Accuracy of a new milk strip cow-side test for diagnosis of hyperketonemia

Abstract

The objectives of this study were to determine the accuracy of the PortaBHB^TM milk strip for detection of hyperketonemia in early lactation cows, and to compare the agreement of results from quarter and composite samples. A total of 577 Holstein cows of all parities, from 88 commercial herds, were sampled once during this study. Cows were sampled simultaneously for blood and milk between one and 60 days-in-milk. Blood samples were collected from coccygeal vessels, and were analyzed on-farm using an electronic ?-hydroxybutyrate (BHBA) hand-held meter. Milk samples were collected from one quarter (n=577), as well as from four quarters (composite; n=299). All milk samples were tested using PortaBHB^TM milk strips (0, 50, 100, 200, and 500 æmol/L). Blood BHBA concentration was considered to be the gold standard, with hyperketonemia denned as a blood BHBA concentration ?1400 æmol/L. With this cut-point, the true prevalence of hyperketonemia in the study population was 24.6%. Using a threshold of 100 æmol/L, sensitivity and specificity of PortaBHB^TM milk strips was 89.2% and 79.6%, respectively. Using a threshold of 200 æmol/L, sensitivity and specificity of PortaBHB^TM milk strips was 40.3% and 99.5%, respectively. The kappa coefficient for agreement between milk results from quarter and composite samples, using a threshold of 100 æmol/L, was 0.95. Study results suggest that PortaBHB^TM has good accuracy, and that there is no benefit to collect milk samples from four quarters compared with sampling one quarter.