Repository logo

Estimating diagnostic test performance and examining effective testing strategies for the control of Johne's disease in western Canadian cow-calf herds



Journal Title

Journal ISSN

Volume Title






Degree Level



Johne’s disease is the clinical manifestation of persistent infection with the bacterium Mycobacterium avium subspecies paratuberculosis in cattle. Clinical signs include severe diarrhea and gradual weight loss with maintained appetite until the point of emaciation. Infected animals remain in a subclinical state either indefinitely or until clinical signs emerge after a period of 2-6 years. During subclinical infection, animals are capable of transmitting disease by shedding Mycobacterium avium subspecies paratuberculosis (MAP) in their feces on a continuous or intermittent basis. Currently, there is no effective treatment or vaccine for Johne’s disease, therefore control methods are limited to testing and removing infected animals from the herd. Presently, most of the research on the control of Johne’s disease has been focused on dairy cattle, due to the higher disease prevalence and therefore greater perceived threat of the disease within the dairy industry. However, Johne’s disease is of increasing importance to the beef cow-calf sector due to the ongoing consolidation of cow-calf herds in western Canada resulting in greater risk of infectious disease transmission. This thesis aims to address a gap in the current literature related to the performance of diagnostic tests for detecting MAP in Canadian beef herds and effective test and cull strategies for reducing disease prevalence in western Canadian cow-calf herds. In Chapter 2, the performance of three commonly used diagnostic testing methods was estimated along with the true prevalence of MAP infection in Canadian cow-calf herds. Blood and fecal samples were collected from 3171 cows on 159 cow-calf operations. All blood samples were analyzed with serum ELISA and all fecal samples were analyzed in pools of 5 samples with PCR. A subset of 913 samples from positive and negative herds were also tested individually with PCR. Bayesian latent class models were used to estimate the sensitivity and specificity of each testing method and the true prevalence of MAP. Multivariable regression analysis was used to determine associations between Johne’s disease risk factors and animal and herd positivity status. Results from this study found that the prevalence of MAP was higher in eastern Canadian herds compared to western Canadian herds at both the animal and herd level. Bayesian latent class model analysis yielded a sensitivity and specificity of 96% and 98% respectively for individual PCR at the animal level followed by pooled PCR (Se=54%, Sp=100%) and then ELISA (Se=36%, Sp=98%). Based on analysis of 20 samples per herd, ELISA had a higher sensitivity (Se=79%, Sp=90%) compared to pooled PCR (Se=43%, Sp=99%) for identifying herds with at least one positive sample. Cows from herds with dairy cattle on-farm and cows from herds that had animals show symptoms of Johne’s disease in the previous three years were more likely to test positive for MAP. Herds with animals demonstrating symptoms of Johne’s disease in the previous three years and with a higher number of breeding females were more likely to be MAP positive. In the third chapter, an agent-based simulation model was developed to study potential options for the control of Johne’s disease in cow-calf herds under typical management practices in western Canada. The diagnostic test sensitivity and specificity values generated in Chapter 1 were included in the model as well as the disease prevalence estimates. Two baseline, non-testing scenarios and 26 testing scenarios consisting of different testing methods, frequencies and combinations of animals targeted for testing were compared using the model. The testing scenarios were ranked based on the within-herd prevalence after a 10-year testing period. The cost of each testing scenario was also calculated to determine the most effective and least-cost testing scenarios. The results of this study found that after a 10-year period, 7 testing scenarios reduced the within herd prevalence below 2% and 4 testing scenarios reduced the within-herd disease prevalence to 1% or lower. Individual PCR testing scenarios were the most effective at reducing disease prevalence after 10-years followed by pooled PCR and then ELISA. Increasing the testing frequency and the number of animals tested had a positive impact on reducing the within-herd disease prevalence after 10 years. Scenarios that were both effective at reducing the prevalence from the initial starting point and were lower in cost included testing a random subset of 50% of the herd with individual PCR every 12 months, restricted testing of cows with less than 4 negative tests every 12 months with individual PCR and whole herd testing with individual PCR every 12 months. A more detailed investigation of the economics associated with testing versus herd infection is needed to determine if testing and culling is a financially attractive option for commercial cow-calf operations.



Johne's disease, beef cattle



Master of Science (M.Sc.)


Large Animal Clinical Sciences


Large Animal Clinical Sciences


Part Of