Comparative age-related responses to serological and faecal tests directed to Mycobacterium avium paratuberculosis ( Map) in French dairy goats

Abstract

The objectives of the present study were to estimate sensitivity and specificity of 6 different tests directed to Mycobacterium avium paratuberculosis (faecal culture, microscopic examination, Agar Gel Immunodiffusion (AGID) and 3 commercial ELISA tests, named A, B and C) in different age categories and to assess the degree of agreement between responses thus obtained. The study was carried out in 12 French commercial dairy herds, which were selected in a population of herds with unknown paratuberculosis status. In each herd, a sample of goats without clinical signs was randomly selected according to 4 age categories. Responses of 412 goats with concomitant results for the 6 tests were analysed. In the 4 age categories, the specificity and the sensitivity of the 6 tests were estimated using Bayesian methods with conditional dependence between tests. Agreement among the results obtained for these tests was determined by calculating the kappa statistic. With the exception of AGID, the proportions of positive responses for each test differed significantly between herds (range 0–27.5%). With the exception of scopy, the proportions of positive responses for each test differed significantly between age categories. Sensitivities and specificities of the 6 tests were very different according to age categories. Specificities were higher than sensitivities. The highest values in sensitivity were obtained in goats between 2 and 3 years, with the best value for ELISA C (76.6%, 95% CI of 46.1–96.1%). In the other age categories, sensitivities of the 6 tests were too low to have practical use. Only scopy, faecal culture and ELISA B have positive responses in category 1 (goats of less than 1 year). Agreements between the different pairs of tests were low, except between the 3 ELISA tests which were in “good” or “quite good” agreement. This study showed that detection of infection with Map was efficient only in goats between 2 and 3 years, with an ELISA test. For goats in other age categories, no test was sensitive enough to detect goats infected with Map. In a Map infection control program, ELISA tests, performed on goats between 2 and 3 years, can be used to define herd status and to estimate prevalence but they should not be used at goat level to select goats for culling, due to the low positive predictive value.