Evolution Of The Sassi Essay, Research Paper

The emergence, psychometric soundness, and continued development of the Substance Abuse Subtle Screening Inventory (SASSI), (Miller, 1985, 1994, & 1997) is reviewed from its original version through it current, third revision. The SASSI is a screening tool designed to detect chemical dependency in individuals, regardless of attempts to conceal, minimize or deny substance use involvement. The SASSI s nine subscales are based on clusters of items that were empirically derived and validated through an extensive period of testing and refinement with different groups of individuals clinically diagnosed as chemically dependent, as well as normative groups of nonaddicts. Shortcomings detected through various research projects are addressed on an ongoing basis with the result of producing an increasingly accurate and reliable instrument with each revision.

The Evolution of the Substance Abuse Subtle Screening Inventory:

Development and Revisions of a Criterion Referenced Instrument Designed to Detect Substance AbuseWithout Regard to Defensiveness Issues

That substance abuse is one of the most pressing problems facing the world today is not news. This particular social problem reaches across all lines of socioeconomic status, race, age, sex, and education level. According to figures from the U.S. Department of Health and Human Services, drug abuse kills 14,000 Americans each year, and places a nearly $70 billion dollar burden on the taxpayers in unnecessary health care costs, additional law enforcement, auto accidents, crime, lost productivity, incarceration and other criminal justice system costs. (National Clearinghouse on Alcohol & Drug Information, 1997). According to a study prepared for the National Institute on Drug Abuse and the National Institute on Alcohol Abuse and Alcoholism (1992), the estimated total economic cost of alcohol and drug abuse for 1992 was $245.7 billion, an increase in excess of 50% over 1985 cost estimates. This increase was attributed to four primary contributors: (1) the epidemic of heavy cocaine use, (2) the HIV epidemic, (3) an eightfold increase in state and federal incarcerations for drug offenses, and (4) a threefold increase in crimes attributed to drugs. (NIDA & NIAAA, 1992).

The director of the NIAAA has noted that, while alcohol is a legal drug, used by many without any problem, for persons with the clinical disorders of alcohol abuse or dependence, drinking can lead to profound health, social, or economic consequences that affect not only the person but other persons and society at large. (National Institute of Health, 2000). In order to address this critical need, NIAAA supports, among other projects, studies on screening for alcohol problems and brief intervention. (NIH, 2000). The U.S. Department of Health and Human Services is likewise committed to finding solutions to this prevalent issue, supporting treatment for individuals and noting that the need for treatment continues to grow annually. (Substance Abuse and Mental Health Services Administration, 2000).

Accurate and comprehensive assessment is fundamental to both treatment of, and research on, alcohol problems. (NIH, 1995). Particularly in the current environment of stringent controls on health care costs and service utilization, the clinician is deeply concerned about issues such as ease of administration, scoring, and interpretation, as well as cost, time, and acceptability of the measure by managed care organizations. (Allen, Litten, & Anton, 1992).

While better assessment of alcoholic patients does not ensure more specific or more effective treatment, chances for successful rehabilitation are clearly enhanced if specific patient needs can be more accurately identified and if treatment can be tailored accordingly. (Allen, 1991, p. 183).

Initial Development of the SASSI

In response to the need for an accurate, simply administered substance abuse screening tool that would be time and cost effective, the Substance Abuse Subtle Screening Inventory (SASSI), was introduced in 1985. (Miller, 1985). Four primary objectives were established in the development of the SASSI: 1) administration not restricted to professionals; 2) quick administration; 3) objective scoring; and 4) simplified interpretation through decision rules a form of cookbook approach. (Miller, 1985, p. 3-3).

Because the original concept was to develop a criterion referenced instrument that would effectively distinguish substance abusing/dependent individuals, early research involved extensive research into the selection of appropriate items. This was followed by validation efforts to determine whether the items themselves would distinguish abusers from non-abusers. Initial item selection focused on obtaining content that would cover a wide variety of substance abuse issues. Early abuse indicators and pre-disposing factors, as well as chronic abuse factors, were incorporated into this process. (Miller, 1985).

