Survey on paratuberculosis prevalence in dairy herds of the Lombardia Region (Italy) Arrigoni, N. 1 , Belletti, G.L. 1 , Cammi, G. 1 , Losini, I. 1 , Taddei, R. 1 and Tamba M. 2 1 Istituto Zooprofilattico Sperimentale della Lombardia e dell’Emilia-Romagna, Centro di Referenza Nazionale per la Paratubercolosi – Piacenza, Italy 2 Istituto Zooprofilattico Sperimentale della Lombardia e dell’Emilia-Romagna, Centro Emiliano-Romagnolo Epidemiologia Veterinaria- Bologna, Italy 391 dairy herds, stratified for size and province, were selected with the purpose of estimating paratuberculosis herd prevalence in the Lombardia region, a Northern Italian Region where 15.000 dairy herds and 45% of the national milk production are concentrated. The prevalence of infected herds was estimated by a commercial ELISA test (Institut Pourquier) on individual blood samples of 38.487 dairy cattle over 12 months. The percentage of herds showing at least one seropositive animal was 43.7%, while the percentage of seropositive cows was 2.6%. On the basis of the performances (sensititity 45%, specificity 99%) of ELISA test on blood, the herds that didn’t reach a herd specificity of 95% were tested for confirmation by fecal culture on the seropositive cows. The resulting estimated herd prevalence was 19.2%. Introduction Paratuberculosis, an infectious and contagious disease caused by Mycobacterium avium subsp. paratuberculosis (Map), is reported all over the world. In Europe, the herd prevalence varies between 7 and 55% (FIL-IDF, 2001). Besides causing economic losses to dairy and beef herds, Map is regarded as a possible infectious agent of Crohn’s Disease in man. Given that the herd prevalence estimation is fundamental for planning a control programme, we have carried out a survey on an Italian northern region, Lombardia, in which 42% of the overall Italian milk production is concentrated. Material and methods In the period between October 2003 and March 2005, we carried out a serological survey for the detection of Map antibodies in blood samples of 38,478 cows over one year of age, belonging to 391 dairy herds of the Lombardia Region; the tests for paratuberculosis were carried out on the samples collected by the Veterinary National Health Service for Brucellosis and Leukosis eradication programmes. The 391 sampled herds, corresponding to nearly 3% of the 15,111 dairy herds of the Lombardia Region, were stratified for geographical distribution and size. In the selected herds, the number of dairy cows over 12 months varied between 10 and 584. The blood samples were submitted to a screening analysis using a commercial ELISA kit (Institut Pourquier, Montpellier, France); all the reactive (positive and inconclusive) samples were analysed using a verification ELISA test (Institut Pourquier, Montpellier, France). It was established that the herds with a minimum number of seropositive animals (over a statistical threshold, see table 1) were considered as infected, taking into consideration the following parameters: • Sensitivity 45% • Specificity 99% • Lower prevalence in infected herds 5% • Confidence limits 95% • Minimum herd-level specificity 95% The threshold was related to the consistency of the herd and is reported in the following table.