Human immunodeficiency virus surveillance at a sexually transmitted disease clinic in Copenhagen.

During four study periods, each of 3-4 weeks' duration in 1987 and 1988, all patients attending the Copenhagen outpatient Venereal Disease Clinic were encouraged to have a test for human immunodeficiency virus (HIV) antibody. Anonymous testing or testing for research purposes only without any data was also offered. Patients known to be HIV antibody-positive were not re-tested. A total of 1753 patients (1118 men, 635 women) were counselled and testing recommended. Testing was refused by 255 patients, 57 of whom had been tested (all negative) within the previous 3 months. HIV serology was therefore unknown in 11.7% (198 of 1696 persons), including 120 of 922 heterosexual men (13%), 16 of 159 homo/bisexual men (10%) and 62 of 615 women (10%). There was no substantial difference between the four study periods in the numbers refusing the test. Sixteen of the 1498 patients tested were HIV antibody-positive, including one woman and 3 men with a history of intravenous drug abuse and 11 homo/bisexual men; one heterosexual man with HIV antibodies had no known risk behaviour. It is concluded that HIV testing with informed consent at a sexually transmitted disease clinic will not include all persons, and therefore will not exclude a self-selection bias. Surveillance studies using unlinked HIV testing are therefore necessary and it is recommended that they should be performed in Denmark. In this study, unlinked HIV testing of the serum specimens taken for syphilis serology would have decreased the percentage of 'non-attenders' from 12.4% to 0.8% during the final study period.