Evaluation of simple rapid HIV assays and development of national rapid HIV test algorithms in Dar es Salaam, Tanzania.

ABSTRACT

BACKGROUND: Suitable algorithms based on a combination of two or more simple rapid HIV assays have been shown to have a diagnostic accuracy comparable to double enzyme-linked immunosorbent assay (ELISA) or double ELISA with Western Blot strategies. The aims of this study were to evaluate the performance of five simple rapid HIV assays using whole blood samples from HIV-infected patients, pregnant women, voluntary counseling and testing attendees and blood donors, and to formulate an alternative confirmatory strategy based on rapid HIV testing algorithms suitable for use in Tanzania. METHODS: Five rapid HIV assays Determine HIV-1/2 (Inverness Medical), SD Bioline HIV 1/2 3.0 (Standard Diagnostics Inc.), First Response HIV Card 1-2.0 (PMC Medical India Pvt Ltd), HIV1/2 Stat-Pak Dipstick (Chembio Diagnostic System, Inc) and Uni-Gold HIV-1/2 (Trinity Biotech) were evaluated between June and September 2006 using 1433 whole blood samples from hospital patients, pregnant women, voluntary counseling and testing attendees and blood donors. All samples that were reactive on all or any of the five rapid assays and 10% of non-reactive samples were tested on a confirmatory Inno-Lia HIV I/II immunoblot assay (Immunogenetics). RESULTS: Three hundred and ninety samples were confirmed HIV-1 antibody positive, while 1043 were HIV negative. The sensitivity at initial testing of Determine, SD Bioline and Uni-Gold was 100% (95% CI; 99.1-100) while First Response and Stat-Pak had sensitivity of 99.5% (95% CI; 98.2-99.9) and 97.7% (95% CI; 95.7-98.9), respectively, which increased to 100% (95% CI; 99.1-100) on repeat testing. The initial specificity of the Uni-Gold assay was 100% (95% CI; 99.6-100) while specificities were 99.6% (95% CI; 99-99.9), 99.4% (95% CI; 98.8-99.7), 99.6% (95% CI; 99-99.9) and 99.8% (95% CI; 99.3-99.9) for Determine, SD Bioline, First Response and Stat-Pak assays, respectively. There was no any sample which was concordantly false positive in Uni-Gold, Determine and SD Bioline assays. CONCLUSION: An alternative confirmatory HIV testing strategy based on initial testing on either SD Bioline or Determine assays followed by testing of reactive samples on the Determine or SD Bioline gave 100% sensitivity (95% CI; 99.1-100) and 100% specificity (95% CI; 96-99.1) with Uni-Gold as tiebreaker for discordant results.