Voucher-Based Incentives in a Prevention Setting

Homelessness is a significant problem plaguing American cities and homeless substance abusers face increased risks. Homeless, substance abusing men who have sex with men (MSM) suffer 80% HIV seroprevalence, engage in the sex trades and resist treatment for substance abuse. For almost 30 years, the Van Ness Recovery House (VNRH) has been providing treatment services to homeless substance-abusing MSMs in Hollywood, CA. In 1994, VNRH began providing community-based HIV prevention services to non-treatment seeking homeless substance-abusing MSMs through the VNPD, its prevention division. Contingency management interventions providing positive incentives for behavior change may be particularly well suited for this disenfranchised, high-risk cohort. Specifically, voucher-based incentive therapies (VBRT) may be particularly effective since they have established potency for increasing prosocial behaviors that successfully compete with taking drugs and for reducing drug use. A randomized, controlled trial will assign 144 non-treatment seeking homeless substance-abusing MSMs at VNPD to either VBRT or control groups (72/group) for 24 weeks, with planned follow up at 7, 9 and 12 months from randomization. The VBRT group will earn vouchers in exchange for completing prosocial and healthy behaviors and/or submitting drug-negative urine and alcohol-negative breath samples. Vouchers will be redeemable for goods located in an onsite voucher store. The control group will receive feedback regarding behaviors performed and urinalysis and breath alcohol tests, but will not receive voucher points for these behaviors. We will assess the efficacy of the VBRT intervention for increasing prosocial and healthy behavior and reducing substance abuse among these non-treatment seeking homeless substance-abusing MSMs receiving services at VNPD. We will also assess the impact of VBRT on other measures of therapeutic change consistent with a harm reduction approach, including reduction of psychiatric symptoms, decreased injection drug use and high-risk sexual behavior, increased participation in prevention programming, improvement in different domains of overall functioning (medical/social/vocational), and increased readiness to change. Additionally, we will examine whether baseline participant characteristics predict VBRT outcomes. Applying VBRT to non-treatment seeking homeless substance-abusing MSMs as well as integrating VBRT into a well-established prevention program with a harm reduction philosophy are both highly innovative. Our reinforcement contingencies will also shape behavioral steps towards more complex behaviors, employing a basic operant tenet that has not received sufficient attention. Overall, our proposed research has the potential to have a significant impact on the public health of the Los Angeles homeless, gay and bisexual community. If VBRT is efficacious for motivating non-treatment seeking homeless substance-abusing MSMs in a community-based prevention program to increase prosocial and healthy behaviors and decrease drug/alcohol use, established prevention programs may modify their approaches to include contingency management, and use it to address the staggering public health problems HGMSAs face on a daily basis.

Inclusion criteria

• Self-identified, non-treatment seeking, men who have sex with men (MSM) at least 18 years of age and eligible to participate in the VNPD Men's Program. To be eligible to participate in the VNPD Men's Program, a person must report that they have sex with men (or men and women) and actively use drugs or alcohol.
• DSM-IV diagnosis of substance dependence as determined by structured clinical interview for DSM-IV (SCID) on at least one of the following substances: alcohol, marijuana, methamphetamine, cocaine, crack, MDMA (ecstasy), opiates, or dissociative anesthetics (PCP, ketamine). These substances were chosen because they are the ones commonly used by individuals contacted by the VNPD's street outreach workers.
• Ability to understand and provide written informed consent. Also see section D6.1 for consent procedures used with vulnerable populations.
• Able to provide sufficient locator information for follow-up and/or known to VNPD outreach workers.

Exclusion criteria

Psychiatric health issues requiring a higher level of care.