S. Michele Owen, C Yang, T Spira, CY Ou, CP Pau, B Parekh, F Cowart, D. Kuehl, S Kennedy, D Rudolph, D Candal , W Luo, N Delatorre, S Masciotra, T Barnett, R Lal and JS McDougal,
Laboratory Branch, NCHSTP, Centers for Disease Control and Preveniton
Background/Objective: The current algorithm for diagnosis of HIV infection involves a screening EIA followed by a supplemental test, Western Blot (WB) or Immunofluorescence (IFA) for confirmation. Given the inherent problems associated with WB (cost, subjectivity, indeterminate results) and the increased number of FDA approved tests for HIV (Nucleic Acid Test (NAT), rapid test, improved EIAs), a study was undertaken to compare commercially available HIV diagnostic tools to address whether new algorithms for HIV screening/diagnosis are appropriate.
Methods: Tests used in the study included 6 EIAs, 4 Rapid tests and 2 NAT based tests as well as 2 WBs. 1002 plasma samples obtained from Boston Biomedica (BBI) that represents U.S. blood/ plasma donors, and 268 non-U.S. samples were used. All or a subset of the samples were evaluated with each test and sensitivity and specificity were calculated relative to the current standard of EIA and WB. Furthermore, the testing data were used to assess potential new algorithms for HIV screening/diagnosis.
Results: The range of sensitivity observed in the study for all tests was from 92.6% to 99.4 % and the specificity observed ranged from 95.8%- 99.4%. When serological testing data were combined with NAT data, to address the question if NAT testing can eliminate or reduce WBs (as described in the proposed Blood Blank Screening algorithm) the number of indeterminate WB was reduced. However, samples were identified that yielded discordant results when serological and NAT- based tests were compared.
Conclusions: All of the commercially available, FDA approved, assays for detection of HIV gave comparable levels of sensitivity and specificity. These data suggest that the proposed Blood Bank Screening algorithm that incorporates NAT based testing would likely decrease the number of indeterminate WB results. However, our data indicates that NAT testing can not completely replace WB as samples were identified that were serology positive and NAT negative.
Presenter: S. Michele Owen, CDC NCHSTP DHAP smo2@cdc.gov

Last Update: April 8, 2005