According the Substance Abuse and Mental Health Services Administration (2013), an estimated 23.9 million Americans are current illicit drug users. These numbers raise serious concerns given the devastating consequences of addiction borne by individuals, families, communities and society as a whole. Furthermore, The National Institute on Drug Abuse (2012) estimates that addiction costs the nation in excess of $600 billion annually. Given this, it is important to evaluate instruments and measures utilized by treatment professionals to assess and diagnose substance use disorders. The purpose of this paper shall be to undertake such an evaluation of two widely used assessment protocols – the Substance Abuse Subtle Screening Inventory (SASSI) and the Addiction Severity Index (ASI) – by appraising the reliability, validity, feasibility, accessibility, and overall appropriateness of these measures in screening and assessing adults with substance use disorders.
On September 21, 2014, a search for literature was conducted utilizing the following databases: Academic Search Premier, PsycINFO (EBSCO), Google Scholar, and PubMed. During the initial search, keywords utilized were assessment, instrument, and substance abuse. A review of the literature subsequent to this search revealed that the SASSI and ASI are two of the most widely used assessment protocols for adults with substance use disorders. As such, separate searches were conducted in the above-described databases specific to these instruments: the first search included the keywords SASSI, substance abuse, reliability, and validity; and the second search included the key words addiction severity index, substance abuse, reliability, and validity. Search results were limited from publication date 2000 to present and peer-reviewed, scholarly articles. The abstracts and/or full articles were reviewed to determine relevance to the purpose of this paper prior to inclusion herein.
Substance Abuse Subtle Screening Inventory (SASSI)
Screening instruments can be helpful during assessment to aid in determining if an individual has a problem, is at risk for developing a substance use disorder, or whether there is need for further assessment. Such instruments can generally be divided into two categories – direct and indirect scales. Direct scales ask questions that are obviously related to substance abuse (i.e. are face valid) by referencing associated signs and symptoms of substance use disorder (eg. “I have used alcohol or ‘pot’ too much or too often”) (Feldstein & Miller, 2007; Miller, Woodson, Howell, & Shields, 2009). One of the concerns with direct scales is that they are susceptible to falsification (Feldstein & Miller, 2007). In contrast, indirect scales include questions that do not appear to be related to substance abuse (eg. “I am rarely at a loss for words”) but which are included because the normative sample differentially endorsed these items (Miller et al., 2009). Indirect scales are thought to detect substance use disorders in spite of the respondent’s denial, attempt to conceal, or lack of awareness (Feldstein & Miller, 2007).
The SASSI, which was originally introduced in 1988, combines two direct scales with several indirect scales. The 10 scales embedded in the SASSI-3 include the total scale, Face Valid Alcohol (FVA), Face Valid Other Drugs (FVOD), Symptoms (SYM), Obvious Attributes (OAT), Subtle Attributes (SAT), Defensiveness (DEF), Supplemental Addiction Measure (SAM), Family History (FAM), and Tendency to Involvement in Correctional Setting (COR) (Miller et al., 2009). The instrument is comprised of 93 questions on two sides of a single sheet (Miller & Lazowski, 1999). On the front side there are 67 “true/false” items with both direct and indirect content, while the back side contains 26 Likert-scaled questions which directly query substance use and negative consequences (Miller & Lazowski, 1999). It is claimed that the SASSI is “the most frequently used empirically based screen in drug and alcohol treatment centers” in the United States (Feldstein & Miller, 2007, p. 42). If cut-offs are exceeded for the decision rules, the individual is designated as “high degree of probability of having substance dependence disorder” (Feldstein & Miller, 2007).
