Unlinked anonymous HIV screening of pregnant women in a low-prevalence population

Unlinked anonymous screening was optimized and used for a survey of HIV prevalence in Finland. In 1993, 66,170 serum samples, covering more than 99% of pregnant women in the country were tested in pools of 5 sera. The pools tested were coded but it was possible to trace the region from where the samples had been collected. A total of 5 positive samples were found corresponding to a prevalence of 1/13,000 (0.01%) which is in accordance with prevalence estimates from earlier serosurveys and the national HIV registry. Information from voluntary regional HIV-testing programmes for pregnant women indicated that probably at least 3 of these positive cases had been identified by voluntary testing. Nine commercial EIAs were evaluated to investigate the effect of pooling on sensitivity and specificity of anti-HIV detection. Only 3 EIAs detected all pools with known positive samples. Individual sera from repeatedly positive pools were tested using the same EIA and positive results were confirmed by Western blot. This algorithm saved more than 80% of the cost compared to the conventional test algorithm used in diagnostic laboratories. The savings were mainly due to the reduced number of primary and supplemental tests.