New scales to assess change in the Addiction Severity Index for the opioid, cocaine, and alcohol dependent.

Abstract

Addiction Research and Treatment Center Analyses were performed to construct and confirm the validity of new conjoint intake and 6-month follow-up scales for the Addiction Severity Index (A. T. McLellan, L. Luborsky, G. E. Woody, & C. P. O'Brien, 1980) applied to a diverse sample of substance dependence patients (N = 1,008). A multistage scaling strategy identified 5 psychometrically integral addiction problem scales. Exploratory item and factor analyses, confLrrnatory oblique item clustering, and variance partitioning verified that the scales comprised relatively little common variance and that each retained a substantial amount of unique and reliable variance. Resulting scales (Psychiatric, Drug, Alcohol, Family, and Legal Problems, respectively) were highly internally consistent and structurally stable overall, at intake and follow-up and across gender, age, ethnicity, and substance abuse categories. Concurrent and predictive validity over 2 years were supported for clinical subsamples based on comorbid psychopathology and mood, HIV risk behaviors, personality indices, urine toxicology, and criminal records. The Addiction Severity Index (ASI) is probably the most commonly used instrument in the field of substance abuse and has been a Arthur I. Alterman, Terry G. Cook, David Metzger, Megan J. Rutherford, and John S. Cacciola, Depart- ment of Psychiatry, University of Pennsylvania, and Treatment Research Center, Veterans Affairs Medical Center, Philadelphia; Paul A. McDermott, Graduate School of Education, University of Pennsylvania; Lawrence S. Brown, Jr., Addiction Research and Treatment Center, Brooklyn, New York. Megan J. Rutherford is now at the Alcohol and Drug Institute, University of Washington. This research was supported by the National Institute on Drug Abuse Grant (DA05186), the National Institute of Alcohol Abuse and Alcohofism Grant AA08480, and a Department of Veterans Affairs grant. Correspondence concerning this article should be addressed to Arthur I. Alterman, Treatment Research Center, 3900 Chestnut Street, Philadelphia, Pennsylvania 19104. Electronic mail may be sent to Alterman@Research.TRC.UPenn.Edu. remarkably useful multidimensional clinical and evaluation tool (Cacciola et al., 1997; Stoffel- mayr, Mavis, & Kasim, 1994). At the same time, research in the past decade has begun to demonstrate some limitations in the two sets of ASI problem indices used to summarize respon- dent status. One set, the interviewer severity ratings (ISRs), is the interviewer's determina- tion of a summary score based on lifetime and recent problems that describes the need for treatment at intake in each of the ASI's seven problem areas. However, the interrater reliabil- ity of ISRs has been shown to be modest when applied by raters in the field (Alterman, Brown, Zaballero, & McKay, 1994; Hodgins & EI- Guebaly, 1992; Stoffelmayr et al., 1994). Our group has recently developed a set of psycho- metrically sound, standardized ASI intake scales based on both lifetime and recent problem items that are designed as a reliable and objective alternative to the ISRs (McDermott et al., 1996). The composite scores (CSs) are arithmeti- cally derived scales, used both at intake and at 233