EVALUATION OF 3 ASSAYS FOR FAILURE OF PASSIVE TRANSFER IN CALVES

This study examined the sensitivity. specificity, predictive values, a nd classification accuracy of 3 commonly used screening tests for fail ure of passive transfer: the sodium sulfite turbidity test, the zinc s ulfate turbidity test, and refractometry relative to serum immunoglobu lin G(1)(IgG(1)) concentrations determined by radial immunodiffusion. Serum samples were obtained from 242 calves ranging from 1 to 8 days o f age. Using a serum concentration of 1,000 mg/dL IgG(1) to define ade quate passive transfer, the zinc sulfate test had a sensitivity of 1.0 0 and a specificity of 0.52 in the detection of inadequate passive tra nsfer. The endpoint of the test appeared to be higher than desired; ca lves testing negative had mean serum IgG(1) concentration of 955 mg/dL and a large proportion of calves with adequate passive transfer were misclassified as positive for failure of passive transfer. Using the q ualitative zinc sulfate test, the percentage of calves correctly class ified with regard to passive transfer status was less than that observ ed with either the sodium sulfite test or refractometry. The sensitivi ty of the sodium sulfite assay was 0.85 at a 1+ endpoint and 1.00 at a 2 or 3+ endpoint. The specificity of the sodium sulfite assay varied from 0.87 at a 1+ endpoint and 0.56 at a 2+ endpoint. The sensitivity and specificity of refractometry varied from 0.01 to 1.00 depending on the choice of endpoint. Refractometry correctly classified the larges t proportion of calves with regard to their passive transfer status at test endpoints of 5.0 and 5.5 g/dL, 83% and 82% respectively. The hig hest percentages of carves correctly classified occurred with the sodi um sulfite test using a 1+ endpoint (86.30%) and refractometry using a 5.0 g/dL endpoint (83.00%). A regression equation was developed that permitted calculation of an optimal endpoint for refractometric determ inations of total serum protein concentration. A serum protein concent ration of 5.2 g/dL was equivalent to 1,000 mg/dL serum IgG(1). Optimal selection of tests for passive transfer status in calves will be gove rned by the prevalence of failure of passive transfer, test performanc e, and the anticipated costs of classification errors.