Relevance of antibody content and test format in HIV testing of pooled sera.

Abstract

To determine the effects of HIV-1 antibody level and test-format characteristics on testing pooled sera. This study was designed with a laboratory exercise followed by test observations on serosurveillance samples. Sera with low, medium and high (n = 22, 12 and 20, respectively) antibody titers were pooled with HIV-1-negative sera and tested with two enzyme-linked immunosorbent assays (ELISA) and a particle agglutination test. The same kits were used to test single and pooled (batches of five, 10 and 20) samples collected from 3000 blood donors and sex workers. These samples were then seeded with 50 varying antibody-containing sera and similarly tested. Initial reactivities, sensitivities, and specificities for all test kits were calculated and compared. In the laboratory exercise, all reactive pools of five were detected. False-negative pools in batches of 10 and 20 with low antibody titers were noted with one or both ELISA, but not with the particle agglutination method. Testing 3000 samples revealed three confirmed reactive samples and 100% sensitivity/specificity for all kits, for both single and pooled sera testing. Increased initial reactivity (IR) was noted for the two ELISA. Examinations of pools of the seeded 3000 samples with the two ELISA showed false-negative reactivity with pools of 10 and 20 when pools contained low antibody sera (sensitivities and specificities of 92-97.9% and 98.1-100%, respectively). Again, increased IR was seen with the ELISA. False-negative pool and increased IR was not seen with the agglutination test (sensitivity/specificity 100%). We recommend the use of the particle agglutination assay for testing pooled sera of batches of 20 or less. Components of reactive pools should then be tested and reactive samples should undergo supplementary testing. Pooled samples tested by ELISA should not exceed five per batch. Retesting of reactive pools, testing of its components, and supplemental test(s) of reactive sera should then follow. The optimum pool size for most laboratories is five, with the best technical and economic performance seen with the particle agglutination assay.