Not so different? Comparison of risk profile of gay men who acquired HIV while travelling with those who acquired HIV in Australia.

ISSUE ADDRESSED: Many countries now identify HIV and international mobility as a priority issue within a global and shared epidemic, including Australia. To support health promotion in this complex area, we investigated recent HIV infections that occurred among Australian gay men while travelling and compared to HIV infections that occurred in Australia. METHODS: 446 gay men recently diagnosed with HIV completed an on-line survey regarding the high risk event (HRE) where they believed that they acquired HIV. Those who acquired HIV while in their usual place of residence (308 men), those who were travelling within Australia (59 men), and those who were travelling overseas (79 men) were compared. RESULTS: Those who acquired HIV while overseas had very similar risk profiles, sexual behaviour, and made similar assumptions about their partners and their own HIV status, as those who acquired HIV in Australia. Only HIV status disclosure at the HRE differed across locations (P = .030). Three quarters (74.7%) of the men who acquired HIV while overseas were not diagnosed until they returned to Australia. CONCLUSIONS: Our findings challenge the idea that there are necessarily differences in behaviour and assumptions for HIV transmission in Australia and overseas. However, the men travelling may be in communities where HIV status is less commonly disclosed, and where HIV prevalence is higher. SO WHAT?: A deeper understanding of contextual factors may be required for HIV prevention and health promotion strategies targeting gay men travelling to locations with different cultural, HIV prevalence, and HIV testing considerations. This would also identify opportunities for new tools such as Pre-Exposure Prophylaxis and self-testing.

High-risk sexual behaviours among gay and bisexual men: comparing event-level casual sex encounters among seroconverters and non-seroconverters.

BACKGROUND: With increasing use of non-condom-based HIV risk reduction strategies by gay and bisexual men (GBM), we compared occasions of condomless anal intercourse with casual partners (CLAIC) that resulted in HIV transmission and similar occasions when HIV transmission did not occur. METHODS: We compared two demographically similar samples of Australian GBM. The HIV Seroconversion Study (SCS) was an online cross-sectional survey of GBM recently diagnosed with HIV. The Pleasure and Sexual Health (PASH) study was an online cross sectional survey of GBM generally. Using logistic regression, we compared accounts of CLAIC reported by men in SCS as being the event which led to them acquiring HIV, with recent CLAIC reported by HIV-negative men in PASH. RESULTS: In SCS, 85.1% of men reported receptive CLAIC, including 51.8% with ejaculation; 32.1% reported having previously met this partner and 28.6% believed this partner to be HIV-negative. Among HIV-negative men in PASH reporting recent CLAIC, 65.5% reported receptive CLAIC, including 29.9% with ejaculation; 59.3% reported having previously met this partner and 70.1% believed this partner to be HIV-negative. CONCLUSIONS: While both groups of men engaged in CLAIC, how they engaged in CLAIC differed, and the context in which they did so was different. A generic measure of CLAIC conceals the critical elements of HIV risk, particularly the role of receptive CLAIC, among GBM that distinguish those who seroconverted and those who did not. Detailed information about the context and nature of the practise of CLAIC is required for a more complete understanding of HIV risk among GBM.

In Australia, Most HIV Infections Among Gay and Bisexual Men are Attributable to Sex with 'New' Partners.

It has been estimated that the majority of global HIV infections among gay and bisexual men (GBM) can be attributed to sex within a committed relationship. In Australia, however, negotiated safety, whereby HIV-negative regular partners agree to discard condoms with each other but commit to consistent condom use with other partners, has been promoted as a key component of the HIV prevention response. We asked GBM recently diagnosed with HIV to describe their relationship to the person they believed to be the source of their infection ('source person'). The majority (66.1%) ascribed their infection to a casual partner. A further 23.3% ascribed their infection to a non-committed and non-romantic partner (or 'fuckbuddy'). Only 10.6% believed they had acquired their HIV from a 'boyfriend' in the context of a committed romantic relationship, and 51.7% of these occurred within the first 3 months following their first sexual contact. Most men (61.5%) believed they had acquired their HIV infection on the first occasion they had sex with the source person. In the Australian context, negotiated safety appears to have minimised infections between regular partners. However, many HIV infections between regular partners may not be in the context of a romantic committed relationship, and yet this distinction between types of regular partners has been all but ignored. Furthermore, in this sample, most infections occurred on the occasion of first meeting, suggesting that the most useful indicators of risk may be the characteristics, contexts, and lengths of sexual partnerships and how sex is negotiated, rather than how GBM categorize their partner. Findings suggest more new HIV infections occur in new partnerships, than in established relationships.

