When serosorting is seroguessing

Gay men attempt to reduce their risks of acquiring or transmitting HIV using information about HIV status, the Eighth AIDS Impact Conference in Marseille heard last week, but use a complex, fallible and sometime contradictory mix of guessing and rationalisation to do so.

As a result, said Peter Keogh of Sigma Research in London, who was one of the presenters in the session on ‘Risk Perception’: “The notion of who should be responsible for safer sex, and what responsibility consists of, becomes ever more fraught.”

Over the past few years there has been increasing interest in understanding what gay men do to avoid HIV infection and transmission when they know their own and their partner’s HIV status. This behaviour has been dubbed `serosorting` by some researchers, because it often involves choosing to have unprotected anal intercourse only with partners of the same HIV status.

We put ‘knew’ in quotes because when Zablotska re-interviewed 178 positive and 249 negative gay men about whether their knowledge of partners’ HIV status came from overt disclosure, she found that a quarter of the HIV-positive men and 40% of the negative men who said they ‘knew’ their HIV partner’s HIV status had in fact guessed it.

She also found that, in HIV-negative men at least, the proportion of sex acts that were condomless was actually higher when men guessed their partner’s status than when they had discussed it.

Amongst the men who said they had UAI at all, if partners had neither discussed nor assumed their partner’s status, she found, 30% of the sex had by HIV-positive men, and 34% of the sex had by HIV-negative men, was unprotected. If HIV status was overtly discussed, then 87% of the sex had by HIV-positive men and 58% of the sex had by HIV-negative men was unprotected.

But on the occasions when men assumed their partner’s status, then 78% of the sex had by HIV-positive men and 61% of the sex had by HIV-negative men had was unprotected.

HIV-positive men were 5.2 times more likely to have unprotected sex if they knew their partner’s HIV status, and 3.2 times more likely if they assumed it, Zablotska said. HIV-negative men were 1.9 times more likely to have unprotected sex if they knew their partner’s HIV status, and 2.1 times more likely if they assumed it – and they assumed it nearly twice as often as positive men.

At last week’s conference Iryna Zablotska of the University of New South Wales in Australia introduced a new term – ‘seroguessing’ – for what a lot of gay men actually do.

She said that the proportion of Australian gay men, both with and without HIV, who both had unprotected sex and who attempted to serosort, in other words restrict it to men of the same HIV status, had increased between 2001 and 2006.

Zablotska was analysing figures derived from two national surveys of gay men, the Positive Health cohort, a group of 549 HIV-positive men established in 2006, and the HIM (Health In Men) survey, an annual gay men’s health survey established in 2001. Between them, these comprised just over 2000 men. She looked at the reported figures for unprotected anal sex (UAI) amongst the men and the figures for UAI between same-status men.

She then re-interviewed a subset of 427 gay men to find out whether their ‘knowledge’ of their partner’s HIV status was in fact knowledge, based on open discussion, or a guess.

Between 2001 and 2006 the proportion of HIV-positive men who had unprotected sex (with casual partners) increased from 32% to 40% and HIV-negative men from 12% to 16%.

At the same time the proportion of men who said they restricted casual sex (protected or not) to partners of their own status increased from 21% to 34% in HIV-positive men. The proportion of HIV-negative men who said they restricted sex (protected or not) to other negative men increased from 9% to 15% during this period.

But in HIV-negative men especially, knowledge of a casual partner’s HIV status was very much the exception rather than the rule: 80% said at least some of their sexual encounters were with partners of unknown status.

Amongst HIV-positive men, if it was ‘known’ that the other partner had HIV, then condom use became a rarity. The proportion of anal sex between two known HIV-positive partners that involved condoms decreased from 43% in 2001 to only 7% in 2006. This, however, was within a context in which only 38% of HIV-positive men said they always disclosed their HIV status themselves.

Overt serosorting – agreeing to have unprotected sex with a partner of the same ‘known’ HIV status – was less common amongst HIV-negative men. This is hardly surprising as ‘knowledge’ of HIV status is intrinsically a less certain thing if men are negative, as it only dates from the last HIV test. However it did seem to be increasing. The proportion of anal sex that was unprotected between casual partners who ‘knew’ they were both negative increased from 5% in 2001 to 22% in 2006 (in apparent contrast to figures from London – but see Elford below).

