Voluntary testing to measure HIV prevalence in sexually transmitted disease clinics

Voluntary HIV testing was used to study the extent of HIV-1 infection in patients attending sexually transmitted disease (STD) clinics in England and Wales between 1985 and 1990. Homosexual and bisexual men and 10-20% of heterosexual men and women were invited to complete a study record and have an HIV-1 antibody test. The rate of newly diagnosed HIV-1 infection was higher in homosexual and bisexual men than in heterosexual clinic attenders. It was also higher in patients attending clinics in the South East compared with those attending clinics in other regions. From 1988 onwards, HIV infection was identified in heterosexual men and women who did not report behavioural risk factors associated with increased risk of HIV transmission. In the early years of the study, the proportion that agreed to complete a study record and have an HIV-1 antibody test was high in all groups. This proportion declined in those attending clinics in the South East, particularly among heterosexual men and women, less than 50% of whom agreed to take part in the study in 1989 and 1990. The decline in acceptance rate made voluntary testing unsuitable for monitoring trends in HIV infection. Unlinked anonymous HIV testing, which minimises the effect of participation bias, has become the method of choice for monitoring the prevalence of HIV infection.