Validity and Reliability
Data from a number of studies reveal that internal consistency for the direct, face valid scales to be relatively high (Cronbach’s alpha ranging from 0.86 to 0.94) (Feldstein & Miller, 2007; Gray, 2001). However, the internal consistency for the SASSI subtle scales is generally lower than the direct scales and highly variable (Cronbach’s alpha ranging from 0.03 to 0.82) (Feldstein & Miller, 2007; Gray, 2001; Miller et al., 2009). SASSI classifications have been found to converge with results from direct screening instruments such as the CAGE and MAST, though the contributions of the subtle scales appear at times to undermine its convergent validity (Feldstein & Miller, 2007). There are some concerns as well related to divergent validity – SASSI subtle scales correlate positively with conduct disorder, depression, social anxiety, general distress and trauma, and suicidal ideation or attempts (Feldstein & Miller, 2007; Gray, 2001). The SASSI detects roughly 7 in 10 cases that have actually been diagnosed with substance use disorder (Feldstein & Miller, 2007). This is comparable to the sensitivity of direct, face valid scales such as the CASE, MAST, and AUDIT (Feldstein & Miller, 2007). Furthermore, Feldstein & Miller (2007) argue that the SASSI “is no more sensitive or accurate, or less susceptible to falsification . . . than simpler direct scales available in the public domain” (p. 47). Finally, test-retest correlations average from r=0.66 at two weeks to 0.61 at four weeks with direct scales showing more stability than the subtle, indirect scales which purport to measure stable traits (Feldstein & Miller, 2007).
Feasibility, Accessibility, and Overall Appropriateness
According to the SASSI Institute (2014), pencil and paper starter kits may be ordered for $135 for the small kit or for $235 for the large kit. The instrument takes approximately 15 minutes to administer and score (The SASSI Institute, 2014). Even though the cost is not necessarily prohibitive and the instrument is easy to administer and score, it may not be the best choice for treatment professionals. First, the assumption that persons with substance use disorder will attempt to falsify screening and assessment instruments in disproportionate numbers – and hence the need for indirect screening and assessment – is false (Gray, 2001). Second, as previously noted, the SASSI is no more sensitive or accurate than direct scales that are available for free in the public domain. Finally – and of paramount importance — studies have consistently reported a positive correlation between SASSI scores and ethnicity with minorities being significantly more likely to be designated as having a high probability of substance use dependence than Caucasians (Feldstein & Miller, 2007).
Addiction Severity Index (ASI)
Developed more than 25 years ago, the ASI is a semi-structured clinical research interview designed to measure client status in seven functional domains: alcohol and drug use, medical and psychiatric health, employment/self-support, family relations, and illegal activity (McClellan, Cacciola, Alterman, Rikoon, & Carise, 2006). For each of the domains, information is gathered in relation to the client’s lifetime and recent history (30 days and 6 months) in order to assess duration, severity, frequency, and intensity of problems and to monitor change in subsequent administrations of the instrument (McClellan et al., 2006). The instrument is comprised of two summary indices for each domain – interviewer severity ratings (ISRs) and composite scores (CSs) (Makela, 2004). For the ISRs, interviewers follow what is described as a complicated procedure to determine a rating of the severity and ‘need for additional treatment’ in each of the domains (Makela, 2004). The ASI is widely used – it has become standard in most clinical trials involving addiction, and it is part of the standard clinical assessment in 20 states and 50 U.S. cities, as well as the Veteran’s Administration, the Indian Health Service, and the federal prison system (McClellan et al., 2006).
Validity and Reliability
Inter-rater reliability for the ASI is relatively high though the domains of drugs, family/social, and psychiatric appear unstable in a number of studies (Makela, 2004). Because composite scores are not as influenced by subjective judgment, the inter-rater reliability of these scores is consistently high (Makela, 2004). The authors of the ASI report high test-retest reliability, and this is generally supported in the literature, except in cases where the instrument was administered to special populations such as those with mental health issues and homeless patients (Makela, 2004). As for internal consistency, three of the seven composite scores (medical status, alcohol use, and psychiatric status) produce reliable results whereas the other domains appear to be more variable and produce low internal consistencies (Makela, 2004). Regarding discriminant validity, the correlations between ASI domains are usually low, supporting the independence of the derived scales, with the exception of the drug measures and legal measures, and the psychiatric and family/social measures which are sometimes moderately correlated (Cacciola, Alterman, Habing, & McClellan, 2011). Additionally, correlations are high between the two summary measures with several of the problem areas, though intense training of the interviewers appears to offset, in part, this problem (Makela, 2004). Correlations between the ASI and external criterion variables (such as the CAGE, MAST, and DAST) are by no means uniformly high and range from 0.50 to 0.73 (Cacciola et al., 2011; Makela, 2004).
