School of Public Health

Jennifer J. Carroll

Risk Perception Among HIV Serodiscordant Couples in East Africa Taking Oral PrEP for HIV Prevention


Levels of adherence varied across trials for pre-exposure prophylaxis (PrEP) for HIV prevention. One hypothesis for this inconsistency is that low perceived risk of HIV infection drove low adherence in trials where PrEP produced no reduction in risk. Using a mixed methods approach, we explored the level of perceived risk of incident HIV infection in the Partners PrEP Study, in which adherence was generally high. The Partners PrEP Study followed 4747 serodiscordant couples in Kenya and Uganda. A brief cross-sectional survey assessed perceived risk of HIV via questionnaire at 12 months after enrollment. Logistic regression was used to analyze the relationship between perceived risk and demographic variables, sexual behavior, and other objective measures of risk. 3226 couples from the Partners PrEP Study were included in this analysis. Perceived risk was low across the entire study cohort, with only 15.4% of participants reporting high or moderate perceived risk. Participants objectively assessed to be at higher risk for HIV were more likely to report high perceived risk (OR=1.60, 95%CI: 1.30-1.95, p<0.001), but still remained below 20%. In addition, this study also analyzed transcripts from in-depth interviews and focus groups, which were conducted with 68 individuals from 34 mutually-disclosed serodiscordant partnerships at the Thika, Kenya Partners PrEP Study site. Differences in the perceived risks and benefits of taking PrEP were identified according to participants’ serostatus and gender. These differences can be grouped into three major categories: gendered and unequal control over medical decision making in the home; the management of male sexual drives and the interference of traditional prevention strategies with the fulfillment of those drives; and culture-bound definitions of ‘women’s work. In summary, participants in the Partners PrEP Study reported generally low perceived risk of HIV but had high levels of adherence to PrEP. Perceptions of risk are likely determined by a multi-faceted social and psychological calculus that is too complex for Likert-scale questionnaires to capture in any meaningful way. In interviews, patients articulated complex concerns and interests related to PrEP, which reflected traditional gender roles and gendered power dynamics within the relationship. Successful delivery of PrEP as an HIV prevention strategy for serodiscordant couples should respond to these socio-cultural realities.