Observations on bacteriologic isolation from pairs of quarter-milk samples

Observations are reported on the agreement of bacteriologic isolation between 11,917 pairs of quarter-milk samples collected as either duplicate samples (both samples collected during 1 farm visit) or consecutive samples (2 samples collected at different milkings). Better agreement was found for duplicate samples (95.5%) than for consecutive samples (88.5% for interval of 2 to 7 days and 81.3% for interval of 8 to 16 days). From these observations, it was estimated that for diagnosing an intramammary infection, the examination of a single quarter-milk sample would be expected to have an error of approximately 3%.