Experimental Analysis of HIV Risk Assessment Reactivity in South African Clinics

Randomized clinical trials of HIV prevention interventions typically include extensive baseline assessments of the risk behaviors that interventions target. Such pretests have the potential to direct participants' attention to the factors that put them at risk of HIV infection. Evidence from intervention study control groups, the motivational enhancement and attitude change literature, and our preliminary studies suggest that assessments that lead people to think about behaviors with potentially negative outcomes lead to motivation for risk behavior reduction. As a consequence, HIV intervention research using baseline assessments may not accurately predict the impact of a given HIV prevention intervention outside of the research context, in particular when interventions are evaluated among high-risk, isolated individuals. However, the actual effects as well as public-health consequences and mechanisms of HIV assessment reactivity have never been systematically tested.

The proposed assessment reactivity study will be conducted in South Africa, a country with high levels of HIV prevalence, incidence, and stigma. The Republic of South Africa's HIV/AIDS epidemic has emerged recently relative to other sub-Saharan countries, and yet 21.5% of South Africans aged 15-49 -- more than 1 in 5 people -- are infected with HIV (UNAIDS, 2004). Despite the common presence of HIV in South Africa, it remains largely a hidden disease and HIV-related stigma interferes with peoples' ability to protect themselves through open discussion or spontaneous reflection on personal risk. Under these conditions, it is reasonable to expect a sizable effect from directing people's attention to their HIV relevant behaviors, an effect that could be incorrectly ascribed to brief interventions used in low-resource settings. The conditions in South Africa, therefore, heighten the clinical and public health significance of the assessment reactivity issue in that (a) it is vital to understand how reactivity may affect the results of behavioral intervention trials, and (b) it may be possible to capitalize on assessment reactivity through brief, assessment-based interventions.

This study will allow us to achieve our primary aim:

1. To separate the effect of a brief HIV-risk reduction intervention from the potential effects of two types of baseline assessment commonly applied in HIV intervention research: a detailed, calendar-based assessment and a general frequency assessment. The detailed calendar-based assessment will potentially lead participants to observe themselves in specific risk-behavior situations, providing them with the opportunity to reassess the risk involved in those situations in light of their health goals and thus motivating them for behavior change. The general frequency assessment will merely prompt participants to estimate the overall frequency with which they engaged in several HIV-related behaviors. The effect of these assessments on participants' sexual behavior will be examined alone and in combination with the HIV-risk reduction intervention.

In addition, this project has two important secondary aims:

1. To generate urgently needed additional efficacy data for a promising HIV risk reduction intervention culturally tailored for use and found to be effective in South Africa
2. To determine potential HIV preventive effects of different types of HIV risk behavior assessments on risk behavior reduction in South Africa for future intervention development

To achieve the above aims, the investigators will:

1. Recruit 1500 patients (50% women), age 18-44, from the Spencer Rd. STD Clinic, Cape Town, South Africa.
2. Randomly assign participants within gender to one of 6 study arms in a modified Solomon Four-Group (SFG) design. The 2 (intervention: standard of care vs. an efficacious theory based HIV prevention intervention) x 3 (baseline face-to-face interview: none vs. basic single-item frequency questions vs. detailed partner-by-partner calendar-based interview) factorial design is depicted below: