitive to hazardous and binge drinking that has not yet produced guilt, criticism or morning drinking. Performance is documented to be weaker in women, prenatal populations, older adults and students.
The items are face-valid and easily faked where patients are motivated to conceal drinking. There is no severity measure and no sensitivity to change. Validated cutoff behaviour also varies considerably across settings, with reported sensitivities at ≥2 ranging from roughly 14% to 84% for at-risk drinking.
See how Aisel removes friction where it costs most. A 20-minute walkthrough tailored to your clinic.
We value your privacy
We use cookies to analyse site usage and improve your experience. Analytics and embedded media (e.g. YouTube) only load if you accept. Read our cookie policy.