Current issues of ACP Journal Club are published in Annals of Internal Medicine


Review: Some alcohol screening tests have acceptable test properties for use in general clinical populations of U.S. women

ACP J Club. 1999 Jan-Feb;130:17. doi:10.7326/ACPJC-1999-130-1-017

Source Citation

Bradley KA, Boyd-Wickizer J, Powell SH, Burman ML. Alcohol screening questionnaires in women. A critical review. JAMA. 1998 Jul 8;280:166-71.



Are alcohol screening questionnaires sensitive and specific for identifying heavy drinking or alcohol abuse or dependence in women in general clinical populations in the United States?

Data sources

English-language studies were identified using MEDLINE (to 1997) with a combination of the search terms alcoholism, alcohol-drinking, screening, CAGE, BMAST, T-ACE, TACE, TWEAK, MAST, SMAST, SAAST, and AUDIT; Social Science and Science Citation Indexes; and bibliographies of relevant papers.

Study selection

Studies were selected if they compared a brief alcohol screening questionnaire (≤ 10 items) with an appropriate criterion standard for alcohol abuse or dependence (i.e., based on Diagnostic and Statistical Manual of Mental Disorders or International Classification of Diseases) or heavy drinking (i.e., based on the Alcohol Use Disorder and Associated Disabilities Interview Schedule or a timeline follow-back interview) in a generalizable clinical population of U.S. women.

Data extraction

Sensitivity, specificity, and area under receiver-operating characteristic curve.

Main results

13 of 759 articles met the inclusion criteria. They evaluated 8 screening questionnaires in various settings (emergency department, primary care, obstetric ward) and different populations (black, white, Hispanic) and assessed test properties using various cut-points. Data were not meta-analyzed because of study heterogeneity. Test properties for 7 tests are listed in the Table. Test sensitivity varied by race or ethnicity and tended to be lower for women than for men at equivalent cut-points.


Some brief alcohol screening questionnaires have acceptable test properties for women, although test sensitivities varied by race or ethnicity and tended to be lower for women than for men. Using lower-than-usual thresholds for a positive screening result may therefore be appropriate when administering alcohol screening questionnaires to women.

Sources of funding: Health Services Research and Development Field Program and Medicine Service; Veterans Affairs Puget Sound Health Care System.

For correspondence: Dr. K.A. Bradley, Health Services Research and Development, Veterans Affairs Puget Sound Health Care System (Seattle Division), 1660 South Columbian Way, Mail Stop 152, Seattle, WA 98108, USA. FAX 206-764-2935.

Table. Test characteristics of alcohol screening questionnaires for women

Test with cut-points Sensitivity Specificity Area under ROC curve*
CAGE ≥ 2 (7 studies) 38% to 83% 92% to 96%
CAGE ≥ 1 (4 studies) 59% to 89% 81% to 86% 0.76 to 0.92
AUDIT ≥ 8 (3 studies) 59% to 66% 93% to 97%
AUDIT ≥ 7 (1 study) 70% 95% 0.87
TWEAK ≥ 3 (4 studies) 67% to 80% 90% to 94%
TWEAK ≥ 2 (2 studies) 87% to 91% 77% to 87% 0.90 (1 study)
BMAST ≥ 6 (2 studies) 23% to 53% 97% to 99%
BMAST ≥ 4 (1 study) 57% 90% 0.75
Trauma scale ≥ 2 (3 studies) 40% to 53% 80% to 93% 0.59 (1 study)
T-ACE ≥ 1 (2 studies) 76% to 91% 70% to 79% 0.84 to 0.89 (1 study)
NET ≥ 1 (1 study) 71% 86% 0.79

*ROC = receiver-operating characteristic.


Dozens of studies confirm the efficacy of brief interventions for alcohol problems (1). Yet translating this evidence into practice has been difficult because of limited physician time and expertise and underrecognition of alcoholism (2). Although the CAGE questions are almost 25 years old, screening tools have only recently been tested in women, minorities, and the elderly and used to detect hazardous drinkers (persons who drink detrimental amounts with no consequences) (3, 4).

The systematic review by Bradley and colleagues finds that alcohol screening questionnaires may be less sensitive for problem drinking in women than in men. The authors also point out that their conclusions should be viewed as tentative, because unfortunately, there were few studies to review.

Alcohol problems are more likely to be detected by standardized screening tools (2). For sex-specific screening, the evidence supports using TWEAK (with questions about Tolerance, Worried, Eye openers, Amnesia, and Kut down; 2 points as a positive test result) in women: It is brief, works in diverse populations, and detects current hazardous drinking (particularly important in women who are pregnant or considering pregnancy). If one instrument is to be used for all patients, we need a brief, sensitive test that is valid in diverse populations and identifies both current and past hazardous and problem drinking.

Richard Saitz, MD, MPH
Boston University School of MedicineBoston Medical CenterBoston, Massachusetts, USA


1. Bien TH, Miller WR, Tonigan JS. Brief interventions for alcohol problems: a review. Addiction. 1993;88:31535.

2. Rydon P, Redman S, Sanson-Fisher RW, Reid AL. Detection of alcohol-related problems in general practice. J Stud Alcohol. 1992;53:197-202.

3. O'Connor PG, Schottenfeld RS. Patients with alcohol problems. N Engl J Med. 1998;338:592-602.

4. Kitchens JM. Does this patient have an alcohol problem? JAMA. 1994;272:1782-7.