The objective of this study was to determine the sensitivity of culturing pooled samples containing varying numbers of individual droppings for detection of Salmonella in commercial egg-laying flocks relative to the within-flock prevalence. A laboratory experiment was performed to directly measure the effect of diluting positive with negative faeces on the sensitivity of detection, and thus provide priors for a Bayesian model of pooled sampling. Pooled samples made up of different numbers of individual faecal droppings were collected from 20 flocks, and in addition bulked faeces and dust were also sampled using an in-house method that involved collecting 10 dust and 10 faeces samples into jars with buffered peptone water. The results from these flocks were analysed using Bayesian methods for diagnostic test evaluation in the absence of a gold standard, and the sensitivity of each pooled sample type was estimated relative to the within-flock prevalence. The sensitivity of pooled samples depended on the within-flock prevalence, and increased as the prevalence increased. The sensitivity of pooled sampling tended to increase with the number of droppings in the pool, and overall there was a higher proportion of positive samples from the pools of 20, 60 and the in-house faeces pooling method compared to the pools of 10, 5 and the individual droppings. Dust samples were more sensitive than the faeces samples, and so the inclusion of dust in sampling schemes is recommended.
Crown Copyright © 2011. Published by Elsevier B.V. All rights reserved.