The Road Less Travelled: Exploring Gay and Bisexual Men's Explanations of 'Uncommon' Routes of HIV Transmission.

Although there are practices other than condomless anal intercourse that may result in HIV transmission among gay and bisexual men, very little is known about these 'uncommon' transmission explanations. To address this topic, the free text survey responses from 465 HIV positive gay men in Australia were thematically analysed; 123 participants offered uncommon explanations for their seroconversion. Men described several sexual acts they believed led to infection, categorised as adventurous sex (e.g., fisting) and foreplay (e.g., oral sex). Participants also identified mediating factors associated with their seroconversion, either internal (e.g., cum/pre-cum) or external (e.g., sores, illness) to sex. Finally, contextual forces associated with infection were also explored, namely physical spaces (e.g., sex on premises venues) or mental states (e.g., depression). While some uncommon explanations are unlikely to have resulted in HIV transmission, these accounts reveal the diverse and intersecting ways that men attempt to make sense of their seroconversion.

The Meaning of 'Regular Partner' in HIV Research Among Gay and Bisexual Men: Implications of an Australian Cross-Sectional Survey.

Estimates of the proportion of HIV infections coming from within regular sexual relationships among gay and bisexual men (GBM) vary widely. Research surveys use various partner type categories, but there is little understanding of how men classify their partners. We conducted an online cross-sectional survey of Australian GBM exploring sexual relationships, including 2057 men reporting on 2566 regular partnerships. Just over half of the partnerships were considered 'relationships', while the remainder were non-romantic 'fuckbuddy'-style arrangements. In multivariable analysis, factors associated with considering the partnership a 'relationship' were: using a 'romantic' descriptor, partnership length, monogamous agreements, any condomless anal sex with each other, love, and commitment. The category of 'regular partner' can mask diverse partnership types, which have different meanings to GBM, associated behaviours, and HIV risks. Certain HIV prevention techniques may be more suited to particular types of partnerships. 'Fuckbuddy' arrangements need to be more explicitly acknowledged in HIV prevention.

Australian Gay Men Describe the Details of Their HIV Infection Through a Cross-Sectional Web-Based Survey.

BACKGROUND: With emerging opportunities for preventing human immunodeficiency virus (HIV) transmission, it remains important to identify those at greatest risk of infection and to describe and understand the contexts in which transmissions occur. Some gay and bisexual men with recently diagnosed HIV infection are initially unable to identify high-risk behaviors that would explain their HIV infection. We explored whether Web-based data collection could assist them in identifying the circumstances of their infection. OBJECTIVE: To assess the capacity of a Web-based survey to collect reliable self-report data on the event to which gay and bisexual men ascribe their HIV infection. METHODS: The HIV Seroconversion Study included a Web-based survey of gay and bisexual men with recently diagnosed HIV infection in Australia. Participants were asked if they could identify and describe the event they believe led to their infection. Men were also asked about their sexual and other risk practices during the 6 months before their diagnosis. RESULTS: Most (403/506, 79.6%) gay and bisexual men with newly diagnosed HIV infection were able to identify and describe the circumstances that likely led to their infection. Among those who were initially unable to identify possible exposure events, many could nonetheless provide sensible information that ostensibly explained their seroconversion. Free-text responses allowed men to provide more detailed and contextual information, whereas questions about the totality of their sexual behavior before diagnosis provided opportunities for men to describe their sexual risk behavior in general. Overall, 84.0% indicated having engaged in condomless anal intercourse before their HIV diagnosis, including 71.8% in the receptive position. CONCLUSIONS: This study demonstrates the effectiveness of using Internet-based technologies to capture sensitive information about the circumstances in which HIV infection occurs among gay and bisexual men. By providing a range of opportunities for relaying experience, this research reveals some of the complexity in how individuals come to understand and explain their HIV infection. These findings may assist in obtaining detailed sexual history in the clinical setting.

Gay and bisexual men's interest in marriage: an Australian perspective.

