Evaluation of an agar gel immunodiffusion test to detect infection of cattle with bluetongue viruses in Queensland, Australia.

An agar gel immunodiffusion (AGID) test to detect group-specific antibodies to infection of cattle by bluetongue viruses was evaluated using field collected sera in Queensland, Australia. The AGID test was compared to the serum neutralisation (SN) test used to detect serotype-specific bluetongue virus antibodies. The AGID test was found to be highly sensitive (95% confidence interval, 80.7-100%) but to have moderate specificity (95% confidence interval, 59.3-79.6%), relative to the SN test. The correlation between AGID and SN test results was 0.41 (P < 0.001). Using likelihood ratio estimates the chance (posttest probability) of an AGID test positive serum sample being positive (titre > or = 1:20) in the SN test was estimated to be 87% and 90% for 2+ and 3+ AGID test reactions, respectively. A serum sample was estimated to have a probability of 40% of being positive in the SN test if it gave a 1+ reaction in the AGID test. Results suggest that the AGID test should ideally be used as a screening test in serological surveys and surveillance programs. If the test is to be used for identification of individually infected cattle, a cutoff point of > or = 2+ should be employed.