Predicting DSM-IV dependence diagnoses from Addiction Severity Index composite scores

This study, using data from the Drug Evaluation Network System and a study conducted through the Center for Studies on Addiction of the University of Pennsylvania/Philadelphia Veterans Administration Medical Center, sought to determine the potential of the Addiction Severity Index (ASI) to serve as a screening instrument for Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) substance dependence. A significant positive correlation was found between ASI composite scores (CSs) and DSM-IV diagnoses of dependence in both the alcohol (r > .7) and drug (r > .5) domains (p < .01). Receiver operating characteristic analyses were run to predict DSM-IV alcohol and drug dependence diagnoses from the respective ASI CSs. Results showed good to strong prediction; ASI CSs identified dependent clients with approximately 85% sensitivity and 80% specificity. We recommend strategies for using ASI CSs as a diagnostic screening instrument in both research and treatment delivery environments.