Send to

Choose Destination
Bull World Health Organ. 1991;69(6):753-60.

Comparative evaluation of 36 commercial assays for detecting antibodies to HIV.

Author information

Department of Microbiology, Institute of Tropical Medicine, Antwerp, Belgium.


Summarized are the results of an assessment of the major operational characteristics of 36 commercially available assays for detection of antibodies to human immunodeficiency virus type 1 (HIV-1) and/or type 2 (HIV-2). For this purpose, 20 enzyme-linked immunosorbent assays (ELISAs), 11 simple immunoassays with visual reading, four supplemental assays, and one discriminatory assay were assessed using a panel of 537 sera (65% of which were of African, 26% of European, and 9% of South American origin); the prevalence of HIV-1 was 39.1% and of HIV-2, 15.7%. The following operational parameters of the assays were investigated: ease of performance; suitability for use in small blood collection centres; sensitivity and specificity; positive predictive values at different prevalences; inter-reader variability for simple assays whose results were read visually; the proportion of indeterminate results; and, for some of the ELISA assays, delta-values, as quantitative measures of sensitivity and specificity. The results will be of use to health policy decision-makers, managers of national AIDS prevention and control programmes, directors of blood banks, and laboratory specialists in the selection of appropriate HIV antibody assays.


Microbiologists at the Institute of Tropical Medicine in Antwerp, Belgium used sera from 537 people (65% Africans, 26% Europeans, and 9% South Americans) to compare commercial assays for detecting HIV antibodies. These assays included 20 ELISAs, 11 simple assays, 4 supplemental assays, and 1 discriminatory assay. HIV-1 seroprevalence was 39.1% and 15.7% for HIV-2. Basically the sensitivity of the assays were very good and equal, except the sensitivity of the Peptide HIV ELISA assay was considerably lower. In fact, on the most part, the sensitivities were higher than the specificities. But only 4 assays had significantly lower specificities than sensitivities. A high number of false positive reactions occurred in the African sera with these 4 assays which emphasizes the need to use African sera to evaluate HIV antibody kits. The higher the positive and negative Delta values the more likely the assay can accurately identify antibody positive and antibody negative sera respectively. The Elavia Mixt (HIV-1+2) and the Wellcozyme HIV-1+2 had the highest positive Delta value while Abbott recombinant HIV-1/HIV-2 EIA had the lowest positive value. Du Pont HIV-1/HIV-2 had the lowest negative value. No significant difference existed in determining sensitivity and specificity by visually reading the results between the simple assays. Interreader variability ranged from 0.8-31.7% with Recodot and Genie HIV-1 and HIV-2 having the highest variability. Further no significant difference existed in sensitivity and specificity between the ELISAs. The Ancoscreen supplemental assay did not have high sensitivity and specificity when used on African sera. Further INNO-LIA HIV-1/HIV-2 Ab test detected both HIV-1 and HIV-2 antibodies at the same time and ranked lower in indeterminant results than the Western blot.

[Indexed for MEDLINE]
Free PMC Article

Supplemental Content

Full text links

Icon for PubMed Central
Loading ...
Support Center