Abstract Body

Background: In resource-limited settings, the World Health Organization recommends using two HIV rapid tests for HIV diagnosis. If the test results are discordant, a third tie-breaker test is often used to determine HIV status. We evaluated the HIV status of adults with discordant rapid tests and evaluated the performance of different tie-breaker tests. Samples were obtained from a population-based survey (NIHM Project Accept [HPTN 043]). The frequency of discordant rapid tests in the survey was 2.1% in Tanzania and 0.3% in South Africa.

Methods: Plasma samples (N=173) were classified as HIV positive or HIV negative using a rigorous testing algorithm adapted from the HIV testing algorithm currently recommended by the United States Centers for Disease Control and Prevention. The testing algorithm included two fourth-generation immunoassays (BioRad Combo assay and Abbott Combo assay), an HIV-1/HIV-2 differentiation assay, and two HIV RNA tests. Selected samples were further characterized using a panel of laboratory assays that included a qualitative assay for antiretroviral (ARV) drug detection. Sensitivity, specificity and accuracy were determined for the following tie-breaker tests: a third-generation HIV rapid test, a third-generation enzyme immunoassay (EIA), the BioRad Combo assay, and the Abbott Combo assay.

Results: Twenty-nine (16.8%) of the 173 samples were classified as HIV positive. One sample was from an acute infection. Three (10.7%) of the remaining 28 HIV positive samples had detectable HIV RNA. ARV drugs were detected in only one sample (viral load: 1,280 copies/ml). Sensitivity of the tie-breaker tests was: HIV rapid test: 8.3%; third-generation EIA: 24.1%; BioRad Combo assay: 93.1%; Abbott Combo assay: 96.6%. Specificity of the tie-breaker tests was: HIV rapid test: 96.1%; third-generation EIA: 98.6%; BioRad Combo assay: 88.2%; Abbott Combo assay: 90.3%. Accuracy of the tests ranged from 79.5-91.3%.

Conclusions: Significant differences were observed in the performance of tie-breaker tests for determining HIV status in individuals with discordant HIV rapid test results. Most HIV infections were missed using an HIV rapid test or third-generation EIA as a tie-breaker. In this cohort, most HIV-infected individuals with discordant HIV rapid test results were virally suppressed in the absence of ARV drugs. More rigorous testing approaches, rather than a single tie-breaker assay, may be needed to resolve the HIV status in individuals with discordant rapid test results.