Abstract
Five different antigens were evaluated in enzyme-linked immunosorbent assay (ELISA) tests for the detection of avian pneumovirus (APV) antibodies. Two of the 5 antigens were prepared from recent APV isolates from Minnesota. The 2 older isolates were passage 63 of a strain currently used as a live, attenuated vaccine and a Colorado strain isolated for the first time in the United States and currently used in an ELISA test. The fifth antigen is based on an APV recombinant N-protein. Basic parameters and positive-negative threshold of the assays were established for all 5 antigens on the basis of data obtained by testing 46 known negative and 46 known positive serum samples. Subsequently, 449 field samples were tested by all 5 ELISAs. The optical density difference (ODD) was calculated by subtracting optical density of the sample in the negative antigen well from that in the positive antigen well. In the current ELISA test based on the Colorado strain, an ODD of 0.2 is considered to be the cutoff value to classify samples as negative or positive. In this study, however, use of different cutoffs, based on ODD of negative control plus 3 SD or values estimated from Receiver operating characteristic analysis, was considered to be more appropriate for the various antigens used. Overall person-to-person and day-to-day variability was found to be large for all tests using either ODD or sample to positive ratio to report results. In addition, results suggest that antigenicity of the APV isolates in the United States has not changed between 1997 and 2000.