Evaluating Assessment Instruments Used to Measure Substance Abuse and Dependence

Title: Alcohol Use Disorders Identification Test By: Babor, Thomas F., de la Fuente, Juan Ramon, Saunders, John, Grant, Marcus, Aasland, O. G., 19930101, Vol. 14

Database: Mental Measurements Yearbook with Tests in Print

Review of the Alcohol Use Disorders Identification Test by PHILIP ASH, Director, Ash, Blackstone and Cates, Blacksburg, VA: The Alcohol Use Disorders Identification Test (AUDIT) (Allen & Columbus, 1995, pp. 260-261) was developed in a six-country (Australia, Bulgaria, Kenya, Mexico, Norway, USA) World Health Organization collaborative project to design a screening instrument to identify people whose alcohol consumption has become hazardous and harmful to their health. It should be noted that this is the first instrument of its kind to be derived on the basis of a cross-national study (Babor, de la Fuente, Saunders, & Grant, 1992). Screening with the AUDIT provides data useful in deciding about treatment alternatives such as brief intervention with heavy drinkers or specialized treatment for more seriously ill patients. The AUDIT package includes the AUDIT Core and an optional Clinical Screening Procedure. This procedure includes two questions about traumatic injury (e.g., "Have you injured your head since your eighteenth birthday?"), five on clinical evaluation (e.g., presence of hand tremor), and a blood test (Serum GGT). Although the Clinical Screening Procedure does not refer directly to problems with alcohol, it may elicit relevant information where patients are defensive toward alcohol-specific questions. To develop the AUDIT Core, a 150-item assessment schedule was administered to subjects (N = 1888) attending representative primary health care facilities in the participating countries. A 10-item self-report questionnaire (the AUDIT) was selected to cover the domains of alcohol consumption, drinking behavior, and alcohol-related problems. It may be paper-and-pencil self-administered or administered in an interview in 2 minutes. Questions 1-3 measure quantitative alcohol consumption, 4-6 drinking behavior, 7-8 adverse reactions, and 9-10 alcohol-related problems. Responses to each question are scored from 0 to 4, for a maximum possible AUDIT score of 40. Among those diagnosed as having hazardous or harmful alcohol use, 92% had an AUDIT score of 8 or more, and 94% of those with nonhazardous consumption had a score of less than 8 (Saunders, Aasland, Babor, de la Fuente, & Grant, 1993). According to the authors' research report (Saunders et al., 1993, p. 799), the differences between the AUDIT and most other existing questionnaires include the following: (a) it tries to identify problem drinkers at the "less severe end of the spectrum" rather than those with established dependence or alcoholism; (b) it emphasizes hazardous consumption and frequency of intoxication rather than drinking behavior itself and its consequences; and (c) it refers to alcohol experiences in the past year as well as over the patient's lifetime, improving relevance to current drinking status, and it does not require the test-taker to identify himself or herself as a problem drinker. Two questions (9-10) have 3 responses (scored 0, 2, 4). The remaining questions (1-8) have 4-choice responses based upon frequency (from, e.g., "never," "1 or 2" to, e.g., "daily or almost daily," "10 or more"). They yield item scores from 0 (lowest frequency choice) to 4 (highest frequency choice). The authors anticipated that using specific frequency continua will reduce underreporting of adverse effects. The manual (Babor et al., 1992) is a 28-page document that introduces the AUDIT by discussing the advantages of screening for alcohol problems, the developmental history of the AUDIT, applications of the AUDIT to the early identification of alcohol-related problems, and the scoring and interpretation of the AUDIT. The authors point out that although alcohol screening tests have most often been used to identify who may probably be alcohol abusers (case finding), the AUDIT is directed at screening known drinkers to deal with their problems and treatment approaches. The AUDIT is claimed to be the first screening test specifically for use in primary care cases. The seven text chapters of the manual are followed by a list of references and six appendices covering research guidelines, validity evidence in the AUDIT, etc., including a copy of the AUDIT. Extensive research has been undertaken on the reliability, validity, and other psychometric characteristics of the AUDIT. Both test-retest and internal consistency measures (Fleming, Barry, & MacDonald, 1991) have shown satisfactory reliability. High intrascale reliabilities (alpha coefficient mean values of .93 and .81) were found among the drinking patients' drinking behavior and adverse psychological reactions domains (Saunders et al., 1993, pp. 794-795). Concurrent, construct, and discriminant validities of the AUDIT were assessed by Bohn, Babor, and Kranzler (1995, p. 425ff). Significant concurrent validities were found against other alcoholism measures such as the MAST (Michigan Alcohol Screening Test) and the MacAndrews scales (r = .31 to r = .887). Coefficients for the AUDIT Core were consistently higher than for the AUDIT Clinical. Construct validities for five risk factors, four drinking consequences, and three drinking attitudes showed significant correlations (r = .27 to r = .88) for 11 of the 12 measures for AUDIT Core for male subjects (n = 107), but fewer significant correlations for AUDIT Core for female subjects (n = 91), or for either sex for AUDIT Clinical scores. An analysis of discriminant validity found a significant difference between nondrinkers and harmful drinkers, but no significant gender or gender x drinker group difference. Similar validity evidence was reported by Babor et al. (1992, p. 21). SUMMARY. Overall, the AUDIT is a useful device for discriminating between alcoholics and medical patients and in the early detection of hazardous or harmful drinking, and is more successful than the Michigan Alcohol Screening Test (MAST) in discriminating hazardous drinkers from nonhazardous drinkers. A well-written manual and substantial published supporting research commend the instrument for serious consideration in the assessment of people with difficult alcohol problems. Its multinational origins and translations also commend it as a device for conducting cross-cultural alcoholism studies. REVIEWER'S REFERENCES Fleming, M. F., Barry, K. L., & MacDonald, R. (1991). The Alcohol Use Disorders Identification Test (AUDIT) in a college sample. International Journal of the Addictions, 26, 1173-1185. Babor, T. F., de la Fuente, J. R., Saunders, J., & Grant, M. (1992). Programme on Substance Abuse: AUDIT--The Alcohol Use Disorders Test: Guidelines for Use in Primary Health Care (an update of WHO Document No. WHO/MNH/DAT/89.4 under the same title) [Switzerland]: World Health Organization. Saunders, J. B., Aasland, O. G., Babor, T. F., de la Fuente, J. R., & Grant, M. (1993). Development of the Alcohol Use Disorders Identification Test (AUDIT): WHO collaborative project on early detection of persons with harmful alcohol consumption--II. Addiction, 88, 791-804. Allen, J. P., & Columbus, M. (Eds.) (1995). Assessing alcohol problems: A guide for clinicians and researchers. Washington, DC: National Institute on Alcohol Abuse and Alcoholism. Bohn, M. J., Babor, T. F., & Kranzler, H. R. (1995). The Alcohol Use Disorders Identification Test (AUDIT): Validation of a screening instrument for use in medical settings. Journal of Studies on Alcohol, 58(4), 423-432.

Review of the Alcohol Use Disorders Identification Test by HERBERT BISCHOFF, Licensed Psychologist, Psychology Resources, Anchorage, AK: The Alcohol Use Disorders Identification Test (AUDIT) is a screening procedure produced by a multicultural collaboration of six nations under the sponsorship of the World Health Organization (WHO). The AUDIT is intended to be used primarily by health care providers to detect harmful or hazardous drinking habits. The AUDIT is not a diagnostic tool, and the authors are clear about its limitations and intended uses. The AUDIT is made up of two components. The first part is a core 10-item questionnaire that addresses drinking patterns including indications of "hazardous alcohol consumption, evidence of dependence symptoms, and harmful alcohol consumption." The second part, which is administered only in cases of a "positive" on the first part, is a clinical screening instrument of eight items covering trauma history, clinical examination, and a blood test. The intended use of the AUDIT is to detect, or more appropriately screen, persons who may be engaging in dangerous drinking habits, referred to as either harmful, hazardous, or evident of dependence. Referring to the ICD-10 the authors define harmful drinking as, "a pattern of use which is already causing damage to health." Hazardous drinking refers to "an established pattern of use carrying with it a high risk of future damage to health, physical or mental, but which has not yet resulted in significant medical or psychiatric ill effects." Evidence of dependence is defined by symptoms, patterns