Authors' objectives

To describe the performance of screening questionnaires for heavy drinking and/or alcohol abuse or dependence in general clinical populations of women in the United States.

Searching

MEDLINE was searched from 1966 to July 1997 using the following search terms: 'alcoholism' or 'alcohol-drinking' and 'CAGE', 'BMAST', 'T-ACE', 'TACE', 'TWEAK', 'MAST', 'SMAST', 'SAAST' or 'AUDIT'. A second MEDLINE search was conducted from 1996 to July 1997 to identify more recent screening articles; this used the terms 'alcoholism' or 'alcohol-drinking' and 'screening'. In addition, the Social Sciences Citation Index and the Science Citation Index were searched for material on screening questionnaires without acronyms and more recently developed questionnaires. The bibliographies of the retrieved papers were examined. The searches were restricted to publications in the English language.

Study selection

Assessment of study quality

The included studies were assessed in light of the following methodological limitations:

the sampling strategy was not described or only patients with positive screening results were interviewed;

the reported sensitivity was based on fewer than 20 women meeting the diagnostic criteria for alcohol abuse or dependence, or the sensitivity was not reported;

the items were used in more than one questionnaire, sometimes with changes in the timeframe or wording;

multiple alcohol screening questionnaires were administered at one time;

screening questionnaires and comparison standards were administered by the same interviewer or at one sitting;

criterion standards were not interviewer administered;

the racial and/or ethnic makeup of the study population was not explicitly stated, or the data for black and white patients were combined.

Two reviewers independently coded each study for the presence or absence of the methodological limitations.

Data extraction

Data on the sensitivity and specificity of the questionnaires, and the areas under the ROC curves, were extracted by one reviewer and checked by a second.

Methods of synthesis

Results of the review

Nine studies evaluating eight brief screening questionnaires were included. The overall number of women screened was approximately 10,476; of these, at least 723 met the diagnostic criteria for alcohol abuse or dependence, according to the reference test.

The CAGE questionnaire had areas under the ROC curves of 0.84 to 0.92 for alcohol abuse and dependence in predominantly black populations of women, but when using the traditional cut point of 2 or more drinks per day, low sensitivities (38 to 50%) resulted in predominantly white female populations. The TWEAK and AUDIT questionnaires had areas under the ROC curves of 0.87 to 0.93 for past-year alcohol abuse or dependence in black or white women, but had sensitivities of less than 80% at traditional cut points. For detecting heavy drinking, the AUDIT questionnaire had areas under ROC curves of at least 0.87 in female primary care patients. The TWEAK and T-ACE questionnaires had higher areas under ROC curves (0.84 to 0.87) than the CAGE questionnaire (0.76 to 0.78) for the detection of heavy drinking before pregnancy was recognised in black obstetric patients.

Authors' conclusions

The CAGE questionnaire was relatively insensitive in predominantly white female populations. The TWEAK and AUDIT questionnaires performed adequately in black or white women when using lower cut points than usual.

CRD commentary

The research question, search strategy, inclusion criteria and validity assessment were adequately explained. Some details of the studies were tabulated. The search strategy was fairly limited and it is possible that additional material may have been identified had other databases and languages been accessed. Also, there was no mention of an attempt to identify unpublished papers. A narrative synthesis would seem appropriate given the authors' reservations concerning between-study heterogeneity; some additional exploration of the possible sources of heterogeneity would have been helpful. The authors' conclusions appear to follow on from the results of the review.

Implications of the review for practice and research

Research: The authors state that additional research is needed to assess the performance of alcohol screening questionnaires in diverse clinical populations of women; a gender-specific modification of the AUDIT questionnaire should be evaluated; the performance of the AUDIT, CAGE, and TWEAK questionnaires should be further evaluated in women, using lower cut points than have been recommended for men; and future studies should strive to have adequate numbers of women, in order to stratify analyses by race, ethnicity and age.

Bibliographic details

Bradley K A, Boyd-Wickizer J, Powell S H, Burman M L. Alcohol screening questionnaires in women: a critical review. JAMA 1998; 280(2): 166-171. [PubMed: 9669791]

Original Paper URL

http://jama.ama-assn.org/

Other publications of related interest

This additional published commentary may also be of interest. Saitz R. Review: some alcohol screening tests have acceptable test properties for use in general clinical populations of U.S. women. ACP J Club 1999;130:17.

Indexing Status

Subject indexing assigned by NLM

MeSH

Adult; African Americans; Alcoholism /ethnology /prevention & control; European Continental Ancestry Group; Female; Humans; Male; Mass Screening; Questionnaires; ROC Curve; Sensitivity and Specificity; United States /epidemiology

AccessionNumber

11998008736

Database entry date

30/11/2003

Record Status

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.