Understanding bovine paratuberculosis in beef herds: Testing strategies and potential outcomes

摘要

The objective of this investigation was to obtain test performance values for a commercial ELISA test and fecal culture in beef cattle for Johne's Disease; as well as to identify the potential economic consequences of using those tests in serial and parallel testing schemes. To accomplish these objectives, we conducted a cohort study (n=693) in a beef cow/calf operation (n=780). All the animals included in the study were tested for three consecutive years using ELISA. In the third year, fecal culture was also conducted with the ELISA test in a parallel scheme. Test results were evaluated using statistical (Chi square), Bayesian distribution, and Receiver Operating Curve (ROC). Under the situations evaluated, the ELISA test showed a sensitivity of 0.083; specificity, 0.920; positive predictive value, 0.509; negative predictive value, 0.500; positive likelihood ratio, 1.033; negative likelihood ratio, 0.997; diagnostic odds ratio, 1.036; and kappa, 0.003. Bayesian estimations for the ELISA test indicated a low positive predictive value (0.035), with a high negative predictive value (0.999). The obtained ROC showed a r~2 of 0.866. Our results are indicative of less than ideal test performance from the ELISA test, when used in conjunction with fecal testing in a parallel testing scheme. Conversely, results of yearly testing using ELISA in a serial testing scheme showed better agreement of results. The economic implications of making management decisions based on the testing schemes were evaluated using pay-off tables.