A Comparison of the Alcohol Dependence Scale and Clinical Diagnosis of Alcohol Dependence in Male Medical Outpatients

Abstract

The Alcohol Dependence Scale (ADS) is a brief self-report instrument designed to assess the alcohol dependence syndrome. Although previous studies have reported its validity, it has only been studied in populations who were presenting for assessment or treatment of alcohol problems, and recent studies have shown lower mean scores than would be expected. To assess its utility in a primary medical setting, we retrospectively examined our experience in two groups of medical outpatients: an early intervention group and a group of medically ill alcoholics. We administered the ADS to 61 male veterans being referred to a special clinic for alcohol-related medical problems. Subjects were also assessed for alcohol dependence by experienced clinicians using a DSM-III-R checklist. ADS scores overall were much lower than would be expected, especially in the medically ill alcoholics. When subjects were dichotomized into dependent or nondependent, there was poor agreement between the two methods. The correlation between the scores was significant, however, and lowering the cut-point on the ADS from 13/14 to 2/3 improved agreement. Medically ill alcoholics who are not presenting for assessment or treatment of alcohol problems may either lack insight into their drinking habits or may be reluctant to disclose them. Although conclusions from this study are tentative until more formal studies are done, it may be most prudent to use other instruments, such as the Michigan Alcoholism Screening Test, in medical outpatients. If the ADS is used, lowering the cut-score to 2/3 may reduce false negatives. Further research on the use of the ADS in different settings and with different populations is clearly needed.