Main content area

Predicting intramammary infection status at drying off using indirect testing of milk samples

Gohary, K, McDougall, S
New Zealand veterinary journal 2018 v.66 no.6 pp. 312-318
California mastitis test, bovine mastitis, cows, dry period (lactation), electrical conductivity, graphs, herds, lactation, milk, pathogens, prediction, somatic cell count, New Zealand
To evaluate the Rapid Mastitis test (RMT, or California Mastitis test) and electrical conductivity (EC) at drying off when used alone or in combination with herd test data (maximum or last herd test somatic cell counts (SCC) before drying off), to define cows or quarters with intramammary infection, using microbiological culture as the gold standard. Quarter-level milk samples (n=609) from clinically healthy cows (n=153), in three herds in the Waikato region of New Zealand, were tested at drying off using the RMT and EC, and were collected for microbiological culture. The maximum SCC and the SCC at the last herd test of the preceding lactation were determined for each cow. The sensitivity, specificity, and positive and negative predictive values for each test were calculated for different cut-points, using microbiological culture as the gold standard. Receiver operating characteristic curves were generated and the area under the curve (AUC) was calculated for each test. The same parameters were calculated for combinations of two tests in parallel or in series. Infection with any pathogen was detected in 62/153 (40.5%) cows and 99/609 (16.3%) quarters, and with major pathogens in 7/153 (4.6%) cows and 8/609 (1.3%) quarters. When predicting infection with any pathogen at the cow-level, the coefficient of agreement was highest for SCC (<0.32) and RMT (<0.28) and lowest for EC (<0.12); the AUC for RMT and EC when used singly ranged from 0.57–0.69, and in combination with SCC ranged from 0.68–0.75. AUC were similar for tests that used either the last or the maximum SCC. When evaluated singly, RMT and EC had only low to moderate diagnostic utility compared to bacteriological culture. When they were combined with SCC and interpreted in parallel, the results were improved, but only moderately. For herds that conduct herd testing, a single herd test late in lactation was as predictive of intramammary infection at drying off as multiple herd tests through the lactation. For herds that do not conduct herd testing, RMT has greater utility than EC.