STUDY OF MASTITIS IN DAIRY SHEEP USING SURVIVAL ANALYSIS

Mastitis is the most prevalent disease present in livestock species leading to economic loss. In dairy sheep, it caused mainly from bacterial infections. The aim of this work was to investigate the risk of having mastitis in Valle del Belice dairy ewes during the first lactation, due to environmental or contagious pathogens, using a survival analysis approach. All test‐day records from primiparous ewes were collected from five flocks. All test‐day were grouped in two data sets, one with mastitis due to environmental pathogens (ENV) and the other with mastitis due to contagious pathogens (CON). In this analysis the follow up period of a ewe was the lactation, consequently all the record began with lambing. Ewes without infection at the end of the first lactation were treated as right censored. The analysis was performed with the Survival kit v.6.1 (Ducrocq et al., 2010). Cox model was used: h(t; x_m,z_m )=h_0 (t)exp{x_m^' (t)┤β+z_m^i u where: h_0 (t) is the baseline hazard; β contains the time‐independent covariates: litter size (LS, 2 classes), age at first lambing (AFL, 4 classes); and timedependent covariates: milk production within flock (MK, 3 classes); somatic cell count within flock (SCC, 3 classes); EP: environmental pathogens, that was only considered when analyzing contagious pathogens (2 classes) and CP: contagious pathogens, that was considered only when analyzing environmental pathogens (2 classes); u contains the time‐independent random effect of FYS: flock‐year‐season (46 classes). For ENV and CON about 48% and 85% of the records were right censored, respectively. Chi‐square test approximations based on likelihood ratio test were calculated for all the effects in the model. For the ENV, the time‐dependent fixed effects: MK, SCC and CP were significant. A higher risk of being infected (hazard ratio, HR, 1.41) was observed for ewes with low MK compared to reference class (i.e. MK=2). Higher HRs were found for animals that had a medium and high level of SCC (1.80 and 2.23, respectively) compared to the reference class of low level of SCC. A lower HR was observed for animals that have been infected with contagious pathogens (0.08). For CON, SCC and EP were significant:. A higher HR was found for animals that had a medium (1.79) and high (3.96) level of SCC compared to the reference class of low level of SCC. A lower HR was observed for animals that have been infected with environmental pathogens (0.09). Results suggest that selection for decreased SCC may be effective to reduce mastitis incidence and the breeding goal should favor ewes with lowest observed SCC. Ducrocq, V., Solkner, J. and Meszaros G. (2010). Survival Kit v6 a Software Package for Survival Analysis. In: 9th World Cong. Genet. Appl. Livest. Prod., Leipzig, Germany.