Sensitivity and specificity of two serological tests for the detection of ovine paratuberculosis

Objective To determine the sensitivity and specificity of an absorbed ELISA and an AGID test for the detection of clinical and subclinical paratubercu losis in sheep. Design By testing a panel of sera from 1257 Australian Merino and crossbred sheep greater than 1 year of age, of which 1137 sheep were not infected wi th Mycobacterium avium subsp paratuberculosis and 120 sheep had paratubercu losis. Procedure Sera were collected from 457 sheep in Victoria and 800 sheep in W estern Australia. Presence of M a paratuberculosis infection in Victorian s heep was determined by histological examination of intestinal tissues, wher eas sheep from Western Australia were presumed to be free of Johne's diseas e. The ability of an absorbed ELISA to discriminate between infected and un infected sheep was described by test sensitivity and specificity, the distr ibution of ELISA OD, and the area under a receiver operating characteristic curve. Results The absorbed ELISA had a specificity of 98.2 to 99.5% (CI) and a se nsitivity of 35 to 54% (CI). In sheep from infected flocks in Victoria, the AGID test had a specificity of 99 to 100% (CI) and a sensitivity of 38 to 56% (CI). The sensitivity of serological tests was higher in sheep with a b ody condition representative of the lower quintile of their flock of origin . Conclusion The AGID test and absorbed ELISA are useful tests for the detect ion of ovine paratuberculosis. Although the tests had a similar accuracy, t hey detected different subpopulations of infected sheep with only moderate overlap. The AGID test had a higher specificity than the absorbed ELISA.