Background: Efficacy studies of HIV prevention require high HIV incidence.  For this reason, many HIV prevention studies are being conducted in high incidence developing country settings. We examined HIV incidence in homosexual men in a developed country setting to assess the suitability of this population for HIV prevention trials, and to assess the use of questionnaire-based risk criteria for the definition of high incidence groups.

Methods: The HIM study is a community-based cohort study of HIV-negative gay men in Sydney, Australia, and commenced in 2001. All participants undergo annual HIV testing and interviews twice a year. HIV infection is calculated using the person-year (PY) method. To define a group at high risk of HIV from questionnaire-based reported criteria, subgroups of cohort members with an annual incidence of at least 2% were defined.  The highest incidence group was first chosen. Then, incidence in the remaining high incidence groups was re-calculated among those who did not report the first risk. The next highest incidence group was then chosen. This process was repeated until annual incidence in all remaining sub-groups was less than 2%.  

Results: A total of 1,427 participants were enrolled. By the end of 2004, there had been 24 HIV seroconversions and 2,721 PY of follow up. HIV incidence was 0.88 per 100 PY. Based on behavioural self-report, subgroups of cohort members were defined. These subgroups were then ranked by incidence as below.

Risk group                                                                    Incidence          % of total

                                                                                                Per 100py        Person-years

1.      Unprotected anal intercourse (UAI) with HIV+ partner       6.7                   4.4%

2.      HIV positive regular partner                                                3.5                   3.5%

3.      Erection pills more than monthly                                          3.2                   3.2%

4.      Receptive UAI with a casual partner                                   3.0                   12%

5.      Recent anal sexually transmitted infection (STI)                    3.0                   9.8%

6.      Previous post exposure prophylaxis                                     2.9                   11%

7.      Partners last 6 months         > 50                                         2.1                   5.3%

A group at very high risk comprising 13% of the total PY was defined with an incidence of HIV of 4.0%.   This group comprised men who reported UAI with a known HIV positive partner or having a recent anal STI.

Discussion: In Australian homosexual men, although overall HIV incidence is less than 1%, groups of individuals with very high HIV incidence can be identified by self-reported criteria. These criteria can be used as trial eligibility criteria, making efficacy trials of methods of HIV prevention feasible, even in a population with a relatively low overall HIV incidence.