In contrast to the figures from Australia, the UK’s Jonathan Elford reported no apparent increase in serosorting behaviour amongst HIV-negative men between 1998 and 2005. In his latest unveiling of figures from his annual survey of gay men using London gyms, Elford said that the proportion of HIV-positive men who restricted unprotected sex to casual partners of the same HIV status increased from 6.6% to 17.7% during this period. However the proportion of HIV-negative men who reported overt serosorting remained flat, at around 1.5% to 2%.

This is an apparent contrast to the Australian figures, but it’s important to note that Elford and Zablotska were asking different questions. Elford was asking the negative men both whether they had unprotected sex and, if they did, whether they restricted it to men they were sure were negative. Zablotska was asking negative men what proportion of sex was unprotected if HIV-negative men knew their partner’s status for sure - and only 15% even thought they knew for sure.

This means that the proportion of HIV-negative men in Australia who were overtly serosorting – both establishing casual partners’ HIV status and restricting unprotected sex to them - only increased from 0.45% to 3.3% during the same period.

The other presentations in the session were qualitative studies of the thinking behind the ‘seroguessing’ – the calculations and rationalisations gay men use to ascertain or estimate status and decide whether to have unprotected sex.

A presentation by Maria Luisa Cosmaro from the Italian Association Against AIDS (LILA) combined a qualitative survey of 20 in-depth interviews with gay men who acknowledged risky sex with a larger quantitative survey of 322 gay men.

The qualitative sample was selected from gay men aged 18-55, 30% of them in a steady relationship, and was chosen so that a third each of the sample was HIV-positive, tested HIV-negative, and untested. In the quantitative survey a larger proportion of men were in a steady relationship (43%) and a larger proportion had tested HIV-negative (68%, against 11.5% positive and 20.5% untested).

In the quantitative survey, consistent condom use was uncommon. Twenty-one per cent said they ‘often’ used condoms, 42% ‘sometimes’ and 37% ‘rarely’.

Most (95%) went to gay clubs, 63% went to cruising areas (21% frequently) and 78% used gay chatlines (42% frequently). Of the men who used chatlines, 80% had had sex with someone they met online, and 20% ‘often’ did.

Italian gay men had high levels of knowledge and usage of post-exposure prophylaxis (PEP). Three-quarters of them had heard about it, 10% had taken it and 2% had taken it ‘often’. They also took HIV seriously; three-quarters of them said they ‘feared AIDS’, and 48% of them added ‘despite the availability of HAART’.

When asked to describe what they would regard as their main risky behaviour (they could describe more than one), 47% said ‘sex (or anal sex) without a condom,’ 20% contact with body fluids, and 15.5% either ‘promiscuity’ or ‘little attention paid to choice of sexual partner’. Five per cent said ‘sex with people looking unhealthy’ 2.5% said ‘thoughtlessness due to drug use’ and 2% said oral sex without a condom. Six per cent said that all the sex they had was risky and nearly 12% said they did not know what was risky.

The most important part of the research was probably the qualitative interviewing, where men were asked to describe why they had chosen to have unprotected or high-risk sex and these were then sorted into (non-exclusive) categories.

These revealed a level of knowledge about sexual partners and about how one could judge HIV risk that Cosmaro described as ‘very superficial’:

• 46% said they felt safe if they chose sexual partners carefully, without defining this further

• 43.5% said they did not insist on condoms for fear of rejection

• 43% said they felt they were safe because they didn’t have sex very often

• 43% said that they would think that a partner proposing sex without a condom would likely be HIV-negative

• 36% said condoms were either unpleasant or ruined the excitement

• 33% said they relied on partners saying they were HIV-negative (one said “If he is negative, then I am negative”

• 19% said they “were HIV-positive and want to enjoy life.” Cosmaro later commented that this apparently irresponsible attitude was tempered by the complementary rationalisation that the HIV-negative men used: That someone willing to have sex without a condom must already ve HIV-positive

• 12% said they were HIV-negative and simply had faith they always would be.

