Abstract
While screening tools have been developed to identify adolescents likely to have alcohol use disorders (AUDs), none of the available methods are optimal for general medical settings. This study explored the sensitivity and specificity of the frequency of drinking episodes in the prior month as an initial screen for AUDs. The subjects were 219 adolescents (ages 12 through 18) systematically recruited from the community, who participated in a baseline assessment as well as 1-year, 3-year, and 5-year follow-up visits. Subjects completed a self-report form indicating their frequency of use of different substances in the month prior to the assessment. DSM-IV AUD diagnoses were determined by SCID. At baseline, 10 of 219 subjects met DSM-IV criteria for an AUD. At a threshold of 3 or more drinking episodes in the past month, the screen was 90% sensitive, correctly classifying 9 of 10 AUD cases, and 83.7% specific, correctly classifying 175 of 209 cases without AUDs. The diminishing specificity of this screen over the follow-up assessments indicated that this method may be useful for adolescents, but not for young adults. These results indicate that an assessment of the frequency of alcohol use in the prior month provides an initial screen with acceptable sensitivity and specificity for use with adolescents.