Scientific Abstract
For surveillance, the inevitable question is, “How many samples must I collect for detection?”
The vast majority of research on this question has focused on oral fluid sampling in pens of ~25 pigs. This work has shown that one rope will provide an oral fluid sample representing ~80% of the animals in a pen holding 25 pigs in a 30-minute sampling period (Pol et al., 2017; White et al., 2014). However, there is no “industry standard” pen size and the industry is generally evolving toward larger pen sizes. To optimally sample swine populations for FAD detection, we need guidelines for oral fluid sampling across the range of pen sizes coming into modern production systems.
The objective of this study was to develop oral fluid sampling guidelines for contemporary pens sizes. Specifically, the goal was to determine how samples should be collected from pens of various sizes, e.g., 20, 40, 60, 100, etc., in order to obtain a specimen that represents the infection status of the pigs in the pen. It is important to note that this project did not involve test performance assessment. Test performance is pathogen- and assay-specific; in contrast, sampling guidelines should be universal.
Research Approach
Prior research on oral fluid sampling used infectious agents and/or modified live vaccines to model detection. This approach was highly effective, but the use of uncontrolled infectious agents in the field is impractical, hazardous, and entails legal risks. Therefore, we used two measures to evaluate sampling efficacy in this field study: (1) detection of a safe, non-infectious surrogate (flunixin meglumine, FM) in oral fluid from a subset of FM-inoculated pigs, and (2) pig behavior, i.e., counts of the interactions of FM-inoculated pigs with ropes hung in the pen. FM has been FDA-approved for use in market swine since 2006, with a withdrawal period of 12 days (Burr et al., 2006).
Key Findings
The study was conducted in 72 pens holding from 21 to 98 pigs by hanging 4 ropes in each pen for 20 minutes, observing pig behavior and then collecting oral fluids for FM analysis. Depending on pen size, each pen had 3 to 12 FM-inoculated pigs, i.e., “prevalence” was held relatively constant and pen size was varied.
Analysis of pig behavior data showed that between 33% and 100% of the FM-inoculated animals interacted with one or more ropes hung in the pens (mean 79%) in a 20-minute observation period. More specifically (Figure 2) in 70 of 72 pens (97.2%), ≥ 50% of injected pigs interacted with at least one rope. The 2 pens in which < 50% of the pigs interacted with a rope were smaller pens (≤ 25 head).
- Given a fix “prevalence” (FM-inoculated pigs), logistic regression analysis showed that pen size had no effect (p = 0.54) on the proportion of FM pigs that interacted with the rope. These results were surprising and counter intuitive. That is, most had assumed that detection would be more difficult in larger pen; such was not the case.
- This study was done in simple quadrangular pens holding <1 00 pigs. Whether pens with more complex designs and holding larger populations conform to these observations could be easily determined using the methodology described herein.
