Standardisation of subjectively scored HIV immunoassays: developing a quality assurance program to assist in reproducible interpretation of results using an anti-HIV particle agglutination assay as a model.

Immunoassays such as particle agglutination assays, rapid tests and western or line blots are scored or read subjectively. These readings display intra- and inter-reader variability, as well as intra- and inter-laboratory variability. In the present study the consistency of scoring was assessed between readers both within and between two groups of scientists using the Serodia anti-HIV particle agglutination assay as an example of an assay scored subjectively. An anti-HIV positive sample in eight serial dilutions made to yield a full range of results expected for the assay was presented 12 times (96 test wells). Each dilution was placed randomly in a plate and tested with the Serodia anti-HIV particle agglutination assay then photographed. Participants in the two groups each scored the photographed plate independently and twice, 2 h apart. Each well was assigned a status (the consensus result of the four most experienced Australian readers) and each participant's results were compared with this status. The average percentage of wells assessed as 'correct' for the Group A participants was 86% (range 56-98%) and for the Group B participants was 67% 'correct' (range 46-88%). In general, strongly positive and negative wells were scored 'correctly'. The highest variations between scores were seen in the borderline positive dilutions +/- region. A quality assessment program based on the method used to obtain these results will be instituted in order to improve the consistency of scoring assays read subjectively.