Sensitivity and specificity of standard and rapid HIV-antibody tests evaluated by seroconversion and non-seroconversion low-titre panels.

BACKGROUND AND OBJECTIVES: The aim of this study is to compare the relative sensitivity and specificity of commercial HIV-antibody assays using seroconversion, non-seroconversion panels, and negative blood donor samples.
MATERIALS AND METHODS: We evaluated the sensitivity of five standard ELISA HIV-antibody assays: Vironostika HIV Uni-Form II, Abbott recombinant HIV-1/HIV-2 third-generation EIA, Biotest Anti-HIV-1/-2 recombinant, Recombigen HIV-1/ HIV-2 EIA and Wellcozyme HIV 1 + 2 (VK54/55), and three rapid screening tests, Capillus HIV-1/HIV-2, Abott Test Pack HIV-1/HIV-2 third-generation EIA, and Sensy-Test HIV 1/2. All tests were assessed using four panels of plasma samples obtained from individuals who were seroconverting and a low-titre HIV-antibody panel of samples. Specificity of the standard screening tests was determined on 3.500 HIV-antibody-negative blood donor samples.
RESULTS: There was no statistically significant difference in sensitivity between the five standard ELISA tests. One of these tests was significantly less specific than the others. The standard ELISA tests detected all the low-titre HIV-antibody-positive samples. Two of the rapid screening tests were significantly less sensitive on the seroconversion panels and all three tests failed to detect at least one of the positive samples in the low-titre panel.
CONCLUSIONS: The additional risk of using one or other of the standard ELISA tests under review of not detecting all HIV-positive units of blood is not statistically significant. Using some of the rapid screening tests will, however, add a significant additional risk. A rapid screening test should therefore be adopted only after careful consideration of the effect of a possible lack of sensitivity on the safety of the blood supply.