APPLICATION OF COMPUTER TECHNOLOGY TO ASSESSMENT OF RISK BEHAVIORS AMONG PARTICIPANTS IN THE VACCINE PREPAREDNESS STUDY OF HIVNET

(HIVNET Protocol 005)

WHAT DID WE HOPE TO LEARN FROM THIS STUDY? 

This study tested a new approach to studying behavior--Audio-Computer Assisted Self-Interview technology (A-CASI). A-CASI allows participants to answer potentially embarrassing or sensitive questions by interacting with a computer rather than an interviewer. Participants read the questions on a computer screen and/or listen to questions spoken by a recorded human voice. They answer questions using the computer keyboard (other systems not being studied here use a touch screen). The spoken version of the interview makes this technology usable by people who do not read well. A-CASI also has the potential advantage of consistency in interviews and questionnaires because it eliminates any variations in the way different interviewers handle the process. 

This study looked at whether the amount of risky behavior reported during computer interviews is different from the results of standard face-to-face interviews about sexual behavior and drug use. The study also compared the cost of A-CASI to that of the standard face-to-face interview.

WHO PARTICIPATED IN THE STUDY?

The study involved 2,200 participants who were also enrolled in the Vaccine Preparedness Study (VPS). Participants included injection drug users and men who have sex with men.

WHERE WAS THE STUDY CONDUCTED? 

• Howard Brown Health Center, Chicago, IL
• University of Pennsylvania, Philadelphia, PA
• New York Blood Center, New York, NY
• University of Washington, Seattle, WA
• New York University, New York, NY
• San Francisco AIDS Office, San Francisco, CA

HOW WAS THE STUDY CONDUCTED?

 VPS participants were randomly assigned to either: 1) complete the A-CASI version of the VPS questionnaire, or 2) complete the standard, face-to-face VPS interview. A total of 600 injection drug users and men who have sex with men from six VPS sites were assigned to the A-CASI arm of the study. Participants who came in for their VPS Month 12 Visit after enrollment began were randomized at their VPS Month 12 Visit, and remained in the same group at the Month 18 visit. Participants who had completed their VPS Month 12 visit when study enrollment began were randomly assigned at their VPS Month 18 Visit.

WHAT WAS THE TIME FRAME FOR THE STUDY?

 Enrollment for the study began in September 1996, and the data collection portion of the study was completed in August 1997. The two methods are now being compared to see whether the type of interview (A-CASI or face-to-face) had any effect on participants' reporting of risk behaviors. Preliminary analysis has shown that A-CASI was accepted by participants, and A-CASI is being incorporated into plans for future studies that include interviews about risk behavior. A manuscript is in development.