bacteriological prevalence of hogs delivered by all producers, BPREV*, is allowed to vary in this
analysis from a high of 3.5% to a low of 0.5%.
The three Salmonella control packages available to producers are those defined by King,
Backus, and van der Gaag (2007). They include only reversible control measures. All three packages
contain basic control measures like control of rodents, hygiene protocols, disinfecting and cleaning.
Package 2 adds strict all-in/all-out procedures and separate routes for different suppliers. Package 3
adds acidification of feed and/or water, a highly effective but expensive measure. Costs for farm
control packages 1, 2 and 3 are estimated to be €0.72, €1.14, and €2.92 per hog (King Backus, and van der
Gaag, 2007). Farm-level serological prevalence probability distributions for the three control packages
- i.e., h(prevt∣xt) - are shown in table 1. The prevalence distribution for Package 1 is quite dispersed
and is centered around 50%. The distribution for Package 2 is more concentrated and has most of its
probability mass at levels below 30%. The distribution for Package 3 is still more concentrated at the
lower end of the prevalence range, with the probability of a test result of zero exceeding 70%. These
discrete prevalence distributions are used to determine elements of the state transition matrices required
for solution of the producer’s dynamic programming problem.
Three plant level Salmonella control packages are considered in this study. Package 1 contains
strict hygienic practices at slaughter & processing (cleaning and disinfection) induced by the investor
owner or cooperative through monitoring the plant manager’s effort. Package 2 adds acidification of
slaughter equipment in cleaning and disinfecting procedures. Package 3 adds logistic slaughter of sero-
negative pig herds. Logistic slaughter includes testing all pig herds and slaughtering positively tested
herds either at the end of the day or at another location to reduce cross-contamination in the
slaughterhouse. Logistic slaughter only reduces the prevalence of Salmonella effectively when the
delay between sampling pig herds at the farm and delivery to the slaughterhouse is at a minimum.
14