The developers conducted reviews of the Michigan Alcoholism Screening Test (MAST), (Selzer, 1971), and other scales that differentiated substance abusers from the general population. (Miller, 1985). Likewise, instruments measuring substance abuse-related constructs, such as the Sensation Seeking Scale, (Zuckerman, 1979), and the Locus of Control Scale, (Rotter, 1966), were examined. Additional subtle (i.e., non face valid) items were drawn from the MMPI for the specific purpose of countering attempts at faking. (Miller, 1985).

Items were designed so that they appeared to be unrelated to substance use. The bulk of the SASSI consists of these non-face valid items, and a true-false, and scaled (0-3) was used. (Miller, 1985). Nearly one thousand potential items were pre-screened through administration to at least one sample of abusers and one control group. Subsequently, sets of 1-200 prescreened items were administered to small criterion groups. Those items found to significantly differentiate between abusers and control groups were retained and combined with another set of prescreened items and then presented to new criterion groups. This process was repeated, however, it lacked systematic administration of all items across demographic factions. (Miller, 1985).

Early criterion groups were recruited from a broad spectrum of resources. Normative samples were drawn from community adult education groups, church groups, and volunteer community action groups. Substance abuse groups were selected from residential treatment centers, halfway houses, detox centers, referral centers, and community based self-help support groups. (Miller, 1985).

The Provisional SASSI

Final selection of the items that comprised the provisional version of the SASSI involved inclusion of additional items aimed at specific issues, and validating the final content through the combined results of three separate studies. At this stage in the development process, the item pool consisted of a nucleus of items that had differentiated abusers from the same sex control groups during pilot studies, along with a variety of additional items. Of particular significance was the inclusion of all 130 items from the Psychological Screening Inventory (PSI). (Lanyon, 1973).

The PSI s defensiveness and factorial scales were used to identify other items that measured general defensiveness or maladjustment. (Miller, 1985, p. 3-7). Ultimately, the resulting 600+ item pool was used in all three validation studies.

Presciption Drug Group

One of the three studies proposed to generalize the pilot work to an additional group of individuals classified as straight drug abusers . (Miller, 1985, p. 3-7). This term was defined as members of the general population who had become physically or psychologically dependent upon mood altering prescription medications taken to address medical conditions or symptoms. (Miller, 1985, p. 3-7).

Several hundred questionnaires were distributed door to door to individuals selected through census data as a cross-representation of the total metropolitan population. Despite inclusion of self-addressed, stamped envelopes, the return rate was relatively low, (n = 41 males, 57 females). Further compounding the difficulties with this validation study was the inability to verify classifications of individuals as straight drug abusers , however, the group did serve as controls for the third study.

Illicit Drug Group

The data for this study was drawn from an abuser sample of 50 heroin addicts from a long-term residential treatment program, and a control group (n = 50). Both groups were equally split across sexes, and were likewise matched in terms of age and education level.

The previously described 600+ items were administered to both groups. The PSI, which was incorporated into this item pool, contains an Alienation Scale designed to measure chronic dysfunction versus short term situational distress. (Lanyon, 1973). Analysis of the responses to these particular PSI items revealed similar responses from both groups, thus supporting the conclusion that differences in other item responses would more likely be a reflection of differences in degree of chemical abuse, as opposed to general maladjustment. (Miller, 1985).

The remainder of the items was narrowed down to 368, after deleting non-dichotomous items, life-change, and medical symptom items.

Using chi-square analysis, 76 of these 386 items were found to significantly (p 1) under these instruction sets, with RAP elevation occurring in 26% (12/46) of the results from the fake good instruction administration, and in 30% (14/46) of the results under the fake bad instruction set. (Myerholtz & Rosenberg, 1997).

