Evaluation of a competitive enzyme-linked immunosorbent assay to detect infection of cattle in sentinel herds in Queensland, Australia with bluetongue viruses.

A competitive enzyme-linked immunosorbent assay (ELISA) was used to detect serogroup specific antibodies to bluetongue viruses. This test is commercially available and was evaluated with serially collected sera from 10 sentinel herds of cattle maintained in Queensland, Australia during 1994. Determination of an infection during the period of observation was based on the development of a serum neutralisation (SN) test titre > or = 1:8 to any one of 8 bluetongue virus serotypes known to exist in Australia. Using the inhibition value of 40% recommended by the manufacturer to classify cattle as exposed to bluetongue viruses, the ELISA was highly sensitive (100%; 95% confidence interval, 77.9-100%) and moderately specific (86.4%; 95% CI, 77.0-93.0%), relative to the SN test. An inhibition value of 70% maximised both sensitivity (100%, lower 95% CI, 77.9%) and specificity (93.2%, 95% CI, 85.2-98.0%) of the ELISA. The chance (posttest probability) of an animal from which a serum sample had an inhibition value > or = 70 % in the ELISA developing an SN test titre > or = 1:8 was 93.6% (95% CI, 86.4-97.9%). Investigations of the temporal development of ELISA and SN test reactions, showed that the ELISA detected exposure to bluetongue viruses significantly (P = 0.0156) earlier than the SN test. The bluetongue virus ELISA is a useful test in surveillance programs. False positive assay results make it inappropriate for monitoring and diagnosis, unless it is used in conjunction with the SN test.