Abstract
INTRODUCTION: Salmonellosis is the second leading foodborne illness in the European Union. Eggs are still an important source of Salmonella despite an EU-harmonised control programme in laying hen flocks. The objective of our study was to identify the characteristics related to poultry house (location, type of housing system) and sampling (sampler, type and number of samples, date) associated with the detection of Salmonella target serovars (STS) in France (S. enteritidis, S. typhimurium, three monophasic variants of S. typhimurium and S. Kentucky). METHODS: For the first time since the implementation of the EU target prevalence in 2010, we compiled the results of bacteriological detection of Salmonella in French laying hen flocks (108,718 sampling events carried out in 4744 poultry houses). The risk of STS detection was modelled using a mixed logistic regression model taking into account repeated sampling at the poultry house level. RESULTS: An STS was isolated from 737 sampling events (0.68%). Caged flocks had a higher risk (odds ratio (OR) = 1.6, 95% confidence interval [1.2-2.0]) of testing positive compared with on-floor, organic or free-range flocks. The risk of detecting STS was higher when sampling was carried out by the competent authority (CA) (OR = 2.62, [2.2-3.1]) relative to food business operators (FBO), in relation to the risk-based sampling strategy used by the CA. A higher risk of STS detection was associated with taking six samples or more per sampling (OR = 2.8 [2.0-4.0]). A spatial gradient of risk was also described, running from the north-west to the south-east regions of France, in addition to seasonal (third quarter of the year: 2.8 [2.2-3.5], fourth quarter: 2.4 [1.9-3.0], relative to the first quarter) and annual effects (2016: 1.7 [1.2-2.5], 2020: 2.1 [1.5-2.9], 2021: 2.0 [1.4-2.8], relative to 2013). CONCLUSIONS: Our findings are of interest for improving sampling protocols for Salmonella detection in laying hen farms.