The ability of subjects to effectively manipulate their classification virtually tore at one of the primary design features of the SASSI and the SASSI-2, that is, reduced vulnerability to faking that other screening instruments available. Examination of status changes based only on the non-face valid clinical scales revealed similar results, in that the same number of chemically dependent subjects, and 18 out of 32 non-chemically dependent subjects, were able to alter their status without detection through the clinical subscale scores. Accordingly, while the RAP scale may be sensitive to faking attempts, neither the remaining clinical scales nor the face-valid measures were immune to undetected deception. (Myerholtz & Rosenberg, 1997).

The researchers cautioned against assumptions that the SASSI-2 was less vulnerable to faking than other instruments. They did, however, acknowledge the possibility that these results may have been affected by the repeated administrations to the same subjects, and suggested a similar study using a between-subjects design. (Myerholtz & Rosenberg, 1997).

College Students

Myerholtz and Rosenberg (1998) conducted a similar, two-part series of studies a year later, wherein they again assessed the SASSI-2 in terms of its internal consistency, its relationship to the CAGE, MAC and MAST, and its reactivity to instruction sets or response biases. In this study, the previously suggested between-subjects design was used, in an effort to avoid effects of repeat administrations. Separate samples of college students were used in the two parts of this study.

The first part of the study examined the test-retest reliability of the SASSI-2 at two-week and four-week intervals. Two separate samples were used, one for the two week interval study (n = 64), and another for the four week interval study (n = 55).

Of the 64 college students in the two week interval group, nine did not return for the retest session after two weeks, and the sample was reduced to 55. The four week set was likewise reduced from 55 to 47. Demographically, both sets consisted predominantly of white female underclassmen, with similar racial content, mean ages, and age ranges.

Test-retest reliability coefficients for both the two week and four week intervals were moderate to high for the SASSI-2 subscales. Test-retest reliability as to the overall classification as chemically dependent or non-chemically dependent however, was moderate at two weeks, (Phi = .68, p.30). (Lazowski, et al., 1998). Likewise, t tests on the mean age and level of education were not significant (ps >.10).

Modification of the SASSI

The efficacy of each item on the SASSI-2 in identifying respondents with substance use disorders was evaluated with the development sample through discriminant analysis. As the maximum correct classification rate using the discriminant function equation was found to be 97%, this statistic served as the reference point for evaluating revisions to the SASSI decision rules. (Lazowski, Miller, Boye, & Miller, 1998). Six SASSI-2 items, including one research item that was not typically scored, did not discriminate between the criterion groups at the 95% level, and were excluded from consideration in the SASSI-3 decision rules. Five unscored research items, all of which eventually were incorporated into the SASSI-3 SYM scale, were found to be statistically significant in the discriminant analysis.

Following item pool modification on the basis of discriminant analysis, the development sample was used to revise scale cutoffs and decision rules. The goal at this juncture was to maximize accuracy, while maintaining a balance between sensitivity and specificity. (Lazowski, Miller, Boye, & Miller, 1998).

Results

Two-week test-retest stability coefficients obtained with a sample of 40 subjects ranged from .92 to 1.00. (Lazowski, Miller, Boye, & Miller, 1998). While the data showed high short-term stability coefficients, it was limited to the two week interval, and additional research will be necessary to determine the stability of the SASSI-3 over longer periods of time. The internal consistency coefficient was found to be .93. (Lazowski, et al., 1998).

Criterion validity analysis with the development sample revealed that the SASSI-3 achieved a 95% overall consistency between SASSI-3 classifications and clinical diagnoses of substance dependence, with a sensitivity rate of 96%, and specificity of 93%. With regard to the cross-validation sample, the findings were an overall correct classification rate of 97%, with sensitivity of 97% and specificity of 95%.

Overall accuracy in classification without consideration of the face valid scales produced a rate of 83%, whereas overall accuracy with the complete set of decision rules was 96%. This result revealed that the SASSI-3 decision rules, which incorporate the subtle, non-face valid scales, decreased the error rate from 17% to 4%, and eliminated 76% of the misclassifications that were produced when using the face-valid scales alone. (Lazowski, Miller, Boye, & Miller, 1998).