Same-sex marriage is a widely debated issue, including in Australia. This study used an online anonymous survey, with free-text responses, to investigate romantic and sexual relationships among Australian gay and bisexual men. We sought to identify what proportion of such men intended to marry their primary regular partner if marriage was made legally available to same-sex couples in Australia, as well as factors associated with intention or non-intention to marry. Most men in the sample did not intend to marry their primary regular partner. Even among men who considered themselves to be in a 'relationship' with their primary regular partner, less than half intended to marry him. However, many men who would not marry their current primary regular partner agreed that same-sex marriage should be available for gay and bisexual men in Australia. Reasons for intention to marry included a desire for social and legal equality, and ideas about marriage as a rite of passage, an expression of love and the most valued form of relationship in Australia. Those who did not intend to marry their primary regular partner offered a number of reasons, including that the nature of their relationship was incompatible with marriage, and reported a critical position towards marriage as a heteronormative institution.

Factors associated with recent previous HIV testing among a sample of recently HIV-diagnosed gay men in Australia: a cross-sectional study.

OBJECTIVE: Timely HIV testing among recently HIV-infected gay men may enable earlier access to clinical care and changes in behaviour that will reduce onward transmission. We investigated the testing practices of men recently diagnosed with HIV to identify factors associated with recent testing. METHODS: In an online survey of men in Australia recently diagnosed with HIV, participants were asked about their HIV testing history, perceived impediments to testing prior to diagnosis, motivation for testing at the time of diagnosis and a range of demographic and behavioural characteristics. Descriptive statistics were used to compare those men who reported recent HIV testing with those men who had not tested for HIV in the 12 months before their diagnosis. RESULTS: Of 187 men who provided information about their testing history and social connectedness, 6.4% were previously untested for HIV, whereas 65.8% had last tested within the 12 months prior to their diagnosis. Factors associated with having tested more recently were being more socially engaged with other gay men (OR 1.34; 95% CI 1.10 to 1.63; p=0.003) and having greater optimism about HIV health (OR 1.13; 95% CI 1.00 to 1.27; p=0.047). In multivariate analysis, only level of social engagement with other gay men remained independently associated (adjusted OR 1.30; 95% CI 1.07 to 1.59; p=0.003). CONCLUSIONS: Gay community plays a key role in the response to HIV in Australia. Building a sense of community through programmes that support social engagement between gay men may support earlier and more frequent testing.

Online dating among Australian gay and bisexual men: romance or hooking up?

Increasingly, gay and bisexual men (GBM) meet casual sex partners online and this has been associated with sexual risk behavior. How do GBM meet regular partners? This online anonymous survey of 4215 GBM included 2562 men with a primary regular partner (PRP) who were included in these analyses. Mean age of the sample was 38.1 years. 60.3 % had met their PRP at least 2 years earlier. Meeting their PRP online increased from 14.0 % before 2001 to 79.9 % in 2013-2014. At all time points, men who met their PRP online were somewhat older than those who met their PRP offline. Regardless of how they met their PRP, most men met casual sex partners online. Among GBM, meeting sexual and romantic partners online has replaced other methods, for all age groups. The population of GBM who use the internet for this purpose is now equivalent to all sexually active GBM.

Understanding Gay Community Subcultures: Implications for HIV Prevention.

Gay and bisexual men (GBM) who participate in gay community subcultures have different profiles, including differing risk behaviors. We examined men's participation in gay community subcultures, and its association with risk behavior. In a cross-sectional survey, 849 GBM provided information about men in their personal networks. We devised measures of their participation in five subcultural groupings and explored their associations with sexual behavior. We identified five subcultural groupings: sexually adventurous; bear tribes; alternative queer; party scene; and sexually conservative. Higher scores on the sexually adventurous measure was associated with being older, having more gay friends, being HIV-positive, and being more sexually active. It was also independently associated with unprotected anal intercourse with casual partners (AOR 1.82; 95 % CI 1.20-2.76; p = 0.005). HIV prevention strategies need to account for the different subcultural groupings in which GBM participate. Measures of engagement with gay subcultures are useful indicators of differential rates of risk behavior and modes of participation in gay community life. Men in more sexually adventurous subcultures are more likely to engage in sexual risk behavior.

At home and away: gay men and high risk sexual practices.

