ABSTRACT
The Million Clinical Multiaxial Inventory (MCMI versions I, II, and III) includes a scale to assess drug use problems, Scale T-Drug Dependence. Detailed drug use data from a sample of 659 known drug users along with MCMI-II results were examined to determine the operating characteristics of the MCMI-II drug dependence scale. Operating characteristics, sensitivity, specificity, positive predictive power, negative predictive power, and overall diagnostic power were calculated for base rate cutoffs and for the number of prototypic items endorsed to determine the diagnostic efficiency of Scale T-Drug Dependence in identifying regular drug users. Prototypic item cutoffs provided higher levels of diagnostic and positive predictive power than did the standard base rate cutoffs.
Subject(s)
Personality Assessment/statistics & numerical data , Personality Inventory/statistics & numerical data , Substance-Related Disorders/psychology , Adult , District of Columbia , Female , Humans , Interview, Psychological , Male , Middle Aged , Personality Disorders/psychology , Personality Disorders/rehabilitation , Predictive Value of Tests , Psychometrics , Sensitivity and Specificity , Substance-Related Disorders/rehabilitationABSTRACT
This study was conducted to investigate the relationship between the indicators of psychiatric disorders of individuals and their choice of either cocaine or heroin, drugs that differ markedly in their pharmacological effects. Cocaine acts as an intense stimulant, and heroin has profound sedative effects. This investigation examined the relationship between preference for heroin or cocaine and indicators of psychiatric impairment. Data from 282 subjects were grouped according to drug of choice and analyzed. Ninety-three percent of these subjects were African-American, 32% were female, and the average age was 34. Univariate and multivariate statistical analyses, such as discriminant analyses, were used to determine group differences. The results are evaluated and interpreted in relation to both the current empirical findings and to the hypotheses and theories postulated as a result of earlier clinical observations on drug of choice and psychopathology. Discriminant analysis yielded an overall correct classification rate of 75%. The discriminant function suggests that members in the cocaine drug of choice group as contrasted with members in the heroin preference group can be characterized as more socially inhibited and more self-defeating after adjusting for differences in age, duration of use of illicit substances, and marital status. Those who favored cocaine as contrasted with those who favored heroin were more likely to have never married, be younger, and have used illicit substances for a shorter period of time.
