Article Text

Download PDFPDF

Testing for HIV and sexually transmissible infections within a mainly online sample of gay men who engage in group sex
  1. G P Prestage1,2,
  2. J Hudson3,
  3. F Jin1,
  4. N Corrigan4,
  5. P Martin5,
  6. A E Grulich1,
  7. D McInnes6
  1. 1
    National Centre in HIV Epidemiology and Clinical Research, University of New South Wales, Sydney, Australia
  2. 2
    Australian Research Centre in Sex Health and Society, La Trobe University, Melbourne, Australia
  3. 3
    Faculty of Medicine, University of Sydney, Sydney, Australia
  4. 4
    AIDS Council of NSW, Sydney, Australia
  5. 5
    Queensland Association for Health Communities, Brisbane, Australia
  6. 6
    School of Communications, University of Western Sydney, Sydney, Australia
  1. Dr G Prestage, National Centre in HIV, Epidemiology and Clinical Research, 376 Victoria Street, Darlinghurst, NSW 2010, Australia; gprestage{at}


Introduction: Group sex among gay men has been associated with other HIV risk behaviours. Gay men who engage in group sex may be at increased risk of infection with HIV and other sexually transmitted infections (STI).

Methods: The Three or More Study (TOMS) of group sex among gay men utilised an anonymous, self-completed survey about participants’ most recent occasion of group sex with other men and in-depth interviews with a small number of these survey participants. The 436 men who reported having engaged in group sex within the previous month were included in these analyses.

Results: Among 436 men who engaged in group sex within the previous month, 32.5% reported unprotected anal intercourse (UAI) with non-regular, mostly HIV non-seroconcordant partners at this recent group sex encounter (GSE) and the majority reported other sex practices that are risk factors for STI other than HIV. Over one-third reported having been tested for HIV or other STI since their last GSE; those who had engaged in UAI at the GSE were more likely to have been tested (p = 0.008). Men who had a doctor with whom they were able to discuss their group sex activities had received a broader range of STI tests (p = 0.003).

Conclusion: Sex practices that risk the transmission of STI were common within this high-risk sample, whereas awareness of risk and the need for testing was high but not universal. Frank discussion with doctors of patients’ group sex behaviour also enhanced decisions about adequate testing. Gay men in group sex networks are an appropriate priority for sexual health screening.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Increasing rates of sexually transmissible infections (STI) among homosexual men have been reported worldwide over the past decade,1 whereas homosexual transmission of HIV has been increasing.25 Screening guidelines in Australia suggest that homosexually active men should be tested at least annually for STI.5 6 Previously, such guidelines usually only recommended anal testing in those reporting receptive anal intercourse,7 8 but this was recently amended in Australia when it was identified that other non-intercourse anal practices were also risk factors for some STI.9 Australian guidelines now recommend at least annual screening for all homosexually active men, to include pharyngeal swab, anal swab, first catch urine and serology for HIV and syphilis.

Approximately 90% of Australian gay men report having ever been tested for HIV and approximately two-thirds report having been tested in the previous year, a rate that has changed little in recent years.10 The proportion of men tested for other STI in the previous year is almost as high,11 probably because much of this testing is routine sexual health screening. These testing rates are relatively high compared with rates found among homosexual men in many other countries.1214 At least with regard to HIV testing in Australia, there is a strong relationship between risk behaviour and recent testing. Men who report having engaged in unprotected anal intercourse (UAI) with non-regular partners are also more likely to have recently been tested for HIV.13

Group sex among gay men often involves men who are highly sexually active.1518 Gay men who engage in group sex are likely to be at high risk of infection by, and transmission of, HIV and other STI, and their group sex activities appear to be associated with other markers of risk.19 In this paper, we investigated recent sexual health testing among gay men who had engaged in group sex within the month before interview and describe the contexts of such testing.


Homosexually active men in Australia who engage in group sex were invited to complete a survey instrument either at gay community sex-on-premises venues or sex party events in paper and pencil format or in an electronic format online. Online participants were also recruited through web-based advertising on gay commercial websites and through e-mail networks of men who participate in group sex events. Men were eligible to complete the survey if they had had group sex with at least two other men simultaneously within the previous 5 years, regardless of their own HIV serostatus or that of their sexual partners. Apart from recruitment through gay community networks, specific recruitment was also conducted through networks of bisexual men.

