The value of the CAGE in screening for alcohol abuse and alcohol dependence in general clinical populations: a diagnostic meta-analysis

https://doi.org/10.1016/S0895-4356(03)00254-3Get rights and content

Abstract

Objective

To perform a meta-analysis to assess diagnostic characteristics of the CAGE in screening for alcohol abuse or dependence in a general clinical population and to test a new method for pooling of ROC curves.

Methods

Medline search performed over the period 1/1/1974 to 31/12/2001.

Measurement

Calculation of diagnostic values.

Results

We identified 35 articles using the DSM criteria as the gold standard to test the diagnostic value of the CAGE. Only 10 studies could be included for the meta-analysis. With a cutoff point ⩾2, the pooled sensitivity is far better in inpatients (0.87) than in primary care patients (0.71) or ambulatory patients (0.60). The pooled specificity also differs for each group. The likelihood ratios seem to be relatively constant over the populations (overall LR+:3.44;LR−:0.18). We calculated a pooled AUC of 0.87 (95% CI 0.85–0.89). At low specificity values, the sensitivity was homogenous over the studies, and at a low sensitivity, the specificity was heterogenous.

Conclusion

The diagnostic value of the CAGE is of limited value using this test for screening purposes at his recommended cutpoint of ⩾2.

Introduction

Alcohol abuse and dependence affect 6 to 15% of a primary care population [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11] and even a larger percentage, about 15 to 61% [12], of a specialist clinic or hospitalized patient population. However, the problem for clinicians has been one of recognition. Only one-third of patients with alcohol problems is detected by their physician [8], [13], [14]. No single laboratory test or combination of tests has been shown to be appropriate for screening [14], [15], [16]. Although physicians do not easily use them during a routine consultation, questionnaires have better test results in screening for alcohol abuse and dependence [17].

The CAGE questionnaire is one of the most widely studied self-report instruments for alcohol problems [18], [19]. It has demonstrated its superiority over laboratory markers [20], [21], [22]. Sensitivity and specificity of the CAGE are thought to be somewhat lower than other questionnaires (MAST, SAAST, and AUDIT). However, used as a first step before deciding to do a more complex workup, his brevity (four questions) and simplicity (answer yes or no) still makes it a useful screening and case-finding tool, particularly for the busy primary care physician [19].

Studies show that implementation of the CAGE may improve the identification of alcohol problems in different settings [22], [23], [24], [25], [26]; however, application of the CAGE varies widely among physicians (from 38 to 100%) [27]. Data of Jorge and Masur in Brazil demonstrated how an attempt to improve detection of alcohol-dependent patients through the introduction of the CAGE was unsuccessful. Because clinical staff did not agree with the wish to reinforce the search for alcohol problems, this results in a lack of motivation in several physicians [28]. On the other hand, Lawner et al. conducted a successful educational intervention in a primary care practice to improve physicians' use of CAGE questions to screen for alcoholism [29]. The use of the CAGE increased from 6 to 77%. The CAGE can also be used as an oral questionnaire embedded in history taking [30].

Nevertheless, there is no consensus about the diagnostic value of the CAGE. Inconsistent reports for the CAGE and other questionnaires reflect, in part, different study populations, varying reference standards for defining problem drinking and the presence of confounding variables in published validation studies [18], [31]. A recent systematic review performed by Fiellin and his colleagues pointed that the CAGE questions proved superior for detecting alcohol abuse and dependence in primary care [32].

Therefore, we performed a diagnostic meta-analysis, based on guidelines for conducting systematic reviews of studies evaluating the accuracy of diagnostic tests [33], of all published studies to evaluate the diagnostic value of the CAGE in general clinical inpatients, ambulatory medical patients, and primary care patients. To avoid bias resulting from different definitions between studies we restricted ourselves to studies using DSM criteria as the reference standard for alcohol abuse and dependence. By analyzing three different subpopulations and pooling them separately, we also want to avoid clinical heterogeneity. Finally, we tested a new method for pooling ROC curves [34].

Section snippets

Instruments

The CAGE questionnaire was developed from a clinical study undertaken in 1968 at North Carolina Memorial Hospital by Ewing [35] to screen for alcohol abuse and alcohol dependence. It was initially validated by Mayfield and colleagues in a psychiatric service in 1970 [36], and later in different countries and populations [37], [38], [39]. The CAGE is an acronym for each of four questions (Table 1). The questionnaire can be administered in less than 60 sec, and is generally used with a threshold

Results

We identified 35 articles using the DSM criteria as a gold standard to test the screening characteristics of the CAGE in a clinical population. Only 10 studies complied with our inclusion and exclusion criteria [14], [20], [49], [50], [51], [52], [53], [54], [55], [56], [57].

Discussion

The CAGE questionnaire is a widely used diagnostic tool. It is mentioned in more than 200 publications related to alcohol problems. Despite a systematic review of screening for alcohol problems in primary care [32], the most surprising result of this study is the absence of a diagnostic meta-analysis and the finding that only a few studies evaluated the CAGE's diagnostic accuracy in relation to DSM criteria. The questionnaire's diagnostic value also was studied in many articles with other

References (79)

  • A Umbricht-Schneiter et al.

