The ARRIVE Project is a randomized outcome trial designed to evaluate the use of brief MI to address alcohol and other substance use as a mediator of HIV risk behavior among people with serious mental illness living in the community. Alcohol and other substance use is associated with continued sexual risk behavior in a number of AIDS-vulnerable populations, including men and women dually-diagnosed with mental illness and substance abuse. The rationale behind ARRIVE is that by directly addressing substance use concurrently with HIV sexual risk behavior, we may be able to more effectively reduce HIV risk behavior in this population. In ARRIVE, brief alcohol and other drug treatment was used in tandem with a small-group HIV risk reduction intervention based on social learning principles. Participants in this condition attended two 60-minute individualized sessions based on Motivation Enhancement Therapy (MET) used in Project MATCH. This brief intervention was based on principles of motivational psychology, social learning theory, and the stages of change model. The 5-session small group HIV intervention used seven elements common to most effective HIV risk reduction interventions (Kelly, 1995): Understanding factors responsible for risk and behavior changes needed to reduce risk; accurate appraisal of personal level of HIV behavioral risk; development of self-efficacy for implementing risk reduction behavior changes; commitment to initiate personal action to reduce risk; acquisition, through modeling, practice and feedback, or correct condom use skills and sexual communication and negotiation skills; planning of strategies for implementing behavior changes and course of action if lapses occur; and self-reinforcement for behavior change and attempts to change behavior over time.

It was proposed that participants in the experimental condition would experience greater reductions in both HIV risk behavior and alcohol and other substance use than would participants experiencing either of two other comparison intervention conditions: a combination of brief substance use education and HIV skills; or a combination of substance use education and health education. All interventions were time-matched for comparison purposes. A total of 140 participants (94 men, 46 women) entered the study, and completed measures of HIV knowledge, intentions to change risk behaviors, sex and alcohol expectancies, condom attitudes, and the TimeLine FollowBack measure at baseline and 3-, 6-, 9-, and 12-month follow-up points. The TLFB assessed alcohol and other substance use and HIV risk behavior over the past 3 months.

Behavioral count data were analyzed with Generalized Estimating Equations (GEE), using SAS Proc GENMOD. Results demonstrate main effects for time across dependent variables for sexual risk, but no significant condition effects. However, participants in the HIV/Brief MI condition did show a trend for decreasing their drinking, approaching significance. PowerPoint slides will be used to present the method and data, and discussion points. These results, limitations of the study, and the implications for MI in addressing these risk behaviors will be discussed.