Comparison of six methods for the detection of antibody to cytomegalovirus.

Five commercial assays were compared to a standardized complement fixation (CF) test for the detection of antibody to cytomegalovirus. Two hundred and thirty serum specimens were analyzed. In addition, nine pairs of acute- and convalescent-phase sera were tested by two of the commercial assays. The assays were compared as to sensitivity, specificity, and positive and negative predictive value, as well as incidence of false-positive and -negative results. Samples which did not agree in all the assays were retested and tested with an indirect fluorescent-antibody assay. Of 228 specimens, 103 (45.2%) were positive by CF. Of the 230 samples, 2 (0.9%) were inconclusive by CF and readable in the other assays. Of the 230 specimens, 97 (42.2%) were positive by an enzyme immunoassay (EIA; Litton Bionetics), 100 (43.5%) were positive by a second EIA (Abbott Laboratories), 104 (45.2%) were positive by a third EIA (M. A. Bioproducts). One hundred and eight (47.0%) were positive by indirect hemagglutination (IHA; Cetus Corporation), and 110 (47.8%) were positive by latex agglutination (LA; Hynson, Westcott and Dunning). Sensitivity and specificity were similar with all the assays (93 to 100%). The greater numbers of positive results by IHA and LA were confirmed by repeat CF testing at less than 1:8 dilution, and by indirect fluorescent-antibody assay. Acute- and convalescent-phase serum pairs showed a significant rise in antibody titer when tested by anticomplement immunofluorescence, IHA, and LA. There was good agreement among the assays, with LA having the highest sensitivity.