False negative results in enzyme linked immunosorbent assays using synthetic HIV antigens.

Abstract

The sensitivity of 12 commonly used anti-HIV-1/HIV-2 enzyme immunoassays was evaluated. The assays, each of which utilises at least one synthetic HIV antigen, were tested against a panel of 1092 specimens previously designated anti-HIV positive. In a total of 13 104 tests there were eight false negative results attributable to assay insensitivity: three were on two serum samples collected close to seroconverison and five on another serum specimen. These eight false negative results arose in seven different assays. Five other false results were attributable to technical error. This false negativity rate indicates that all of the assays performed adequately and leads to an estimate of one false negative result in a thousand tests in routine diagnostic practice. Because of the antigenic heterogeneity of HIV strains, similar evaluations would be required in several regions before this satisfactory level of sensitivity in anti-HIV assays incorporating synthetic antigens could be said to be universal.