Comparison of sensitivities and specificities of latex agglutination and an enzyme-linked immunosorbent assay for detection of antibodies to the human immunodeficiency virus in African sera.

Simple and rapid methods for the identification of Human Immunodeficiency Virus (HIV)-contaminated blood donations are urgently required for both routine screening in the developing world situations where blood bank facilities are inadequate or non-existent and in the developed world in situations where out-of-hours urgent screening is required (e.g. organ donors). This report critiques an investigation into the usefulness of a latex agglutination (LA) assay for the detection of antibodies to HIV in Zaire. 771 children and 711 adults from the internal medicine wards out patient clinic and blood bank at the Mama Yemo Hospital were tested. Samples from 537 individuals were confirmed to be anti-HIV-positive by repeated ELISA reactivity and Western blot. All diluted serum of whole blood from the 537 confirmed anti-HIV-positives were reactive (100% sensitivity). However undiluted bloods yielded 3 false negatives (99.4% sensitivity). Similarly serum or whole blood diluted 1 in 10 gave rise to the same results: 4 of 945 anti-HIV negatives were falsely reactive (specificity 99.6%) whereas testing undiluted whole blood was marginally worse with 6 false positives arising (specificity 99.4%). The investigators conclude that LA is highly sensitive and specific when diluted serum whole blood or diluted whole blood is used and that it will perform with the same accuracy as a commercial ELISA. However there are drawbacks with LA technology: 1) reading of the reaction is subjective and 2) there is no permanent record of the reaction. This type of approach still may have a role in certain circumstances and the ability to use whole blood as a test sample is certainly a point in favor of LA. Unfortunately the authors do not state if whole blood must be collected with anticoagulant.