Feasibility, Accessibility, and Overall Appropriateness
The ASI is a useful tool in that it broadens the perspective of the assessment to other important domains beyond mere patterns of drug and alcohol use. Furthermore, it combines objective and subjective measures in treatment planning. With an administration time of approximately one hour and relative ease in scoring, the instrument is a nice fit in terms of time allotment and expectations for an initial assessment (McClellan et al., 2006). Furthermore, the instrument is widely accessible and may be obtained for free in the public domain. With that being said, as previously noted, studies have shown that without extensive interviewer training, the reliability of results may suffer (Makela, 2004). To this extent, whether a particular agency or organization should adopt the ASI depends in large part on its means and willingness to properly train staff in the administration thereof. Where such training is not available, simpler direct measures (such as the AUDIT and/or DAST) might be better options though these instruments lack the multi-dimensionality of the ASI.
Early intervention and detection of substance use disorders is of paramount importance. Both the SASSI and ASI are tools that are frequently used by treatment professionals during the initial assessment phase of client intervention. The premise of the SASSI seems to be that subtle, indirect screening is necessary to accurately assess substance users who, by inference, are assumed to be prone to deceit. The problem is that, as previously noted, while there might be some anecdotal evidence that this premise is true, the research does not support the necessity to employ indirect, subtle screening with substance users. Given this, and other issues of validity and reliability of the subtle scales, if the SASSI is utilized, it may be best to give significantly more weight to the face-valid, direct scales in determining whether a problem might exist or whether further assessment is necessary. With that being said, the SASSI is a self-administered and can be taken and scored quickly and thus has some advantages in this regard. A more in-depth approach would be to utilize the semi-structured interview provided by the ASI. Representing a more holistic approach, the ASI assesses a number of domains and, for the most part, has been found to be a psychometrically sound measure. However, because it involves clinical judgment and subjective criteria ratings, it should only be utilized by staff who have received sufficient prior training. Ultimately, where training is unavailable, direct and more simple instruments – such as the MAST, AUDIT and/or DAST – are likely the best options for treatment providers who are looking to support initial clinical impressions with empirical data.
(This paper was originally submitted in September 2014 by the Site Administrator in partial fulfillment of the Masters of Science in Social Work degree at the University of Louisville)
Cacciola, J. S., Alterman, A. I., Habing, B., & McClellan, A. T. (2011). Recent status scores for version 6 of the Addiction Severity Index (ASI-6). Addiction, 106(9), 1588-1602.
Feldstein, S. W., & Miller, W. R. (2007). Does subtle screening for substance abuse work? A review of the Substance Abuse Subtle Screening Inventory (SASSI). Addiction, 102(1), 42-50.
Gray, B. T. (2001). A factor analytic study of the Substance Abuse Subtle Screening Inventory (SASSI). Educational and Psychological Measurement, 61(1), 102-118.
Makela, K. (2004). Studies of the reliability and validity of the Addiction Severity Index. Addiction, 102(1), 42-50.
McClellan, A. T., Cacciola, J. C., Alterman, A. I., Rikoon, S. H., & Carise, D. (2006). The Addiction Severity Index at 25: Origins, contributions, and transitions. The American Journal on Addictions, 15(2), 113-124.
Miller, C. S., Woodson, J., Howell, R. T., & Shields, A. L. (2009). Assessing the reliability of scores produced by the Substance Abuse Subtle Screening Inventory. Substance Use and Misuse, 44(8), 1090-1100.
Miller, F. G., & Lazowski, L. E. (1999). The adult SASSI-3 manual. Springville, IN: The SASSI Institute.
National Institute on Drug Abuse. (2012). Principles of drug addiction treatment: A research-based guide (NIH Publication No. 12-4180). Washington, DC: U.S. Government Printing Office.
Substance Abuse and Mental Health Services Administration. (2013). Results from the 2012 National Survey on Drug Use and Health: Volume I. Summary of National Findings (NSDUH Series H-46, HHS Publication No. SMA 13-4795). Rockville, MD: Substance Abuse and Mental Health Services Administration.
The SASSI Institute. (2014). The SASSI Institute: Early intervention saves lives. Retrieved from https://www.sassi.com/