Improved classification of recent HIV-1 infection by employing a two-stage sensitive/less-sensitive test strategy.

Abstract

Current serologic techniques for the classification of recent HIV-1 infection produce some misclassifications, and, together with the loss to follow-up of individuals, results in decreased enrollment of HIV-infected persons into appropriate intervention programs. We report on the development of a sensitive/less sensitive (S/LS) test strategy that includes a rapid assay to quickly identify persons most likely to have recent infection, followed by an enzyme immunoassay (EIA) with exquisite specificity. The Uni-Gold Recombigen HIV rapid assay (UG; Trinity Biotech, Dublin, Ireland) was procedurally-modified and calibrated as an LS test to differentiate recent (<133 days) from established HIV infections using 178 samples from persons with known dates of infection. This method correctly classified 83.0% of recent infections, but with a high misclassification rate of persons with established infection. By performing the rapid test followed by a modified S/LS EIA, the positive predictive value of the combined results for recent infections was increased to 100%. This two-stage testing algorithm can result in an increased efficiency for the enrollment of recent infection cases over a standard EIA S/LS method alone due to provisional enrollment during an initial testing visit, and because of an increased accuracy for identifying truly recent infections. We conclude that the rapid S/LS assay provides a tool for capturing recent infection cases quickly and is particularly valuable in resource-limited settings, and that the two-stage strategy provides a more accurate identification of persons with recent HIV infection.