Agreement study between two laboratories of immunofluorescence as a confirmatory test for human immunodeficiency virus type 1 antibody screening.

A total of 114 serum specimens from 76 blood donors, 21 patients with acquired immune deficiency syndrome or acquired immune deficiency syndrome-related complex, 7 multiply transfused patients, 3 hemophiliacs, and 7 others were tested for anti-human immunodeficiency virus type 1 (HIV-1) antibody by enzyme immunoassay (EIA) and Western blot (WB) and then blindly tested by immunofluorescence (IF), independently, in two separate laboratories. The IF technique used acetone-fixed HIV-1-infected E cells and uninfected HUT-78 cells mixed at a 1:3 ratio in one spot on a glass slide and uninfected HUT-78 cells (to assess nonspecific fluorescence) alone in a second spot. Of 114 serum specimens, 85 were repeat EIA positive, and 21 of these were WB positive. A total of 129 of 134 of the IF results (included were 20 duplicates) were identical between laboratories, for a Kappa agreement statistic of 0.93. All five IF results discordant between laboratories were EIA repeat positive and WB negative. Included in the study were eight WB-indeterminate sera, of which five blood donor serum specimens and one hemophiliac serum specimen were IF negative and two acquired immune deficiency syndrome serum specimens were IF positive. As a confirmatory test for HIV-1 antibodies, IF provided a faster alternative or supplementary test for confirming EIA results.