Somatic cell count, electrical conductivity, and serum albumin concentration for detecting bovine mastitis.

Cell concentration, electrical conductivity, and serum albumin concentration of milk were evaluated for predicting infection status of quarters in three herds. Probability of misclassifying quarters was lowest for cell concentration. For discriminating quarters infected with Staphylococcus aureus from quarters free from infection, probability of misclassification for cell concentration ranged from 8 to 20% among herds. For electrical conductivity probability of misclassification ranged between 22 and 32%, whereas for serum albumin the range among herds was from 15 to 48%. Among herds there was also considerable variation of the trait, which minimized the probability of misclassification. Cell concentrations for the three herds were 158, 200, and 251 x 10(3) cells/ml whereas for electrical conductivity counts were 4.6, 5.8, and 6.5 mS/cm. For electrical conductivity and serum albumin comparisons within cow by differences and ratios from different quarters yielded little advantage over absolute estimations. For a single sampling, cell concentration provided the most accurate prediction of quarter infection status. We suggest that 200 x 10(3) cells/ml should be the threshold above which quarters are considered likely to be infected with a major pathogen.