Since the introduction in the mid-1980s, HIV testing has gradually improved both in terms of sensitivity and specificity. The so-called fourth generation of tests, combined HIV antigen/antibody assays, has now been introduced. This study compares three automated combined assays with older third-generation antibody assays in large-scale screening. Serum samples from routine screening of blood and plasma donors and clinical samples were investigated for specificity evaluation. Three fourth-generation combination assays from one manufacturer were compared with three older third-generation antibody assays from the same manufacturer. More than 40 000 samples per assay were included. For sensitivity, selected panels of confirmed HIV-1- and HIV-2-positive samples as well as seroconversion samples (HIV-1) from commercial panels and also from patients who appeared during the evaluation were used. The specificities of the fourth-generation tests were 99.91% (AxSYM), 99.95% (ARCHITECT) and 99.97% (PRISM) after repeated testing. Some specificity variation between reagent batches was observed. All HIV-1-positive samples were reactive by the three fourth-generation systems. HIV-1 seroconversion samples and panels were reactive earlier than by antibody-only tests. As for HIV-2 samples, AxSYM failed to detect one (n = 40), whereas PRISM and ARCHITECT detected all (n = 16 for PRISM and n = 52 for ARCHITECT). The new HIV antigen/antibody combination assay systems were found to have high sensitivity and specificity. The instruments provided a rational and easy way of testing at large scale.

Since the start of screening in 1994, there has been no increased seroprevalence of HTLV-1/2 among blood donors in Sweden. Seroprevalence among Swedish IVF patients is 10 times higher than among blood donors. This finding is comparable to a 2003 European seroprevalence study of pregnant women in 7 countries. However, the possibility that the IVF group includes individuals with a higher risk of acquiring sexually transmitted infections, including HTLV, than the general population cannot be ruled out.

In resourced settings, where the reverse sequence algorithm is preferred for screening, an automated treponemal immunoassay for initial screening subsequently followed by the TrepSure test or TP-PA assay as a second treponemal assay appear highly effective. Finally, a quantitative highly sensitive non-treponemal assay, e.g. the Macro-Vue RPR Card test, could then be used as a supplementary test to evaluate activity of the syphilis infection.