The major way in which HIV testing is already affecting sexual behaviour has been described as `negotiated safety' by Australian researchers. `Negotiated safety' refers to an agreement between partners about sexual practices which takes into account the HIV antibody status of both partners. Australian and British researchers observed that gay men who had taken the HIV test were having unprotected sex, and doing so with other men who also knew their antibody status. These men had worked out agreements about sexual activity outside the relationship that were designed to minimise the risk of infection within the relationship if it was not sexually exclusive.

These strategies were a consequence of the availability of HIV antibody testing from 1985 onward, but they were not promoted by HIV prevention campaigns. Indeed, most HIV prevention campaigns around the world have consistently promoted the message “Positive or negative, it's all the same” as a means of encouraging solidarity between infected and uninfected men.

Nevertheless, despite such messages, a significant minority of gay men adopted `negotiated safety strategies' because they were preferable to the continued use of condoms.

Since such strategies were first identified many people have argued that it is unreasonable to assume that long–term condom use is sustainable for the majority of gay men (Odets; King). A proportion of gay men will still continue to have unprotected sex, and the job of HIV prevention ought to be to facilitate the safe practice of unprotected sex, they argue.

This has been a difficult notion to accept in some quarters, for it appears to undermine the continued promotion of condom use. Yet research suggests that in practice gay men in particular are already using information about their HIV status to guide sexual practice, and that many gay men are much more careful about condom use with casual partners than they are with primary partners (Coxon). In his review of the SIGMA cohort data, Tony Coxon identified a number of worrying trends that are particularly important to consider in relation to debates about `negotiated safety':

• Those who engage most frequently in unprotected sex generally restrict their unprotected sexual activity to one partner who has the same HIV status.
• But, nearly 50% of all men interviewed invert any notion of `negotiated safety' by having unprotected sex with casual partners and regular partners, or else using condoms with their regular partner whilst having unprotected sex with one–off casual partners.
• Those not tested tend to have the highest–risk sex, primarily with those known to be negative.

However, Coxon also observes:

• The non–use of condoms is often as much a matter of waiting for the partner to object as it is a pre–negotiated condition (i.e. it happens by default).
• “paradoxically the group which overwhelmingly exhibits close conformity to the ideal of choice, responsibility and negotiation...are those already diagnosed HIV- positive.”.

These findings present a number of significant problems. They suggest that much unprotected sex is unnegotiated (verbally, at least), and whilst it may not lead to HIV infection with any greater frequency than unprotected sex with regular partners, it will present significant problems to anyone trying to develop `negotiated safety' approaches to HIV prevention.

For example, negotiated approaches rely on trust. One of the major criticisms of this approach is the degree to which it relies upon the honesty and trust of both partners. Given the degree to which personal relationships are bedevilled by problems of honesty and trust, it is a little difficult to imagine such agreements working well for some people.

Many partnerships have agreements about no sex outside the relationship, but these agreements are broken frequently. Sometimes the other partner will find out, but often they will not, because the partner who has been `playing away' is scared of the consequences for the relationship. If any agreement about unprotected sex outside the relationship is tied up with an agreement about no other sexual activity outside the relationship, this may reduce the incentive for honesty if a lapse in the agreement does occur.

When does negotiated safety work?

Australian researchers recently looked at whether making explicit agreements about unprotected sex had any effect on the likelihood that men would have unprotected sex outside their primary relationship. They looked at 165 men in seroconcordant HIV-negative relationships. 61% had engaged in unprotected anal intercourse at some point. The researchers found that amongst this group the following factors were associated with a lower likelihood of having unprotected sex with casual partners and with a primary partner (Kippax).

• Having an agreement about unprotected sex outside the relationship.
• Agreement included no anal sex with casual partners.

Men who found condom use unacceptable were more likely to have unprotected anal sex with their casual partners. No other demographic or behavioural factors were found to be significant.

Talk, Test, Test, Trust

Australian approaches to negotiated safety have used the slogan `Talk, Test, Test, Trust' to highlight the issues involved in making such a commitment. This approach is useful because it highlights the issue of trust, and makes it a talking point amongst the target audience. It implies that an agreement about unprotected sex is a serious matter.

The Australian approach has also used advertising about the testing procedure to follow if you want to come to an agreement about unprotected sex. This advertising emphasises that negotiated safety is a complex process, and shouldn't depend on assumptions about HIV status or shortcuts to abandoning condom use.

Several other factors are likely to distort the strategy of negotiated safety:

• Ability to negotiate and assert: some individuals may find it less easy to assert their doubts about the chosen strategy.
• Pressures on the gay scene to have sex outside the primary relationship, increasing the potential for slip-ups.
• Unwillingness to wait long enough to go through a demanding testing procedure: unprotected sex may become the norm very quickly in a relationship as a signal that the relationship is intensifying.
• Lapses in condom use before the HIV test window period is over.

Negotiation when testing uptake is low

A significant difference between the UK and Australia is the level of HIV testing within the gay population. In Australia nearly 90% of gay men are thought to have been tested at least once, but in the UK the proportion is closer to 50%. SIGMA Research has found that the proportion of gay men under 30 who have taken an HIV test is greater than the proportion of those over 35, and hence it may be reasonable to assume that this group may already be practising strategies of negotiated safety more frequently than their older counterparts (Weatherburn).

Yet it has also been reported that gay men over 35 are more likely to report unprotected sex with both regular partners and with casual partners (Coxon).

A consequence of the lower rates of HIV testing in the UK, in the view of some researchers, is the tendency of gay men to make assumptions about their own and their partner's HIV status. It should be noted that this `optimistic bias' is also likely to be a function of low HIV prevalence in comparison with the US and Australia. In such circumstances `negotiated safety' strategies need to be approached carefully in order that they do not promote more `guessing' about HIV status (based on the assumption that HIV testing is a community norm). This could occur if campaigns appear to communicate a primary message that unprotected sex within relationships is preferable to condom use, rather than focusing on the issue of how to make unprotected sex within a relationship as safe as possible.

Recent research in Edinburgh and Glasgow suggests that `negotiated safety' in the light of full knowledge about HIV status is still very rare. Research conducted in 1996 shows that only 16% of men questioned were having unprotected anal sex in a relationship were doing so in the light of knowledge about both partner's HIV status (Hart). This compares with around 60% in the Australian study discussed above.

Approaches to `negotiated safety' also need to take into account widespread psychological resistance to the suggestion that it is permissable for some gay men to stop condom use. The use of condoms has become inextricably linked with the idea of sexual intercourse for many gay men, and it can seem threatening to suggest that 100% condom use is no longer necessary when some people have worked so hard to achieve it. For many, 100% condom use may be part of what constitutes being a good, responsible gay man.