3%). Based on this finding, it may seem appropriate to target only these men for HIV prevention trials. However, these men accounted for only 4.8% of total HIM follow-up. Thus, the target population is small and recruiting

sufficient numbers for an HIV prevention trial would probably be difficult. By adding the two next highest risk behaviours (receptive UAI with casual partners and reporting use of oral erectile dysfunction medication and methamphetamines), Vemurafenib datasheet while maintaining an HIV incidence of 2.7 per 100 PY per year, the size of the population at risk was greatly increased to 24% of the cohort. With an HIV incidence of 2.7 per 100 PY, HIV prevention trials among this group of men may be feasible. The sample size necessary for each study arm in a randomized controlled HIV prevention trial to show 50% efficacy of an intervention after 1 year of follow-up would be 1853, assuming a significance level of 95% and a power of 80%. If the entire HIM population was recruited, with an incidence of 0.78 per 100 PY, the sample size would increase to over 6500 per study arm. Men at high BIBF 1120 ic50 risk of HIV in the HIM study were more willing to participate in HIV prevention trials of vaccines and ARVs. Although not quite significant, there was a trend towards greater willingness to participate in rectal microbicide trials among men at higher risk of

HIV infection (P=0.056) when only men who had definite opinions on participation were included. This association GNAT2 between an increased risk of HIV infection and willingness to participate in HIV prevention

trials has been consistently identified in MSM who are potential trial participants, both in Australia [37] and in other countries [34,35,38–41]. The combination of high HIV incidence and increased willingness to participate in trials further indicates the suitability of such a population for prevention trials. This study had the strength of being a large-scale prospective cohort study and was primarily community-based, with only 4% of participants recruited from clinics. Although not necessarily representative of all Australian gay men, a wide variety of recruitment strategies were used to reach a broad sample of the homosexual community. Detailed information on UAI behaviour was collected, which allowed the differentiation of partner- and position-specific practices from all UAI acts and the creation of precise definitions of risk variables. The prospective biannual collection of behavioural data minimized recall bias. There were several limitations in this study. The question on willingness to participate in trials using ARVs to prevent HIV infection potentially included men’s attitudes to PREP and/or NPEP trials. However, as the intervention is the same (oral antiviral therapy), it is feasible that men’s attitudes towards participation in PREP and NPEP trials would be similar.