Cosmaro commented that (in common with the Zablotska survey), an HIV serostatus described as ‘known’ was often nothing of the sort, because men used cues such as outward appearance to decide on someone’s status. She said it was a challenge to prevention that gay men actually sought out what she called “duty free zones” where concern about risk was deliberately left behind. For instance, 42% of the men interviewed agreed that they would regard a gay venue that did not feature HIV prevention posters and literature as more fashionable and ‘sexy’ than one that did.

In contrast to the situation with casual sex, Elford found that HIV-negative men were more likely than HIV-positive men to have unprotected sex with primary partners of the same HIV status than HIV-positive men. This is hardly surprising as HIV-negative men are more likely to find themselves in a concordant relationship than HIV-positive men.

Where the negative status has been ascertained through testing, this situation is better called ‘negotiated safety’ than ‘serosorting’, Elford said. The proportion of HIV-negative men who restricted unprotected sex to a main partner of the same status increased from 12.4% in 1998 to 19% in 2005 and the proportion of HIV-positive men who did the same increased from 5.1% to 10.1%.

Iryna Zablotska’s colleague at the University of New South Wales, Jeanne Ellard, interviewed a group of recently-diagnosed men about ‘Why I thought I wouldn’t get HIV’.

She said she had found that gay men were not complacent about the health risks of HIV and did not want to catch it and that there was little evidence of ‘treatment optimism’ – that people with more positive views of HIV treatment were more likely to have unsafe sex.

However she also found that men used a lot of rationalisations to ‘talk themselves into’ unsafe sex.

One said he’d persuaded himself HIV was rare: “I would think about all the times I’d had unprotected sex and was fine, and concluded that that showed there weren’t many HIV-positive guys in Sydney. I certainly didn’t know any.”

Another rationalised that ‘fit’ guys on the scene must be negative: “If they were HIV-positive, their quality of life would be lower and they wouldn’t be out partying.”

If someone overtly disclosed HIV status then sex was protected, but men more often used guesswork: “When you sense someone doesn’t want to use a condom, then it sends a little trigger to your head that he might be positive,” someone said.

In a paper from the USA, Lisa Eaton of the University of Connecticut found that 36.5% of 628 men interviewed at Gay Pride events agreed with the statement “I’d be more likely to have unprotected sex if my partner told me he had the same HIV status.”

Gay men who answered ‘yes’ to the question were three times more likely to have had anal sex the last time they had sex and 2.2 times more likely to have had unprotected anal sex than ones who said ‘no’; they were also 50% more likely to believe that serosorting was an effective strategy against HIV.

They also had on average a 21% lower score when it came to a series of questions gauging their self-efficacy – their confidence in their ability to achieve goals.

Finally, Peter Keogh of Sigma Research in the UK quoted data that showed that HIV-negative gay men with proximity to the epidemic, i.e. knew or had had sex with someone they knew to be positive, were more realistic about whether HIV-positive men would disclose or not and what their chances were of encountering one.

He interviewed 36 men who had engaged in high risk sex and divided them into ‘high proximity’ men who either knew HIV-positive men or had had sex with one in the past year, and low-proximity men who didn’t and hadn’t.

The low-proximity men expressed contradictory attitudes towards status disclosure. On the one hand they thought people with HIV were unlikely to disclose. “What person’s going to really admit to HIV?” asked one. Another said “Most people don’t talk about HIV – it’s a ‘don’t ask, don’t tell’ attitude.”

On the other hand they assumed that if someone HIV-positive did have sex with them they would disclose: “I would expect disclosure because I’d imagine they’d feel a sense of duty,” said one. “If someone has HIV they should say,” said another.

In contrast the high-proximity men had more realistic attitudes. They assumed that a high proportion of men they knew did have HIV: “I often hear about people I didn’t know were positive,” said one. “If I have sex in a sauna I know that a lot of guys are going to be positive; it’s just statistics,” said another.

They did not expect HIV-positive men to disclose straight away but assumed they would if a relationship started to happen: “I’d expect status to be discussed if guys start wanting something more,” one said. The high-proximity men were less anxious about HIV than the low-proximity men, but more fatalistic: “Sometimes you’re safe; sometimes you’re not,” one shrugged.