Social scientific and public health literature on National Institutes of Health-funded HIV behavioural prevention science often assumes that this body of work has a strong biomedical epistemological orientation. We explore this assumption by conducting a systematic content analysis of all NIH-funded HIV behavioural prevention grants for men who have sex with men between 1989 and 2012. We find that while intervention research strongly favours a biomedical orientation, research into the antecedents of HIV risk practices favours a sociological, interpretive and structural orientation. Thus, with respect to NIH-funded HIV prevention science, there exists a major disjunct in the guiding epistemological orientations of how scientists understand HIV risk, on the one hand, and how they engineer behaviour change in behavioural interventions, on the other. Building on the extant literature, we suggest that the cause of this disjunct is probably attributable not to an NIH-wide positivist orientation, but to the specific standards of evidence used to adjudicate HIV intervention grant awards, including randomised controlled trials and other quantitative measures of intervention efficacy.
