Sampling Techniques for Foodborne Pathogens in Animals and Animal Products

" Simple Random Sampling (SRS). Simple random sampling (SRS) is the primary sampling technique, which also serves as the basis for other techniques. Simple random sampling occurs when a sample size of n (for example, selected subset of eggs) from a population of size N (for example, total number of eggs produced on a farm on a given day) is taken in such a way that every possible sample of size n has the same chance of being selected. Ordinarily, this is interpreted as the equal probability of selection of each element from the population, but this would omit simple random sampling with probabilities that are proportional to size (for example, more eggs selected from large houses on the farm). The use of SRS depends on having a list of all elements that are part of the population (called a list frame) from which to select the sample. The selection of the sample must be conducted using a random procedure. In a random procedure, each element of the population is assigned a unique identification. Random numbers are then chosen from a random number table or by a similar method. Then the sample is chosen by selecting the elements that are associated with the selected random numbers. For example, if the objective were to sample feces from ten cattle in a feedlot pen of one hundred cattle, we would start by creating a list frame of animals in the pen. Each animal could be assigned a unique number between one and one hundred. Ten unique numbers between one and one hundred would be selected from a random number table. The ten cattle represented by these random numbers would then comprise the sample. have serious human health implications, whereas Clostridium perfringens requires much higher spore numbers before human health is of concern. Four primary categories of conventional sampling techniques exist: (1) simple random sampling; (2) stratified random sampling; (3) systematic sampling; and (4) cluster sampling. The remainder of this chapter focuses on the presentation of these four basic sampling techniques and their effectiveness and practicality for sampling foodbome pathogens. The discussion ofeach technique includes a description of important assumptions, characteristics, and application to foodbome pathogens. The material presented here is intended to be an introduction. Many texts (Schaeffer et al. 1986; Kish 1995) are available for more detailed and theoretical presentations of sampling techniques. INTRODUCTION.. Monitorin.g for. foodborne athogens in U.S. ammal. p.opu~atlons pnor t? harvest ~as relied primarily o~ clmlcallsolates submI.tted durincr an outbreak investIgatIon or as part of a dIagnostrc w~rk-up. Some sampling for these pathogens also has occurred as part of the National Animal Health Monitorincr System (Dargatz et aJ. 2000). Post-harvest detection ~f foodborne pathogens has utilized clinical samples and samples from compliance monitoring undertaken by regulatory agencies such as the USDA's Food Safety Inspection Service (FSIS) and the Food and Drug Administration (FDA). Although there is value in linking pathogens with outbreak, diagnosis, and compliance events, there is also a need for increased efforts directed toward collection of reliable and representative samples for the precise and accurate assessment of pathogen levels in animals and in animal products. A well-designed monitoring system based on a sound sampling strategy makes it possible to make inferences concerning population prevalence or incidence levels and trends from a subset of the population. Associated risk factors and interventions can be assessed when the levels are reliably estimated.