Within-herd mathematical modeling of Mycobacterium avium subspecies paratuberculosis to assess the effectiveness of...

Johne's disease (JD) in cattle is caused by Mycobacterium avium subspecies paratuberculosis (MAP) and is characterized by chronic, progressive enteritis that can lead to substantial weight loss, severe diarrhea, and eventual death. Economic losses due to JD are primarily driven by reduced milk production in subclinical and clinically infected cows, but also include reduced value when sold to slaughter, and costs associated with premature culling. Controlling the transmission of JD within a dairy herd can be achieved through proactive calf management practices and reactive test-based culling. While effective, test-and-cull interventions have the potential to result in net economic losses, particularly when the intervention includes culling of low-shedding cattle. Proactive calf management practices have been observed to be effective at controlling within-herd JD prevalence. However, assessing the magnitude of effect of interventions in observational and experimental studies can be difficult due to the pathogenesis of MAP and may take many years of data to provide meaningful results. The limitations of studying JD in nature presents an opportunity to use mathematical modelling techniques to assess the effectiveness of various interventions on the simulated within-herd disease dynamics of JD. The objectives of this study were to build a within-herd compartmental disease model of JD and assess the effectiveness of three interventions: 1) strategic insemination of test-positive low-shedding adult cattle to preferentially breed beef calves, 2) using separate calving areas for low- and high-shedding dams, and 3) test-based culling of low- and high-shedding cows. Model outcomes were compared to a base case model (i.e., no interventions) under four endemic within-herd prevalences. In general, simulations of test-based culling performed best at reducing long-term within-herd prevalence of JD. Strategic insemination and separate calving area interventions were both effective and performed similarly to one another, but even when combined were not as effective as test-and-cull alone. Finally, the results from the separate calving area intervention model suggest that increased dam-calf contact time would not result in a substantial increased within-herd prevalence. Given that some of the modelled populations in this study are very small and prevalence is very low, further work is needed to assess these interventions using discrete, stochastic methods, which may result in different outcomes.