We aimed to describe HIV risk practices of gay men who travel locally, regionally and overseas. We analysed data from the Sydney Gay Community Periodic Survey 2009 about high-risk sexual practices in four locations (locally, while travelling in NSW, Australia and overseas) and with partners of HIV positive, negative and unknown serostatus in each location. Analyses of associations used generalized log-binomial estimation procedures with Type I error of 5 %. Of 1,839 sexually active participants, 70.1 % reported having sex locally. 19.7 % elsewhere in NSW, 20.1 % interstate and 18 % overseas. Unprotected anal intercourse (UAI) was reported by 29.9, 28.6, 21.3 and 19.3 % of men in each location respectively. There was no difference in the levels of UAI locally and elsewhere in NSW, but UAI levels were lower in other Australian locations [adjusted prevalence rate ratio (APRR) = 0.76; 95 % confidence interval (95 % CI) 0.66-0.88] and overseas (APRR = 0.76; 95 % CI 0.65-0.89). UAI was more likely if partners were seroconcordant HIV positive (APRR = 1.67; 95 % CI 1.32-2.11) and less likely if partners were of different HIV serostatus (APRR = 0.39; 95 % CI 0.33-0.47) as compared to seroconcordant HIV negative partners. UAI was associated with group sex and use of party drugs. In this community sample, UAI levels were higher in the local context than in travel destinations, suggesting that familiarity between partners may play a role. High-risk sexual practices can nevertheless contribute to bridging different HIV epidemics and HIV transmission across borders. HIV prevention programs should develop effective approaches to target sexually adventurous gay men and HIV transmission associated with travel.

"It's hard to know what is a risky or not a risky decision": gay men's beliefs about risk during sex.

Gay men increasingly use non condom-based risk reduction strategies to reduce the possibility of HIV transmission. Such strategies rely on men's knowledge and communication with each other, but how they employ these strategies may depend as much on their attitudes toward risk and pleasure. We explored current beliefs about safe sex, sexual desire and risk behavior in an online survey of 2306 Australian gay men. The survey included free text components to explore men's beliefs about risk and pleasure. We conducted a principal components factor analysis on the safe sex belief items in the survey, and thematic analysis of the qualitative material was used to interrogate the concepts underpinning these beliefs. We identified two measures of safe sex beliefs: risk reduction optimism (HRRO; α = 0.703); and viral load optimism (α = 0.674). In multivariate analysis, unprotected anal intercourse with casual partners (UAIC) was associated with HRRO among non HIV-positive men only (p < 0.001), but, regardless of HIV serostatus, UAIC was associated with a belief that serosorting could be an effective risk reduction strategy and with being more sexually adventurous in general. Using the qualitative data we identified four themes in how men think about HIV: 'seeking certainty', 'regretful actions', 'nothing is safe', and 'acting on beliefs'. Each theme interacted with the safe sex beliefs measures to provide a highly contextualised understanding of men's beliefs about safe in specific circumstances. Gay men think about the risk of HIV transmission in qualitatively different ways depending on specific circumstances. While measures of belief about relative risk of HIV transmission are useful indicators of men's propensity to take risk, they oversimplify men's thinking about risk, and fail to account for the role of desire, both in influencing men's thinking about risk, and in how they balance their perception of relative risk against the pursuit of pleasure.

Is optimism enough? Gay men's beliefs about HIV and their perspectives on risk and pleasure.

BACKGROUND: Measures of HIV treatments optimism were developed in response to increased risk behavior among gay men, but were limited in their capacity to help understand gay men's risk behavior. METHODS: We explored current beliefs about HIV health and transmission and sexual desire and risk behavior in an online survey of 2306 Australian gay men. The survey included free text components. We conducted 40 qualitative interviews to explore how men's beliefs affected decisions about risk behavior. We conducted a principal components factor analysis on the optimism belief items in the survey, and thematic analysis of the qualitative material was used to interrogate the concepts underpinning these beliefs. RESULTS: We identified two measures of HIV optimism: Health Optimism (α = 0.791) and Transmission Optimism (α = 0.795). In multivariate analysis, unprotected anal intercourse with casual partners was only associated with HIV transmission optimism regardless of HIV serostatus (P < 0.001). Using the qualitative data, we identified 4 themes in how men think about HIV: "concerned," "unconcerned," "fearful," and "irrelevant." Each theme interpellates the 2 optimism measures. CONCLUSION: HIV optimism remains a useful indicator of gay men's likelihood to take risk, but technical knowledge, experience, desire, and attitudes to risk may all affect how people respond and often in multiple, sometimes contradictory, directions. Men's beliefs about HIV transmission risk in particular may reflect willingness to pursue pleasure over risk, or, alternatively, morbid fear of any risk. Measures of HIV optimism should be complemented by analysis of the complexities of individuals' assessments of both risk and pleasure in specific sexual contexts.

Would gay men change their sexual behavior to reduce syphilis rates?

