Published online by Cambridge University Press: 01 June 2009
The following indirect tests for the detection of subclinical mastitis were applied, under bacteriological control, to a series of milk samples from the individual quarters of cows in full milk: presence of clots, reaction to bromcresol-purple, reaction to brom-thymol-blue, amount and character of centrifuge deposit, catalase content, chloride content, casein number and solids-not-fat content.
These tests detected 12, 61, 14, 55, 49, 77, 91, and 63% respectively of the infected quarters, and diagnosed 0, 16, 0, 3, 34, 0, 8, and 9% of the uninfected quarters as infected.
For each indirect test the mean square contingency was calculated as a measure of agreement between it and the direct bacteriological method. The tests were placed in the descending order of their value for diagnosis thus: casein number, chloride content, centrifuge deposit, solids-not-fat content, brom-cresol-purple reaction, brom-thymol-blue reaction, presence of clots, catalase content.