To test or not to test, could you repeat the question?

We aimed to investigate clinical practice in the offering of HIV tests and subsequent uptake in a central London genitourinary medicine (GUM) clinic. A random sample (n = 330) of attenders at 3 inner-London GUM departments was surveyed. Reasons for and rates of offering of HIV tests were recorded and analysed in relation to demographic, risk group information and uptake. The results were integrated with the latest unlinked, anonymous seroprevalence data for the clinic. After exclusion of patients known to be HIV-positive or to have recently undergone HIV testing, HIV tests were offered to 96% of homo/bisexual men, 55% of heterosexual men and 60% of heterosexual women. Comparison with anonymous HIV seroprevalence data showed an inverse relationship between seroprevalence rates for heterosexual men/women (2.5% vs 1%) and rates of HIV test offering. A lack of research into the policy of offering HIV tests may have resulted in inconsistencies in practice. An evidence based policy should offer HIV tests in line with seroprevalence.