A comparison of diagnostic assays for the detection of type A swine influenza virus from nasal swabs and lungs.

Nasal swabs and lung samples from pigs experimentally infected with H1N1 swine influenza virus (SIV) were examined for the presence of SIV by the indirect fluorescent antibody assay, immunohistochemistry, cell culture virus isolation, egg inoculation, and 2 human enzyme immunoassays (membrane enzyme immu- noassay, microwell enzyme immunoassay). Egg inoculation was considered to be the gold standard for assay evaluation. The 2 human enzyme immunoassays (EIA) and egg inoculation agreed 100% for the prechallenge nasal swabs. Agreement on SIV identification in nasal swabs with egg inoculation following challenge was considered to be good to excellent for membrane EIA (kappa 5 0.85) and microwell EIA (kappa 5 0.86). Agreement on SIV identification in lung tissue with egg inoculation following challenge was good to excellent for membrane EIA (kappa 5 0.75), fair for microwell EIA, fluorescent antibody, and cell culture virus isolation (kappa 5 0.48, 0.64, 0.62, respectively), and poor for immunohistochemistry (kappa 5 0.36). No assay was 100% accurate, including the ''gold standard,'' egg inoculation. In light of this information, it is important to consider clinical signs of disease and a thorough herd history in conjunction with diagnostic results to make a diagnosis of SIV infection. Swine influenza virus (SIV) has recently become recognized as an important pathogen of swine, incrim- inated along with the porcine reproductive and respi- ratory syndrome virus and Mycoplasma hyopneumon- iae as a cause of porcine respiratory disease complex. In 1998, tissues from over 3,000 cases of respiratory disease in swine were examined at the Veterinary Di- agnostic Laboratory at Iowa State University. Evi- dence of SIV infection was detected in 350 cases (P. Halbur, personal communication). The number of SIV cases diagnosed in 1998 was 4 times greater than those diagnosed in 1993. Contributing to the rise in the num- ber of SIV infections was the recognition, in late sum- mer 1998, of the emergence of H 3 N2 SIV. The initial case was identified in North Carolina, but by the end of the year, H3N2 SIV infections had been identified in Iowa, Minnesota, and Texas. 26 Swine influenza viruses submitted to the National Veterinary Services Labo- ratories for serologic typing between September 1, 1998 and February 1, 2000 have been identified as