Before running a trial, it is essential to know something about HIV prevalence and incidence, to understand what may motivate people and couples to take risks, to volunteer for the trial and to stay with it. It is necessary to find out what the likely social consequences of being a volunteer will be, so volunteers can be fully informed and advised.

A further level of feasibility is about the ability of a community to participate in the decision making about running the trial and for volunteers to understand and support the trial protocol. These questions should become easier to answer when they can be posed in relation to specific vaccine candidates and easier still when there has been previous experience of running a trial.

In Europe, there are three main populations in which a trial could be based. The first and most accessible population consists of homosexual and bisexual men attending GUM clinics and testing HIV negative, particularly in cities such as London.

This is the only group for which there is really firm evidence on continuing HIV incidence in the UK (Miller 1995). Among homosexual and bisexual men under the age of 30 attending for repeat HIV tests at selected London GUM clinics between 1988 and 1994, incidence seems to have been between 7 and 10 per cent. (Incidence was substantially lower in older gay men, at around 3.3 per cent, and available evidence for clinics outside London suggests lower figures still.)

In reading these figures, it is important to note that the risk here is determined retrospectively. To recruit people at this level of risk into a vaccine trial would mean knowing in advance who was going to return to the clinic for repeat testing. So prospective studies may still be needed, to prove the feasibility of recruiting and maintaining contact with clinic attenders at genuinely high risk of HIV infection.

There has been one community-based survey of gay men’s readiness to participate in HIV vaccine trials, which was limited by the absence of any product with a known history which anyone is proposing to test in the population surveyed. It nonetheless concluded that a large number of men would need to be asked, to find men prepared to participate (Bolding).

Similar questions were raised, prior to VaxGen’s current trials, as to whether they would be able to recruit and retain volunteers for a product for which there were low expectations of efficacy. In the event, VaxGen proved its doomsayers wrong.

It might also be possible to identify subgroups of gay men at particularly high risk by selectively recruiting HIV-negative gay men recently treated for herpes (Tabet 1996) or other sexually transmissible infections. However, numbers in such categories are likely to be small. Such populations might equally well be recruited into trials of other prevention technologies, such as the preventive use of antiretroviral drugs.

The second population consists of couples, whether same-sex or opposite-sex, where one partner is HIV-positive and the other is HIV-negative.

The size of this population is estimable from the number of people living with HIV and known to services and the proportion of those people who have a continuing sexual relationship with someone who is HIV-negative.

These 'serodiscordant' couples are likely to become increasingly visible to HIV services as the uptake of HIV testing increases as a result of perceived benefits from antiviral treatments and as the number of people living with HIV rises with reduced death rates due to treatments.

In Britain, this is particularly likely to apply to members of refugee and migrant communities with links to high-prevalence countries. Present levels of knowledge, HIV testing and access to medical services are among the lowest of any of the categories of people affected by HIV; a situation that many people are working to change.

Seronegative partners are likely to become particularly well educated about HIV, to be realistic about their own risks, and to be at particular risk of acquiring treatment-resistant strains of HIV.

For all of these reasons, it is argued that seronegative partners should be recruited to vaccine and other prevention studies.

Above all, when vaccines are being developed in Europe for future use in Africa, it would seem foolish to overlook the possibility of engaging African men and women affected by HIV in Eyrioe in the process of developing those vaccines.

In Britain, rates of new infection among injecting drug users appear to be very low although this is not necessarily the case in other parts of Europe.

The total number of HIV cases reported annually in the UK and attributed to injecting drug use has been stable for some years between 175 and 200. In contrast to the pattern among gay men, there is evidence that newly diagnosed cases are becoming older, implying that there are even fewer new infections.

In contrast to the situation with gay men and GUM services, there is no evidence that injecting drug users attracted to services are at higher HIV risk than those who are not attracted to services. It appears from the anonymised testing programmes based in drugs services that there is little or no seroconversion among those who continue to attend them.

It might still be possible to identify subgroups at much higher risk, such as injecting drug users recently infected with hepatitis C or seronegative partners of people known to be seropositive.

In conclusion, the logical place to start any education programme and to begin asking questions about the feasibility of vaccine trials is among gay and bisexual men, especially those living in cities and towns with high HIV prevalence and attending GUM services. And, secondly, among members of African refugee communities, whose understanding and support for global efforts to develop vaccines could be invaluable.