Understanding drug abusers' use of drug abuse treatment and other health care services has ban designated a scientific research priority of the National Institute on Drug Abuse. The primary goal of the proposed five- year project is to produce a natural history of crack-cocaine users which describes and analyzes-their use of health services. Such data can have direct utility for the planning, development, and provision of health care and intervention services. The main analytic goals of this project will be to determine the factors that predict both perceived health care needs and the use of health services, particularly with respect to women and minorities. Using targeted sampling methods proven effective during our Cooperative Agreement for AIDS Community-Based Outreach/Intervention Research (COOP) project, 420 never-Injecting, crack-cocaine abusers in the Dayton/Montgomery County, Ohio, community will be assessed and followed-up at six-month intervals for three years. In addition to background and drug abuse data collected in natural history studies, qualitative and quantitative methods will be used to obtain information on crack users' perceived needs for drug abuse treatment and other health care services as well as their actual use of these services. Subjects will be recruited using a targeted sampling methodology to assure representation of crack users in the sampling area with respect to gender, ethnicity, and length of crack use. This will enable us to make ethnic and gender comparisons of perceived health service needs, barriers to access, and use, The initial assessment and follow-ups will include such standard instrumentations as the Addiction Severity Index (ASl), a modified version of the Risk Behavior Assessment (RBA), the Beck Depression Inventory (BDl), the Symptom Check List (SCL-9O-R), as well as schedules to be developed by project staff and consultants that will assess domains such as health beliefs and behavioral intentions. Retrospective data will be collected using a modified version of the life chart technique developed for use in natural history studies of drug abuse. The Cox proportional hazards model will be used to analyze the likelihood and correlates/determinants of drug abuse treatment utilization and the cessation of crack use and relapse. A Poisson regression will be used to analyze prospective data on the use of general/physical health services and mental health services. A discrete-time hazard model in conjunction with an ordered probability logit model will be used to analyze changes in drug use and other attendant behaviors. Structural equation modeling (SEM) will be used for analyses involving latent variables, including the perceived need for health services.
Showing the most recent 10 out of 14 publications