of behaviors, and debilitating social and occupational consequences of long-term alcohol dependence, as described in the DSM-IV and ICD-10. The AUDIT has also been found to have high levels of concurrent and construct validity. There have been numerous studies on the efficacy of using the AUDIT screening instrument, and it has been established as providing reliable and valid scores of harmful and hazardous drinking patterns with high levels of accuracy in detecting dependence in multinational and widely varied populations (Bohn, Babor, & Kranzler 1995; MacKenzie, Langa & Brown, 1996; Saunders, Aasland, Babor, de la Fuente, & Grant, 1993). Scores from the AUDIT have been found to be highly reliable in identifying persons who currently demonstrate the above patterns of behavior. The AUDIT appears to be a highly useable, quickly administered, and cost efficient tool for health care providers to use in the course of various types of clinical interviews. During the short time the AUDIT has been available, it has proven itself when compared to more traditional methods, such as the Michigan Alcohol Screening Test (MAST), TWEAK, and the MMPI's (Minnesota Multiphasic Personality Inventory) MacAndrew Alcoholism scale. Overall, the AUDIT may be an even more reliable indicator of problem drinking patterns with certain populations, such as ethnic minorities and the long-term unemployed (Cherpitel & Clark, 1995; Claussen & Aasland, 1993; Luckie, White, Miller, & Icenogle, 1995). There are, however, some limitations with the AUDIT and its manual. These include information on how the cutoff point was determined, how results vary from population to population, and how the 10 items were chosen. The information is covered in supplemental material provided by the test authors in the forms of varying research projects that have demonstrated the AUDIT to provide highly reliable and valid scores and to be a useful instrument. Another concern with the test manual is the lack of normative data; nor is there an explanation of how various cultures view harmful or hazardous drinking. This may leave the test administrator confused about the validity of scores from the AUDIT when there are no apparent considerations devoted to the importance of context, other than suggested ideal situations when administering the test. For instance, it is assumed a drink containing 10g of alcohol is the typical quantity per drink cross-culturally. Yet, there is no consideration given to how drinking is viewed culturally, or to the effects of perceived social desirability on the respondents' level of self-reporting. Moreover, there is no discussion on how possibly to control for these potential effects. SUMMARY. In summation, the AUDIT appears to be an easily administered screening tool that provides reliable scores for health care providers to ascertain if an individual is engaging in potentially damaging drinking patterns. The AUDIT's limitations include limited diagnostic utility and problems with the testing manual. It appears to be a useful tool for health care providers to facilitate treatment or interventions for persons already suffering the effects of alcohol misuse or those soon to be suffering the damaging effects of long-term harmful alcohol consumption. REVIEWER'S REFERENCES Claussen, B., & Aasland, O. G. (1993). The Alcohol Use Disorders Identification Test (AUDIT) in a routine health examination of long-term unemployed. Addiction, 88(3), 363-368. Saunders, J. B., Aasland, O. G., Babor, T. F., de la Fuente, J. R., & Grant, M. (1993). Development of the Alcohol Use Disorders Identification Test (AUDIT): WHO collaborative project on early detection of persons with harmful alcohol consumption--II. Addiction, 88, 791-804. Bohn, M. J., Babor, T. F., & Kranzler, H. R. (1995). The Alcohol Use Disorders Identification Test (AUDIT): Validation of a screening instrument for use in medical settings. Journal of Studies on Alcohol, 56(4), 423-432. Cherpitel, C. J., & Clark, W. B. (1995). Ethnic differences in performance of screening instruments for identifying harmful drinking and alcohol dependence in the emergency room. Alcoholism Clinical and Experimental Research, 19(3), 628-634. Luckie, L. F., White, R. E., Miller, W. R., & Icenogle, M. V. (1995). Prevalence of estimates of alcohol problems in a Veterans Administration outpatient population: AUDIT vs. MAST. Journal of Clinical Psychology, 51(3), 422-425. MacKenzie, D. M., Langa, A., & Brown, T. M. (1996). Identifying hazardous or harmful alcohol use in medical admissions: A comparison of AUDIT, CAGE, and Brief MAST. Alcohol and Alcoholism, 31(6), 591-599.