    Alcohol abuse: comparison of two methods for assessing its prevalence and associated morbidity in hospitalized patients

    Am J Med

    (1991)
  • S.M Wiseman et al.

    Assessment of drinking patterns in general practice

    J R Coll Gen Pract

    (1986)
  • M King

    At risk drinking among general practice attenders: prevalence, characteristics and alcohol related problems

    Br J Psychiatry

    (1986)
  • J.L Coulehan et al.

    Recognition of alcoholism and other substance abuse in primary care patients

    Arch Intern Med

    (1987)
  • M Cornel et al.

    The Medical profile of unidentified problem drinkers in general practice: test of a hypthesis

    Alcohol Alcohol

    (1995)
  • Hoeksema SL, Oltheten JTM, Mook JHA, Mulder JD. Huisarts en problematisch alcoholgebruik; een overzicht van onderzoek....
  • H.A.J Van Rens et al.

    Herkenning van problematisch alcoholgebruik in de huisartsenpraktijk

    Huisarts Wet

    (1989)
  • D Huas

    Consommateurs a risques et buveurs excessifs

    Excercer

    (1992)
  • D.G Buchsbaum et al.

    Alcohol consumption patterns in a primary care population

    Alcohol Alcohol

    (1991)
  • A.L Leckman et al.

    Prevalence of alcoholism in a family practice center

    J Fam Pract

    (1984)
  • B Aertgeerts et al.

    Screening properties of questionnaires and laboratory tests for the detection of alcohol abuse or dependence in a general practice population

    Br J Gen Pract

    (2001)
  • K Magruder-Habib et al.

    Alcohol abuse and alcoholism in primary health care settings

    J Fam Pract

    (1991)
  • W.L Adams et al.

    Alcohol abuse in elderly emergency department patients

    J Am Geriatr Soc

    (1992)
  • K.A Bradley et al.

    Variations on the CAGE alcohol screening questionnaire: strengths and limitations in VA general medical patients

    Alcohol Clin Exp Res

    (2001)
  • H.L Hoeksema et al.

    The value of laboratory tests for the screening and recognition of alcohol abuse in primary care patients

    J Fam Pract

    (1993)
  • M Reynaud et al.

    Patients admitted to emergency services for drunkenness: moderate alcohol users or harmful drinkers?

    Am J Psychiatry

    (2001)
  • G.L Phelps et al.

    Bright light in dark places: physician recognition of alcoholism

    JSC Med Assoc

    (1990)
  • S.A Maisto et al.

    Contrasting self-report screens for alcohol problems: a review

    Alcohol Clin Exp Res

    (1995)
  • J.T Hays et al.

    Alcoholism: early diagnosis and intervention

    J Gen Intern Med

    (1987)
  • E Girela et al.

    Comparison of the CAGE questionnaire versus some biochemical markers in the diagnosis of alcoholism

    Alcohol Alcohol

    (1994)
  • O Nilssen et al.

    The CAGE questionnaire and the Short Michigan Alcohol Screening Test in trauma patients: comparison of their correlations with biological alcohol markers

    J Trauma

    (1994)
  • M Olfson et al.

    The detection of alcohol problems in a primary care clinic

    J Community Health

    (1993)
  • R.P Tempier

    Screening for risk factors for alcohol consumption in the Quebec Health Survey

    Can J Public Health

    (1996)
  • D.R Lairson et al.

    Screening for patients with alcohol problems: severity of patients identified by the CAGE

    J Drug Educ

    (1992)
  • A Perdrix et al.

    Detection of alcoholism in the medical office: applicability of the CAGE questionnaire by the practicing physician

    Schweiz Med Wochenschr

    (1995)
  • K Lawner et al.

    Implementation of CAGE alcohol screening in a primary care practice

    Fam Med

    (1997)
  • D.A Fiellin et al.

    Screening for alcohol problems in primary care: a systematic review

    Arch Intern Med

    (2000)
  • J.M Kitchens

    Does this patient have an alcohol problem?

    JAMA

    (1984)
  • J.J Deeks

    Systematic reviews of evaluations of diagnostic and screening tests

    BMJ

    (2001)
  • Cited by (109)

    • Spectrum, Screening, and Diagnosis of Alcohol-related Liver Disease

      2023, Journal of Clinical and Experimental Hepatology
      Citation Excerpt :

      Two or more positive responses constitute a result that may raise suspicion of AUD and should be investigated further.62 The CAGE questionnaire was evaluated by a meta-analysis; a cutoff of more than two positive answers found a sensitivity of 71% and specificity of 90%.63 Table 1 summarizes the current screening tools and diagnostic criteria for AUD.

    • Alcohol-related liver disease: Clinical practice guidelines by the Latin American Association for the Study of the Liver (ALEH)

      2019, Annals of Hepatology
      Citation Excerpt :

      Two or more out of the four possible “yes” responses indicate that the possibility of AUD should be investigated further [64]. A meta-analysis that evaluated the CAGE questionnaire using a cutoff of more than two positive responses, found an overall sensitivity of 71% and specificity of 90% [65]. The diagnosis of ALD relies on a history of significant alcohol intake, clinical features, laboratory abnormalities and on excluding other causes of liver disease.

    View all citing articles on Scopus
    View full text