When and how sexual risk behavior assessments influence HIV risk behaviors and HIV prevention intervention outcomes: An examination of the impact of assessment type and risk awareness

BACKGROUND. Randomized controlled trials (RCT) of HIV prevention interventions include risk behavior assessments to obtain baseline risk estimates, which can affect intervention outcomes by influencing risk awareness. Understanding this influence is necessary to interpret evidence from intervention RCTs. We examined the effects of different risk behavior assessments on sexual risk and HIV intervention outcomes in terms of respondents' risk perceptions. We expected a less personalized (frequency-based) assessment to reduce risk and increase intervention efficacy among risk aware respondents and a more personalized (event-based) assessment to do so among all. METHOD: We conducted a 2 X 2, risk-reduction counseling by assessment RCT with a three-month follow-up in an STI clinic serving youth in South Africa (N= 1723). Participants in the assessment groups received either a frequency-based or an event-based assessment. We examined main and interactive effects of assessments, counseling, and baseline risk perceptions on condom use and partner number. RESULTS: Relative to the not assessed or counseled group, youth with high risk perceptions reported more condom use and fewer partners when they received the assessments (ds = 0.47, -0.26), the counseling (ds = 0.55, -0.32), or both (ds = 0.60, -0.37). Youth with low risk perceptions reported improved condom use when they received the counseling alone (ds = 0.43). Patterns replicated regardless of assessment type. CONCLUSIONS: Risk behavior assessments may reinforce respondents' beliefs about risk, reducing HIV risk among individuals who are risk aware but hindering intervention efforts among those who are not. Results have implications for RCTs and intervention dissemination.