Assessment of the Risk of Foot and Mouth Disease among Beef Cattle at Slaughter from East African Production Systems

Figures

• 

  Figure A1. Sensitivity analysis: Reductions in P1 result in “paradoxical” increases in total risk. Cumulative distribution functions for R1 (cattle infected in the source herd), R2 (cattle infected en route to slaughter), and Ptot (overall probability of infection at slaughter) for two production systems in Uganda under the default and alternative approaches to estimating FMD incidence. Top left: agropastoral, default. Bottom left: agropastoral, animals seropositive with record of vaccination within six months of sampling are called FMD-negative. Top right: pastoral, default. Bottom right: pastoral, viral isolation from probang samples used rather than serology. Vertical gray lines indicate the median value of each curve. (a): Under the scenario where antibodies of recently vaccinated animals are assumed to indicate vaccination rather than infection (bottom panel), the agropastoral system has the largest decrease in prevalence of all systems. This resulted in a reduction in the median value of R1 to 0.14 (from 0.22 in the default scenario (top, black curve)). An increase in R2 (due to more animals eligible for infection) “compensated” for the lower R1, and the median Ptot was slightly…

• 

  Figure 1. Risk pathways for the probability of FMD infection at slaughter among cattle sold for meat in Kenya and Uganda. R1 represents cattle infected at the time of leaving the source herd. R2 represents cattle that acquire new infections between the herd and the time of slaughter. The total probability, Ptot, is the sum of R1 + R2. Each event is conditional on the preceding events.

• 

  Figure 2. Total risk, Kenya. The cumulative distribution functions (a) and probability density functions (b) for the probability of cattle sold for meat arriving at slaughter while infected with FMD from each of four production systems in Kenya. The vertical gray line represents the median value. Distributions are based on 30,000 iterations of the stochastic model. Ptot is the sum of R1 and R2 depicted in Figure 1.

• 

  Figure 3. Total risk, Uganda. The cumulative distribution functions (a) and probability density functions (b) for the probability of cattle sold for meat arriving at slaughter while infected with FMD from each of four production systems in Uganda. The vertical gray line represents the median value. Distributions are based on 30,000 iterations of the stochastic model. Ptot is the sum of R1 and R2 depicted in Figure 1.

• 

  Figure 4. Influential nodes (a) and input variables (b) for feedlot and ranching systems in Kenya. Each node (P1–P6) or input value was divided into percentiles (1, 5, 25, 50, 75, 95, 99) and the conditional mean value of Ptot was calculated when the node was held fixed within each percentile interval while all others varied randomly. Only the top five most influential nodes or inputs were included for each plot. Note that the axes vary for each.