Estimation of the sensitivity of environmental sampling for detection ofSalmonellain commercial layer flocks post-introduction of national control programmes

Abstract

SUMMARYA key element of national control programmes (NCPs) forSalmonellain commercial laying flocks, introduced across the European Union, is the identification of infected flocks and holdings through statutory sampling. It is therefore important to know the sensitivity of the sampling methods, in order to design effective and efficient surveillance forSalmonella. However, improvedSalmonellacontrol in response to the NCP may have influenced key factors that determine the sensitivity of the sampling methods used to detectSalmonellain NCPs. Therefore the aim of this study was to compare estimates of the sensitivity of the sampling methods using data collected before and after the introduction of the NCP, using Bayesian methods. There was a large reduction in the sensitivity of dust in non-cage flocks between the pre-NCP studies (81% of samples positive in positive flocks) and post-NCP studies (10% of samples positive in positive flocks), leading to the conclusion that sampling dust is not recommended for detection ofSalmonellain non-cage flocks. However, cage dust (43% of samples positive in positive flocks) was found to be more sensitive than cage faeces (29% of samples positive in positive flocks). To have a high probability of detection, several NCP-style samples need to be used. For confirmation ofSalmonella, five NCP faecal samples for cage flocks, and three NCP faecal boot swab samples for non-cage flocks would be required to have the equivalent sensitivity of the EU baseline survey method, which was estimated to have an 87% and 75% sensitivity to detectSalmonellaat a 5% within-flock prevalence in cage and non-cage flocks, respectively.