On completion of the survey, men were invited to participate in the qualitative arm of the study for an in-depth interview; 16 of 25 Sydney and Melbourne men who volunteered were interviewed in detail, to provide more nuanced description and analysis of the men’s sexual cultures.


Men were asked about their most recent group sex encounter (GSE). They were asked to describe the sex practices in which they engaged at the GSE, including whether or not they used a condom for anal intercourse, and the perceived HIV serostatus of the men with whom they had sex. They were also asked what types of STI tests they had had since the GSE and whether they discussed their sexual behaviour with their doctors.

Data analyses

For univariate analyses, categorical variables were analysed using χ2 statistics, whereas scales as well as ordinal or continuous variables were analysed using one-way analyses of variance. Multivariate logistic regressions were undertaken to determine independent associations with having been tested for HIV and other STI. All variables significant at the univariate level were tested in the multivariate analyses. Due to the potential for other factors to influence the likelihood of testing over time, we restricted our analysis only to men who had engaged in group sex within the previous month.

The in-depth interview data consist of interview transcripts, so we used texts produced in and through the discursive event of an interview.20 Our analytical orientation in drawing upon these data was to explore whether men’s descriptions of their sexual activity and testing behaviour provided an account of knowledge formations to provide some explanation of the profile of testing patterns in our survey data, particularly as this relates to the motivations for testing among the men we interviewed.


In total, 994 eligible men completed the Three or More Study (TOMS) questionnaire between May 2007 and February 2008, of whom 446 (44.9%) reported engaging in group sex within the previous month. The following analyses are restricted to these 446 men.

Participants were mainly (79.6%) recruited online, but 17.3% were recruited onsite at sex-on-premises venues or privately organised sex parties and 3.1% through networks of men who organise and participate in private group sex parties. The mode of recruitment was not significantly associated with the main demographic items or outcome measures in this paper.

Participants ranged in age from 18 to 67 years, with a mean of 40.0 years. Over half (56.7%) had some university education. Nearly half the respondents were in managerial (13.5%) or professional (31.2%) occupations. Most (81.6%) were of Anglo-Celtic background. The place of residence largely reflected the intensity of local recruitment: 52.0% were from Sydney, 19.7% from Melbourne and 9.0% from Brisbane. Most (85.0%) identified as homosexual.

The 16 men who were interviewed in-depth ranged in age from 25 to 60 years. In terms of ethnic background, education, occupation and sexual identity, they were similar to the survey sample: mostly well-educated, professionally employed Anglo-Australian gay men. Ten lived in Sydney and six lived in Melbourne. Seven reported being HIV positive and the remainder reported being HIV negative. Six were in a regular relationship, four with a man and two with a woman and the remainder were not in a relationship.

Most men (90.1%) indicated that they had ever been tested for HIV. Two-thirds (66.8%) reported being HIV negative and nearly a quarter (23.3%) being HIV positive.

Over a quarter (28.5%) reported that there was a total of three men, including themselves, at the last GSE within the previous month, but 226 men (50.7%) indicated that there were at least four other men in addition to themselves. The majority (63.0%) reported that at least some of the men present were previously known to them. These included: previous sex partners (40.0%), friends (26.6%), their regular partners (20.2%) and other men they had met on another occasion (32.5%). Men reported a broad range of sex practices with partners other than their regular partner at the GSE, including oral–anal contact, oral intercourse with ejaculation, digital penetration and UAI (table 1). Among the 145 men reporting UAI, the majority reported UAI with a non-regular partner they did not believe to be HIV seroconcordant. These included 45 men who reported being HIV positive, 42 men who reported being HIV negative and 12 men whose HIV serostatus was unknown.

Table 1 Sex practices with non-regular partners at most recent GSE within previous month (N  =  446)

Half the men reported having had a sexual health test (any test for STI or HIV) since their most recent GSE within the previous month (table 2). Among men who had been tested since the GSE, the mean number of types of tests reported was 3.30.

