This treatment evaluation study will assess and followup approximately 500 male admissions for cocaine abuse to the West Los Angeles Veterans Administration Substance Abuse Treatment Program. The sample will be accumulated over an 18 month period. Patients will be assigned to immediate inpatient treatment if they exhibit medical complications, psychosis, or severe social dysfunction (approximately 25%) or to the comparative modalities group (approximately 75%). Patients in this latter group will be randomly assigned to one of three types of treatment: 28 day inpatient care followed by outpatient aftercare, outpatient care only; or referral to self-help/mutual support groups. All inpatients (immediate assignment to inpatient or random assignment to inpatient) will have extensive medical, psychiatric and psychosocial assessments. The outpatient and self-help random assignment conditions will have an abbreviated assessment. A major evaluation instrument will be a retrospective longitudinal natural history questionnaire that will obtain data from 12 months before the patients' first regular or dysfunctional involvement with illicit drugs to the time of admission for treatment. The history will be updated at three followup points: six months, one year, and two years post admission. This natural history approach is suitable for time series evaluation with comparisons possible across the four conditions. Treatment effectiveness will be assessed, as well as how the various intervention conditions differentially affect the course of cocaine and other drug abuse and their associated behaviors. The major variables of interest include level of drug use, methods of funding drug purchases, criminal involvement, employment, interpersonal and social functioning, legal status, alcohol use, and use of health resources. Statistical analyses will assess the role of sociocultural and environmental variables in the maintenance, cessation, relapse or """"""""maturing out"""""""" of drug abuse. The differential outcomes of the four conditions may allow a determination of those abuser types who respond better to each modality. For the purposes of this research statistical procedures will be applied such as ANOVA or ANCOVA with repeated measures. Box-Jenkins ARIMA time series, log linear modeling, and causal modeling (LISREL).
Showing the most recent 10 out of 21 publications