BACKGROUND: The community at which public health strategies for reducing syphilis epidemics are potentially targeted may have different considerations with regards to their sexual and health priorities. We aimed to elicit information on the acceptability of behavior change interventions among gay men for reducing syphilis transmission. METHODS: We conducted an online survey (n = 2306 participants) and focus groups to determine whether further sexual behavior change to reduce syphilis is likely to be acceptable to gay men in Australia. RESULTS: One quarter of survey respondents (26%) indicated that they would be highly likely to reduce partner acquisition rates in order to reduce their chances of syphilis infection. However, among the 475 (21%) men who reported greater than 10 partners in the previous 6 months, only 11% indicated being "highly likely" to reduce partner numbers to avoid syphilis. Among 606 (26%) survey respondents who reported not always using condoms in the previous 6 months, 34% indicated being highly likely to always use condoms with casual partners to avoid syphilis. In the focus groups, men indicated little commitment to sexual behavior change but some willingness to consider short-term changes to reduce community syphilis levels. CONCLUSIONS: Interventions promoting partner reduction or increased condom use are unlikely to be adopted on a long-term basis by men at greatest risk. Behavioral interventions alone are unlikely to materially contribute to syphilis prevention among gay men.

Will changes in gay men's sexual behavior reduce syphilis rates?

BACKGROUND: Reducing rates of partner change and increasing condom usage among gay men are obvious targets for potentially reducing syphilis transmission among gay men. METHODS: We developed an agent-based stochastic model to examine syphilis transmission among a population of gay men, representative of gay men in Australia. This model was used to explore the potential impact of changes in sexual behavior over 1 month, 3 month, and indefinite time frames on syphilis epidemics. RESULTS: Simulations of interventions showed that short-term reductions in rates of partner change and increased condom use would have negligible impact on the long-term trends of syphilis epidemics. If no interventions are introduced, then the model forecasts that the syphilis prevalence in the population could continue to rise, with an increase of 80% in the number of men infected with syphilis during the next decade. However, if changes in sexual behavior are maintained in the long-term, then syphilis epidemics can be mitigated. If condom use is sustained at 80% in partnerships that are HIV discordant or of unknown status, then the prevalence of syphilis is estimated to decrease by 9% over 10 years. Similarly, if partner acquisition rates decrease by 25%, then there will be a 22% reduction in syphilis prevalence. CONCLUSIONS: Interventions promoting partner reduction or increased condom use would be ineffective in the short-term, and would have limited prospects for success in the long-term unless very large changes in behavior are sustained. Complementary social research indicates that such long-term changes in behavior are unlikely to be adopted, and therefore other intervention strategies need to be developed to reduce syphilis among gay men.

Chemoprophylaxis is likely to be acceptable and could mitigate syphilis epidemics among populations of gay men.

BACKGROUND: Over the last decade, syphilis epidemics have resurged around the world, particularly among gay men. An innovative public health response could be the use of chemoprophylaxis. We sought out to determine the acceptability of syphilis chemoprophylaxis and its likely population effectiveness if it were adopted. METHODS: We conducted a mixed-methods study. An online survey (n = 2095 participants) and focus groups (n = 23 participants) were conducted to determine whether syphilis chemoprophylaxis is likely to be acceptable to gay men in Australia. We also developed an individual-based mathematical model that simulated a population of gay men, to explore the potential impact of introducing chemoprophylaxis. RESULTS: Of the 2095 gay men surveyed, 52.7% (95% confidence interval, 50.6%-54.8%) indicated that they would be very likely or slightly likely to use chemoprophylaxis to reduce their chance of acquiring syphilis, increasing to 75.8% (95% confidence interval, 74.0%-77.6%) if chemoprophylaxis would help reduce infections in the gay community. In this model, 70% use-effectiveness of chemoprophylaxis used by 50% of gay men is expected to reduce the number of syphilis cases by ∼50% after 12 months and 85% after 10 years. The majority of the prevention efforts can be gained by targeting subpopulations of men with higher sexual activity. CONCLUSIONS: Chemoprophylaxis offers promise as an acceptable and effective intervention for mitigating syphilis epidemics. The outcomes of a planned placebo-controlled syphilis chemoprophylaxis trial are eagerly anticipated.

Sex partying among gay men in Sydney, Melbourne and Brisbane, Australia.