Table 2 Testing for STI since last GSE within previous month

The likelihood of having a sexual health test since the last GSE within the previous month was not associated with age, cultural background or homosexual identity at the univariate level. All factors associated at the univariate level with having been tested (table 3) were included in the multivariate analysis. Engaging in UAI with a non-regular partner at the GSE (adjusted odds ratio (AOR) 1.77, 95% CI 1.16 to 2.70, p = 0.008), having at least four other men at the last GSE (AOR 1.67, 95% CI 1.11 to 2.51, p = 0.014), having previously met any of these men (AOR 1.84, 95% CI 1.19 to 2.84, p = 0.006) and having less than university education (AOR 1.62, 95% CI 1.08 to 2.43, p = 0.019) were independently associated with having been tested. Separately, having engaged in UAI with a non-regular partner at the GSE was also independently associated with each type of test for STI other than HIV. After excluding men who reported already being HIV positive, UAI with a non-regular partner (AOR 2.45, 95% CI 1.42 to 4.23, p = 0.001) and having at least four other men at the last GSE (AOR 1.70, 95% CI 1.05 to 2.80, p = 0.031) were independently associated with having been tested for HIV since their last GSE within the previous month.

Table 3 Univariate analysis of associations with HIV and STI testing since most recent GSE within previous month (N  =  446)

R9 (HIV negative, self-reported), who reported being tested for STI every 3 months, indicated that his lack of diagnosis with any infections was one means of being reassured that he was not putting himself at risk, despite having reported elsewhere in the interview that he had engaged in UAI with partners he did not know to be seroconcordant:

“I just don’t think it is that risky to have unprotected sex, for the insertive partner… And I get tested very regularly, and I haven’t caught anything.” R9

All men were asked whether, and how often, they discussed their group sex behaviour with their doctors. Half (51.8%) indicated that they had discussed this with their doctors, including 17.3% who often did so. Among the 158 men who had been tested for STI since their last GSE within the previous month, those who discussed group sex with their doctors were likely to report having had a greater number of types of STI tests than those who had not discussed group sex with their doctor. The 59 men who had not discussed group sex with their doctors had a mean of 2.58 types of STI tests since that GSE; the 58 men who occasionally discussed group sex with their doctors had a mean of 3.52 tests and the 41 men who often discussed group sex with their doctors had a mean of 4.05 types of STI tests (p<0.001). This trend was also true among men who did not report being HIV positive and who had been tested since the last GSE within the previous month. Among these non-HIV-positive respondents having discussed group sex with the doctor (AOR 3.70, 95% CI 1.57 to 8.74, p = 0.003) and not having previously met any of the men at the GSE (AOR 3.03, 95% CI 1.24 to 7.39, p = 0.015) were independently associated with having received at least four types of tests in the multivariate analysis.

Discussing details of sexual behaviour can be uncomfortable or embarrassing for some patients, and their doctors. R2 (self-reported HIV negative) reported being tested every 3 months. His doctor, however, did not understand his need to be tested so often:

“And for whatever reason, and this might sound a bit overzealous, but I have tests every three months. My doctor tries to get me to only test every six months, and so I alternate his visit with a visit to the Sexual Health Clinic, and that also makes me seem like a bit less of a slut…” R2

R2 also noted that he had one particular partner with whom he regularly engaged in UAI and that this partner was also tested regularly.

Among the 342 men who did not report being HIV positive and had last had group sex within the previous month, 94 (27.5%) indicated they had been tested for HIV since the GSE. Most also reported having tests for STI other than HIV; 18 had been tested only for HIV. Those who had engaged in UAI with a non-regular partner at the GSE were more likely to have been tested for HIV since that GSE than those who had not engaged in UAI (table 4; p<0.001).

Table 4 Testing for HIV and UAI with non-regular partners since last GSE within previous month among men who reported being HIV positive


UAI with non-regular partners, a major risk factor for HIV transmission among homosexual men, was often reported in this sample, as were other sex practices implicated in the transmission of STI other than HIV, such as oral intercourse, oral–anal contact and digital penetration.13 Given the broad range of their sex practices and their large numbers of partners, most men in this sample who had recently engaged in group sex were at high risk of infection with STI. Moreover, a majority of the reported UAI was with partners not believed to be HIV seroconcordant.

