Predictors of bluetongue development in Sardinia (Italy) identification, using multilevel logistic mixed model

Abstract

Objectives: The Bluetongue Virus is one of the most studied ruminant diseases, affecting particularly sheep and goats. This study aims to identify, for the first time, the specific risk factors influencing the disease development in Sardinia, using multilevel logistic regression model, in order to give a contribution to the sanitary programs and favour the early detection. Methods: The data of the present retrospective study, collected from informatics systems of Istituto Zooprofilattico della Sardegna, are referred to all 15,780 Sardinian sheep farms observed for 3 years (2012-2014). The outcome of interest was dichotomous and defined the development of Bluetongue outbreak, after serological test or clinical signs. The effect of several region-specific prognostic factors on disease spread was investigated. Results: The final model indicated that Bluetongue development was significantly associated with an increase in number of animals (P < 0.0001), number of cattle around farm (P < 0.0001), water surface area (P =0.002), and amount of rainfall in the previous days (P < 0.0001). Furthermore, the altitude over 450 MASL (P < 0.0001), the vaccination prophylaxis (P < 0.0001) and the previous outbreak event (P < 0.0001) had a protective effect against the outcome. Conclusion: The results of this study indicated that number of animals and the amount of rainfall were the most important risk factors that affected the Bluetongue development, while the vaccination prophylaxis was found to be an effective measure in decelerating the disease spread.