Group sex is relatively common among gay men, and is a likely risk factor for infection with HIV and other sexually transmissible infections (STI). The 5,432 participants in the 2008 Gay Community Periodic Survey in Sydney, Melbourne and Brisbane, Australia, were asked about their participation in group sex. Nearly half (44.0%) of those who reported sex with casual male partners in the previous 6 months also reported engaging in group sex. Among other factors, using drugs odds ratios (OR) and confidence intervals (CI) (OR = 1.89, CI = 1.61-2.21), being HIV-positive (OR = 1.62, CI = 1.13-2.32), and engaging in anal intercourse, both with a condom (OR = 3.03, CI = 2.46-3.73) and without a condom (OR = 5.68, CI = 4.53-7.12) were independently associated with having engaged in group sex. This study suggests that gay men who engage in group sex represent an important priority for targeted HIV and STI prevention activities and research.

Gay men who engage in group sex are at increased risk of HIV infection and onward transmission.

Among 746 participants in the Three or More Study (TOMS) of gay men who engaged in group sex in the previous 6 months, 22.4% reported unprotected anal intercourse (UAI) with any partners they did not know to be the same HIV serostatus as themselves. Not knowing oneself to be HIV-negative, not having a clear intention to use condoms, and more frequent group sex were independently associated with UAI. This study shows that gay men who engage in group sex represent an important priority for targeted HIV prevention activities and research.

Comparing Australian gay and bisexual men with undiagnosed and recently diagnosed HIV infection to those in the National HIV Registry.

Background Gay and bisexual men (GBM) with recent HIV infection are a key population to inform HIV-prevention. The National HIV Registry (NHR) provides details about all individuals diagnosed with HIV, but it is unclear how. METHODS: Basic sociodemographic characteristics of GBM in the NHR who were diagnosed between 2010 and 2014 were compared with three samples of GBM: men with undiagnosed HIV infection from the Community-Based Study of Undiagnosed HIV and Testing (COUNT) study of HIV prevalence and undiagnosed infection that was conducted during 2013-14; men in the Australian Gay Community Periodic Surveys (GCPS) who were diagnosed with HIV between 2010 and 2014; and men in the HIV Seroconversion Study (SCS) who were also diagnosed in those years. RESULTS: The NHR identified 3629 men who reported male-to-male sex as the exposure risk for their diagnosis between 2010 and 2014. COUNT identified 19 (8.9% of all men who tested HIV positive) men as having undiagnosed HIV. In the GCPS sample, 185 (2.5%) reported being diagnosed with HIV in 2010-14. In total, 367 men in the SCS received their diagnosis during 2010-14. The mean age of men in the NHR (36.8 years) was similar to that in GCPS (36.3 years) and SCS (35.1 years), while undiagnosed men in COUNT were younger (32.6 years), with no other significant differences between the samples. CONCLUSIONS: The undiagnosed men were somewhat younger than diagnosed men. To achieve earlier diagnosis of new HIV infections, improved HIV testing frequency is needed among younger men.

Impact of Peer Support on Behavior Change Among Newly Diagnosed Australian Gay Men.

INTRODUCTION: Treatment as prevention relies on early uptake of HIV treatment, but onward transmission during primary HIV infection may be affected by changes in sexual risk behavior after diagnosis. What factors are associated with these changes in sexual risk behavior among gay and bisexual men? METHODS: We surveyed gay and bisexual men recently diagnosed with HIV about changes to their sexual behavior since their diagnosis. In 2008-2010, 263 men described their sexual behavior during the 4 weeks before, and during the 4 weeks after, their HIV diagnosis. In 2010-2015, 301 men reported how they had changed their sexual behavior since their HIV diagnosis. RESULTS: During 2008-2010, 26.6% engaged in condomless anal intercourse with non-HIV-positive casual partners during the 4 weeks before diagnosis, and 9.7% did so during the 4 weeks after diagnosis (P < 0.001). Only peer support from other people with HIV was associated with this change in behavior (OR = 1.42; 95% confidence interval = 1.07 to 1.88; P = 0.014). Peer support was also associated with partner reduction after diagnosis (P = 0.010). During 2010-2015, 31.9% reported that they had increasingly disclosed their HIV status to sex partners since their diagnosis, and 74.1% reported having reduced the number of men with whom they had sex. Peer support was associated with both these changes in behavior (P = 0.003 and P = 0.015, respectively). CONCLUSIONS: The risk of onward transmission is likely to be less affected by immediate versus early treatment than it is by earlier diagnosis and peer support for those newly diagnosed with HIV. Enhanced peer support may further reduce the likelihood of onward transmission after diagnosis.