Relationship of patient variables and prognosis ratings in an alcoholism treatment program

This study investigated the relationship of selected psychosocial variables, based on cognitive social learning theory, and clinicians' prognoses for patients who had completed treatment for alcoholism. Within 5 days of entering treatment, 145 patients at a Veterans Administration Medical Center alcohol treatment unit were administered the Michigan Alcoholism Screening Test (MAST), the Drug Abuse Screening Test (DAST), the Buss-Durkee Anger Inventory (BDAI), and the Tennessee Self-concept Scale (TSCS). A structured interview was used to obtain sociodemographic information. The test scores and sociodemographic information were the independent variables. Six clinical staff members from the Alcohol Treatment Unit judged the patients' prognoses using a 5-point scale. These ratings served as the dependent variable. Multiple regression and analysis of variance procedures were used to analyze the data. The one-variable regression model indicated that the DAST scores were the best predictor of clinician's ratings of patients' prognoses. There was a slight, but statistically significant, relationship between DAST scores and prognoses, indicating that clinicians tended to give lower prognosis ratings to those patients who had a history of polysubstance abuse than to those patients who did not. The results of the four separate one-way ANOVAs suggested that marital history and employment history could be related to the patient's prognosis ratings. The limitations and implications of the findings are discussed. Suggestions for future research are given.