Additional comparisons revealed that the SASSI-3 was accurate in five diverse types of settings (addiction treatment, dual-diagnosis hospital, general psychiatric hospital, vocational rehabilitation program, and sex offender treatment program), and that there was no significant difference between males and females in terms of accuracy. (Lazowski, Miller, Boye, & Miller, 1998). Likewise, the respondents level of functioning, pursuant to DSM-IV Axis V, (APA, 1994), did not have any significant impact on the accuracy of the SASSI-3.

Overall, the results of the Lazowski, et al., (1998) study support the use of the SASSI-3 as a reliable and valid measurement tool in clinical assessment.

Conclusion

The abuse of and dependence on substances is widely recognized as a serious and pervasive problem. Given the profound consequences of substance use disorders, effective tools for early identification are necessary to interrupt the progression of their associated personal and social costs.

The SASSI was originally designed to be a brief, objective screening tool that would be simple to administer, score and interpret. Intended to identify chemically dependent individuals so that they might be further evaluated, the ability to make such classifications without regard to the defensiveness inherent in many instances of substance abuse or dependence was of critical importance.

Since it initial publication in 1988, the SASSI has been the subject of much fanfare,

as well as criticism. Through years of research, modifications and revisions have been made and evaluated in an effort to rectify any detected shortcomings.

Now in its third revision, the SASSI-3 continues to provide clinicians with a reliable and accurate screening tool that is comparable to other instruments, and which, despite its limitations, has made remarkable progress toward the developers original goals.

References

Allen, J. P. (1991). The interrelationship of alcoholism assessment and treatment. Alcohol Health Research World, 15 (3), 178-185.

Allen, J. P., Litten, R. Z., & Anton, R. (1992). Measures of alcohol consumption in perspective. In Litten, R. Z., & Allen, J.P. (Eds.), Measuring alcohol consumption: Psychosocial and biochemical methods. Totawa, NJ: Humana.

American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders (3rd ed., rev.). Washington, DC: Author.

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author.

Beattie, M. (1996a). Codependent no more: How to stop controlling others & start caring for yourself. Center City, MN: Hazelden Information Education.

Beattie, M. (1996b). The language of letting go. Center City, MN: Hazelden Information Education.

Butcher, J. N., Dahlstrom, W. G., Graham, J. R., Tellegen, A., & Kaemmer, B. (1989). Minnesota Multiphasic Personality Inventory-2 (MMPI-2): Manual for administration and scoring. Minneapolis: University of Minnesota Press.

Chobanian, N. H., & Rossler, J. I. (1995). Predicting the incidence of drug abuse in a population of pregnant DoD beneficiaries utilizing the Substance Abuse Subtle Screening Inventory Analysis & implications. Paper presented at the 34th Annual Meeting of the Armed Forces District, The American College of Obstetricians and Gynecologists, San Diego, CA. [Online]. Available: http://wpmc1.wpafb.af.mil/pages/obgyn-cp/chob.html.

Cooper, S. E., & Robinson, D. A. (1987). Use of the Substance Abuse Subtle Screening Inventory with a college population. College Health, 36, 180-184.

Ellsworth, R. B. (1978). Profile of Adaptation to Life Holistic Scale Manual. Roanoke, VA: Institute for Program Evaluation.

Ewing, J. A. (1984). Detecting alcoholism: The CAGE questionnaire. Journal of the American Medical Association, 252, 1905-1907.

Inciardi, J. A., & McBride, D. C. (1991). Treatment alternatives to street crimes. (DHHS Publication No. ADM 91-1749). Washington, DC: Department of Health & Human Services.

Lanyon, R. I. (1973). Psychological Screening Inventory: Manual. Goshen, NY: Research Psychologists Press.

Lazowski, L. E., Miller, F. G., Boye, M. W., & Miller, G. A. (1998). Efficacy of the Substance Abuce Subtle Screening Inventory-3 (SASSI-3) in identifying substance dependence disorders in clinical settings. Journal of Personality Assessment, 71 (1), 114-128.

MacAndrew