Date of Graduation


Document Type



A significant proportion of clients utilizing substance abuse programs fail to successfully complete a course of treatment. High attrition rates represent a large number of clients who do not receive the full benefit of treatment. Both clients and treatment programs are adversely affected by premature termination. The present research focused on prediction of discharge status from residential substance abuse rehabilitation. A discriminant function analysis resulted in a prediction equation which incorporated nine sociodemographic variables. This equation correctly classified the discharge status of 67% of the prediction sample of 252 subjects. The equation was cross validated in a stratified random sample of 95 subjects. Upon cross validation, the equation correctly classified 71% of the overall sample and 91% of the subjects who had left treatment prematurely. An additional attempt to enhance the efficiency of the discriminant function analysis with MMPI-2 measures of personality, measures of self-efficacy for abstinence, and a measure of readiness for change based in transtheoretical therapy failed. Furthermore, differences on these measures were not identified between program completers and premature terminators. It was concluded that sociodemographic variables can be used to efficiently discriminate program completers from premature terminators. It is believed that the present research finding may enhance the process of treatment matching and substance abuse program design.