HIV and STI testing was relatively common but not universal. These data do not necessarily indicate a strong relationship between engaging in group sex and testing for STI, including HIV. The majority of gay men in Australia report being tested within a 6-month period.14 21 The rate of testing in this high-risk sample may have been due to the men testing for other reasons, such as routine screening, as much as for their involvement in group sex. Nonetheless, men who engaged in sex practices at the GSE that might be considered more high risk for the transmission of HIV and other STI were more likely to have since been tested and to have been more comprehensively tested. Overall testing rates were similar to those found in broader samples of Australian gay men, but recent testing was high. Men who engage in group sex relatively often may be especially conscious of the need for testing. Other studies have found that gay men who have not been tested for HIV are at lower risk than those who have been tested.15 Nonetheless, in our study, many of those who had not been tested had engaged in high-risk behaviours.

Although the majority of men did not discuss their involvement in group sex with their doctors, those who did so, and particularly those who often did so, were more likely to have been more comprehensively tested for a range of STI than were men who did not discuss such issues with their doctor. These data suggest that having a regular doctor with whom patients can discuss such aspects of sexual behaviour is an important factor in obtaining adequate sexual health screening. Embarrassment in discussing these issues, both for the doctor and patient alike, may lead to confusion or inadequate care. R2’s alternating of tests with his doctor and the sexual health clinic was driven by his desire for frequent testing but his comment that this makes him “seem… less of a slut” suggests he is also managing impressions. The word “slut” conveys powerful meaning, but its use here also conveys contradiction between his own embarrassment and his apparent desire to articulate himself through the use of that word. The question of how ideas about personal identity relate to participation in group sex and the potential implications for health outcomes require further investigation.

Nonetheless, testing can have unintended consequences. Some men who test regularly may view this activity in itself as somehow protective, particularly if they repeatedly test negative, reinforcing a belief that their behaviour is low risk. We examined the relationship between testing and group sex within the previous month. The time between the GSE and completing the survey may have been insufficient for testing in some cases, but other men may have been tested too soon for effective diagnosis, particularly for HIV. A negative result in such cases may have been unreliable.

Many respondents had previously met some of the other men at the GSE and these respondents were also more likely to have been tested since the GSE. In some cases, they were even aware of the testing patterns of some of their sexual partners. This suggests that men who are more closely connected to group sex networks may be more open to testing, or may have better access to appropriate screening services. This relative social connectedness may provide opportunities for informal types of contact tracing. Interestingly, though, among non-HIV-positive men who had been tested for STI since the GSE, sex at the GSE with men they had not previously met was associated with more comprehensive STI testing.

This was a sample specifically of men who had recently engaged in group sex. Respondents were predominantly recruited online, and mostly they reported living in Sydney, Australia. Although neither method of enrolment nor geographical location were associated with the main outcome measures, these findings may not apply to other populations. In general, this sample was demographically similar to other samples of Australian gay men,22 23 although HIV prevalence was somewhat higher.24 Also, rates of UAI at a single sexual encounter in this sample were approximately equivalent to what has been reported for any encounters during the previous 6 months among gay men in other Australian samples.25

Our findings confirm that gay men who engage in group sex are at high risk of transmission and infection with HIV and other STI and that considerations of potential risks figure highly in their decisions about testing. One possible factor in improving the quality of this testing is the relationship between such men and their doctors. Promoting an open, non-judgemental and trusting relationship with doctors with whom they can feel safe discussing all aspects of their sexual behaviour may lead to improved testing decisions. Even so, there is relatively good recognition of the risk of STI transmission among these men and their testing behaviour reflects this fact. The willingness of many men in these group sex networks to test for STI when they engage in risk behaviour, and the role doctors can play in this, may provide opportunities to expand these healthy behaviours across these high-risk sexual networks.


The authors thank the AIDS Council of New South Wales, the Victorian AIDS Council and the Queensland Association for Healthy Communities for collaboration with the project and the men who have participated in the TOMS study.



  • Funding: Supported by the New South Wales Health Department (Sydney). The National Centre in HIV Epidemiology and Clinical Research and the Australian Research Centre in Sex Health and Society receive funding from the Commonwealth Department of Health and Ageing. The Three or More Study was partly supported by the New South Wales Health Department. This funding agency had no further role in the study design, in the collection, analysis and interpretation of data, in the writing of the report or in the decision to submit the paper for publication.

  • Competing interests: None.

  • Ethics approval: Ethics approval was obtained.

  • Contributors: GPP and JH designed the TOMS study and wrote the protocol. GPP managed the literature searches and summaries of previous related work. GPP undertook the statistical analysis and wrote the first draft of the manuscript. All authors contributed to and have approved the final manuscript.