Relationship between somatic cell count status and subsequent clinical mastitis in Dutch dairy cows

High composite somatic cell counts (CSCC) in dairy cows may develop into clinical mastitis (CM), suggesting that prevention or intervention of high CSCC may prevent CM later in lactation. The objective of this study was to quantify the relationship between high CSCC in dairy cows and the first subsequent case of CM in the same lactation. Farmer-diagnosed cases of CM and test day CSCC measurements during 1 year of 13,917 cows in 196 randomly selected Dutch dairy herds were available for analysis. Cows were followed in 1 lactation from the first test day postpartum until CM, drying off, culling or end of study. Cox proportional hazards models with time-varying CSCC levels were used to estimate the effect of high CSCC (>= 200,000 cells/ml) on the time until the first case of CM. A shared frailty effect was included to adjust for clustering of cows within herds. The proportion of cows developing CM after a CSCC measurement was 11%. Primiparae with a high CSCC had a 4-fold higher hazard for subsequent CM than primiparae with a low CSCC; multiparae with a high CSCC had a 2-fold higher hazard than multiparae with a low CSCC. Additionally, multiparae with a low CSCC had a 2-fold higher hazard for CM occurrence than primiparae with a low CSCC. Increasing the threshold for high CSCC showed that the risk for CM increased. If the last CSCC before CM was low, CSCC information of 2 preceding test days was more predictive than CSCC information from only the last test day. When the last CSCC was high, CSCC information of 2 preceding test days did not have added predictive value. This study identified that approximately 25% of first subsequent CM cases after a CSCC measurement can potentially be prevented when cows are prevented to get high CSCC or when high CSCC cows are removed from the population. This corresponded with a decrease in the proportion of lactating cows with CM after a CSCC measurement from 11% to 7%. (C) 2011 Elsevier By. All rights reserved.