Computer-based Cognitive Behavioral Therapy for Risky Behaviors in Opioid Dependent Patients

We plan to conduct a randomized pilot trial evaluating the feasibility and promise of computerized CBT as a single module targeting risky sexual behavior in a HIV/Hepatitis C virus(HCV) high risk population. In the context of a drug treatment program, we will randomize 60 methadone maintenance patients to either standard treatment as usual at the program (TAU) or TAU plus the HIV/HCV/STD risk reduction module of CBT4CBT("Stay Safe"). Participants will be assessed at baseline, one month and three month time points. The primary outcome measures will be knowledge regarding risk of infection and transmission of HIV/HCV/STD,self report assessment of risky sexual practices, retention in the study and client satisfaction with the module. Our specific aims are as follows:

1. To determine the feasibility of integrating CBT4CBT/"Stay Safe" into this setting. Feasibility will be determined by completion of assessments at 1 and 3 month time points (retention) and satisfaction ratings by participants randomized to the computer module.
2. Compare knowledge of HIV/HCV/STD transmission and its prevention before and after intervention. This will be assessed by percent correct answers on a quiz administered at each assessment.
3. To evaluate changes in drug and sex risk behaviors by treatment. The primary outcome measure will be number of unprotected sexual encounters on self-assessment instruments for risky sexual behaviors.

We hypothesize that the single session of computer-based CBT will result in reduced risky sexual behaviors and increased knowledge of HIV/HCV/STD transmission with differences in retention by intervention. A successful outcome for this pilot project would result in an effect size of 0.30 or more on outcomes of risk behavior. This intervention will be compared to treatment as usual at the participating site, which currently consists of two group educational sessions on safety regarding risks of HIV/AIDS and hepatitis.

Preliminary Data- The case for computer-based behavioral treatments Computer-assisted therapies offer a relatively novel approach to the dissemination of Evidence Based Therapies (EBT) for behavioral interventions. Existing evidence points to their efficacy and cost-effectiveness. Computer-based interventions offer a number of attractive characteristics for use in primary care and substance treatment settings. Their low cost coupled with high consistency, accessibility and standardization are some associated advantages. Ease of implementation will be critical to reaching patients in remote and rural areas or in small medical offices as primary care providers become a mainstay of substance abuse treatment. Risky sexual behaviors can be targeted without the potential negative associations of face to face behavioral interventions for this highly sensitive domain and thus may be ideally modulated with computer-based therapies. Equivalency to counselor-based education for learning new health behaviors, cost-effectiveness, patient satisfaction and accessibility for illiterate patients are other advantages shown with computer-based therapy.

Computer-based CBT: Carroll et al. at Yale have developed an effective computer-based version of CBT, called "CBT4CBT" and have demonstrated its efficacy in reducing drug use and building coping skills.(38) The program makes extensive use of 'movies' as teaching tools: In each module, the user watches an individual confront a difficult situation relevant to that module's topic; after teaching the key skill through a variety of strategies, the 'movie' repeats but has a different ending because the characters implement the targeted coping skills. In developing CBT4CBT, we sought to develop an engaging version of CBT that could take advantage of the capacity of computer-based learning to convey information via a wide range of media (e.g., text, video, graphics, audio instruction, interactive exercises). The CBT4CBT program is highly user-friendly, requiring no previous experience with computers and minimal use of text-based material (i.e., minimal reading is required), and is highly interactive. In particular, we capitalize on the use of videotaped examples to allow users to actually see examples of individuals utilizing skills and strategies in a range of realistic situations. Viewers are able to watch real-life challenges acted out and safe behaviors modeled, while addressing negative or detrimental thought processes that predispose to unhealthy or risky behavior choices (e.g. making a decision to have sex without a condom). Viewer knowledge base is targeted by didactic portions and role playing is modeled by the actors. The module includes the opportunity for the client to print out and perform 'homework' worksheets, shown to be predictive for successful outcomes in CBT. The module allows for considerable control by the viewer, who can choose the speed of progression through the screens and has the capability to go back to previous screens for reviewing.

Single CBT module application: Our original feasibility, efficacy, durability and cost-effectiveness studies for addictions treatment employed the full version of CBT4CBT, comprised of 7 modules. One independent module focuses entirely on targeting risky sexual and drug use practices (titled "Stay Safe"), but has not been evaluated as an independent module for effects on risk reduction. The "Stay Safe" module was developed by Dr. Kathleen Carroll and her team at Yale in consultation with the Connecticut AIDS Education and Training Center director (Karina Danvers) and others. Its development involved the input of the CBT experts (Dr. Carroll, Dr. Michael Copenhaver), infectious disease specialists and substance abuse treatment patients who volunteered to review the module. Like the substance abuse targeted sections, the script was written to be easily recognizable across many socioeconomic and cultural groups. Professional actors were hired to play the parts of substance users in two separate high risk situations, one sexual and one Intravenous drug use (IDU). Skills are taught with multimedia presentation allowing the user to direct the pace of the module, as well as to view the consequences of various choices during the risky situation. The intervention for the proposed pilot study is this single targeted module, "Stay Safe". The module can be completed on virtually any computer, and thus is ideal for implementation in this, and a wide range of settings. Its expense for clinical use will also be reduced as a result of its brevity. These characteristics will make it easily disseminated, with "real world" applications for settings of substance use treatment and primary care medicine.

Overview: In order to maximize scientific yield from this project, we intentionally designed the study to parallel the landmark Calsyn(5) and Tross(6) randomized trials evaluating effects of the 5-session group CDC behavioral approach for risky sex practices reduction in substance abuse treatment programs. We seek to compare as closely as possible the variable of therapy type (computerized CBT in our study versus group therapy for safer sex skills building in theirs), so we sought to minimize the differences between outcome measures and methods. Use of parallels in design and outcome measures will allow us to 'benchmark' outcomes to reported changes in those studies.

Study Population and Sample: Hartford Dispensary (Hartford, CT) will serve as the site. Drug treatment and medical services are offered on an outpatient basis there. This study will enroll clients undergoing IDU treatment with methadone at the site. These clients come in regularly for scheduled maintenance medication and clinical monitoring as well as standard counseling. Many (roughly 55%) of these patients are already HCV and/or HIV positive and therefore pose a significant risk for the transmission of the virus through routes other than IDU.

Recruitment: Clients will be identified through self-presentation response to advertisements and fliers in the clinic. Drug counseling providers may also ask clients if they are interested in participating in the study. A sample size of 60 (30 patients per group) is feasible and would be sufficient to detect a large effect size on risky sexual behavior of [Cohen's d= 0.3, (alpha.05, power0.8)]. This effect size is smaller than that reported for behavioral studies in the literature, but should be sufficient to evaluate feasibility and promise of the proposed intervention in this pilot.