Identification and Ranking of risk factors for somatic cell count economic penalty in 349 southern Belgium dairy farms

In Belgium, the main economic penalty accounted for bovine milk quality is the bulk milk somatic cell count geometric mean over 3 months reaching more than 400,000 cells/ml. Yet, it is still difficult to make progress on udder health and milk quality because regional risks related to endemic farming practices are not broadly known. Hence, a first step in understanding specific udder health risks associated with herd management has to be a broad ecopathological survey.
A random stratified sample of 349 dairy farms, representing 25% of producers registered for performance recording, was selected with a total of 16,000 cows. Thorough audits recording 400 farming practices were made in each farm by 2 different surveyors during milking. The practices were recorded across four categories: Herd structure, Housing, Milking practices and Herd Management (including Nutrition). Our chosen variable was the geometric mean of the herd composite somatic cell count from the last three months compared to the 400,000 cells/ml European threshold. The sample had a mean somatic cell count of 287,000 cells/ml following a normal distribution between 73,000 and 807,000 cells/ml.
From 19 risk indicators identified through univariate logistic analysis (p<0.15), half were related to milking practices and 5 were underlined by significant odds-ratios (OR) found through multivariate logistic analysis (p<0.05). Therefore, it was found that cubicle housing had the least risk (OR= 0.59 compared with tightened stalls, OR= 0.42 compared with straw stalls); Presence of a calving pen (OR= 0.40), use of post-dipping (OR= 0.50) had a positive impact; whereas pre-dip had a negative impact in our study (OR= 3) though it was not clear if this routine was performed correctly. Stripping also had a bad impact on milk quality whether it was systematic (OR = 1.90) or occasional (OR = 2.43). It was also found that farms with poor milking liner hygiene had more trouble (OR = 2.34).
The results were comparable to other ecopathological studies such as northern and southern American and European studies. This study is a prerequisite in operational veterinary advice in southern Belgium dairy farms, because it provides a cross-sectional study of dairy practices and states on major epidemiological risk factors in dairy management for this region.