Measuring the Effect of HIV Behavioral Interventions: Individual Behavior Change Success vs. HIV Acquisition Risk

• Activity Number: 407
• Type: Contributed
• Date/Time: Tuesday, August 5, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Lillian Lin*+ and Craig B. Borkowf
• Companies: CDC and CDC
• Keywords: HIV behavior change ; HIV risk ; men who have sex with men

Abstract

HIV behavioral interventions seek to reduce HIV acquisition risk, that is, unprotected sex or injection drug use with HIV-positive persons. This risk cannot be measured directly. Commonly used proxy measures include the proportion of sex acts using a condom and the number of sex partners during a specified time period. Standard measures of the effectiveness of a behavioral intervention focus on relative changes, e.g., treating someone who reduces from 50 to 25 sex partners the same as someone who reduces from two partners to one. However, the former has negligibly reduced his HIV acquisition risk whereas the latter has substantially reduced his, especially if his partner is HIV-negative. Moreover, the apparent success of a behavioral intervention may be driven by random changes among the riskiest individuals (the outliers). We deconstruct measures of sex behavior change that are commonly used to evaluate HIV prevention interventions from an infection risk perspective.