(PS1-44) Examination of the Addiction Severity Index in a Sample Not Recruited for Substance Use Problems

Many problems commonly assessed by measures of addiction severity can occur in the absence of substance use disorder (SUD); however, these problems are often used as indicators of SUD severity (e.g., interpersonal problems, legal problems, mental health) in clinical research studies. Despite their assessment in relation to addiction severity, many problems predate or co-occur with use rather than being a direct result of substance use. Many of the problems assessed for addiction severity could be attributed to other stressors (e.g., low socioeconomic status, homelessness) and may occur at high rates in populations of non-substance using individuals. The purpose of the current study is to determine the ability of the Addiction Severity Index (ASI; McLellan et al., 1992), a widely used measure of addiction related problems, to differentiate between problems of people who do and do not use substances, using data from a community sample of women not recruited specifically for a study on substance use.   

Participants were women identified as at-risk for HIV (N = 331); 38.7% identified as Black/African American, 51.1% as White, and 32.9% as Hispanic/Latina; mean age was 44.5 (SD = 11.8); 24.8% were currently homeless, 46.8% had experienced lifetime homelessness; and 40.8% met criteria for past-year SUD, assessed with a diagnostic interview based on DSM-5 Criteria (APA, 2013). Data were from a baseline assessment conducted as part of larger randomized trial evaluating an intervention to improve rates of HIV testing; the ASI was administered prior to randomization/intervention.

Composite scores for medical, employment, legal, family, and psychiatric problems were examined for women with and without SUD diagnoses. Non-parametric tests were used to compare ASI composite scores for individuals who had a past year SUD and those who did not. There were not significant differences between those with and without SUD on the medical and employment composites. In contrast, women with SUD reported significantly greater family (z = 2.55, p = .011), legal (z = 4.33, p < .001), and psychiatric composites (z = 4.28, p < .001) than women without SUD. Non-parametric tests were also used to compare ASI means for those who had any substance use in the past year (n = 271) and those who had none (n = 60). For medical, employment, family, and psychiatric composites, there were no significant differences between groups. Women with substance use in the past year had higher legal composites (z = 2.11, p = .035) than those with no past year use.

Findings suggest that the ASI, a commonly used assessment of problems related to substance use, may be capturing other life problems beyond the scope of substance use. Consequences that have previously been conceptualized as indicators of addiction severity may be better described as co-occurring or life problems that may or may not have direct relationship to substance use if they are being experienced in non-substance use populations. These findings underscore a need for discerning interpretation of addiction severity assessment and particular caution may be needed in interpreted the results of clinical trials that use the ASI and other problem-based assessments of substance use severity as a primary outcome measure.