Use of an enzyme-linked immunosorbent assay for detection of Treponema hyodysenteriae infection in swine.

Discriminate analysis was used to evaluate the enzyme-linked immunosorbent assay (ELISA) for the detection of anti-Treponema hyodysenteriae antibodies in experimentally and naturally infected swine. In trial 1, 26 pigs were randomly divided into three groups (naturally infected, n = 8; experimentally infected, n = 11; and noninfected, n = 7), and samples were collected for 10 weeks. For trial 2, 31 pigs were randomly divided into two groups (naturally infected, n = 22; and noninfected, n = 7), and samples were collected for 20 weeks. Rectal swabs for T. hyodysenteriae isolation were collected daily, and fecal samples for isolation of Salmonella spp. were collected weekly. Serum samples for ELISA evaluation were collected biweekly (trial 1) or weekly (trial 2). Results of discriminate analysis indicated that the ELISA correctly identified 90% or more of the individually infected pigs at prior probabilities of infection ranging from 60 to 90%. The test correctly identified noninfected pigs at a lower rate (61 to 92% range). The mean ELISA titers of naturally infected pigs without clinical signs were not significantly different (P less than 0.05) from the titers of both groups of experimentally infected pigs. Mean ELISA titers of naturally infected pigs without clinical signs were significantly greater than the mean titers of naturally infected pigs with clinical signs. Naturally infected pigs with clinical signs had a mean ELISA titer that was significantly greater than that of noninfected pigs and significantly less than the mean titers of the experimentally infected pigs without clinical signs and the naturally infected pigs without clinical signs.