What changed?

Studies reported on a whole variety of outcomes, none exclusively, and no study included all the outcomes specified by the CDC: unprotected sex; serodiscordant unprotected sex; number of partners; condom use; STI incidence; HIV incidence; and (with IDUs) needle sharing and (with youth) age at sexual debut.

Twelve studies reported a significantly reduced frequency of unprotected sex in participants, and all three studies targeting MSM reported significantly reduced unprotected anal or receptive anal intercourse. Eight studies reported increased condom use. Four out of five studies with drug users reported significantly reduced sexual-risk behaviour and three reported reduced injecting or needle-sharing.

Four reported a significantly reduced incidence of STIs in the twelve months after the study and three reported a significant reduction in the number of partners.

However, only one study reported on the bottom line of HIV prevention, namely HIV incidence. This, the EXPLORE study, was able to do so because it was by far the largest study, recruiting 4295 gay men or 4.5 times as many as the next-largest study. It randomised the subjects into normal three-monthly health monitoring at clinic visits versus monitoring plus a series of one-to-one coaching interviews with a counsellor. It produced a modest (18%) reduction in HIV incidence which was not statistically significant, though larger reductions took place in unprotected anal sex and STI acquisition.

There are significant gaps in the CDC’s review of what works in HIV-prevention programmes. Firstly, and most significantly, the CDC only reviews US-based programmes and what works in the American cultural context may not work in other countries. Secondly,  most intervention studies only take place in one setting and often at only one site, and it is uncertain how generalisable they are either in terms of varying the setting or the target group. This has not stopped NHS funders in the UK from using the CDC’s findings as evidence for which programmes they should fund.

Thirdly, significant groups of people have not been addressed by many of these interventions. These include minority MSM, particularly young minority MSM, transgendered persons, intravenous drug users with HIV, and rural populations (all studies took place in cities with the exception of one study among drug users in rural Puerto Rico). Fourthly, this particular CDC review looked at very few community-based interventions and excluded media and structural interventions.

HIV-prevention programmes generate, as the researchers comment, “a vast and heterogeneous literature” in a field where there are few protocols or best-practice guidelines to refer to. The CDC is at least attempting to inject some evidence and theoretical rigour into an inherently ‘fuzzy’ area of HIV research.