Diagnostic test performance of somatic cell count, lactate dehydrogenase, and N-acetyl-β-d-glucosaminidase for detecting dairy cows with intramammary infection

Abstract

The main objective of this study was to investigate the diagnostic test performance of somatic cell count (SCC), lactate dehydrogenase (LDH), and N-acetyl-β-. d-glucosaminidase (NAGase), analyzed in composite test milking samples, for detecting dairy cows with or without intramammary infection (IMI). A second objective was to investigate whether an adjustment of these udder health indicators according to their associations with different influential factors (i.e., parity, days in milk, and season) improved their test performance. Moreover, we wanted to investigate whether test performance of SCC improved if SCC results from previous adjacent test milkings were included in the model. Such test milking data were not available for LDH or NAGase. In this cross-sectional study, quarter milk samples for bacteriological examination were taken from almost 1,000 cows from 25 dairy herds during 3 consecutive days: the day before test milking, the day of test milking, and the day after test milking. From each cow, a composite test milking sample was analyzed for milk composition, SCC, LDH, and NAGase. Among the cows sampled, 485 were IMI negative and 256 were IMI positive in one or more udder quarters according to the definitions used. The remaining cows had inconclusive IMI status. To assess the test performance of SCC, LDH, and NAGase to identify IMI-negative and IMI-positive cows, univariable generalized estimating equation models were used with the udder health indicator of interest as outcome and IMI status as explanatory variable. From these models, receiver-operator characteristic curves were created and the area under cure (AUC) was calculated. From each model, a cut-off was chosen for calculations of the sensitivity (Se), specificity (Sp), positive predictive value (PPV), negative predictive value (NPV), and accuracy (ACC) for each udder health indicator. The AUC was similar for the adjusted SCC (0.84), nonadjusted SCC (0.83) and geometric mean SCC (0.80-0.81), but much lower for LDH (0.66) and NAGase (0.62). The highest Se, Sp, PPV, NPV, and ACC were obtained using SCC. Adjustment of the udder health indicators for influential factors (e.g., parity) did not improve the test performance markedly, whereas adding information about SCC from previous adjacent test milkings improved the test performance of SCC slightly. In conclusion, of the udder health indicators investigated, SCC had the best overall ability to correctly identify IMI-negative and IMI-positive dairy cows.