The optimization of Salmonella surveillance programmes for pullet and layer farms using between-farm distance as a risk factor

Abstract

Human salmonellosis cases are often caused by Salmonella serovars Enteritidis and Typhimurium and a large percentage of Salmonella outbreaks is associated with the consumption of eggs and egg products. For this reason, many countries implemented general surveillance programmes for the detection and control of Salmonella on pullet and layer farms. The infection risk however varies between farms and the identification of risk factors for Salmonella infection may be used to improve the performance of these surveillance programmes. The aims of this study are therefore to determine 1) whether local farm density is a risk factor for the infection of pullet and layer farms by Salmonella Enteritidis and Typhimurium and 2) whether the sampling effort of surveillance programmes can be reduced by accounting for this risk factor, while still providing sufficient control of target serovars Enteritidis and Typhimurium. To assess the importance of local farm density as a risk factor, we fitted different transmission kernels to Israeli surveillance data during the period from June 2017 to April 2019. The analysis showed that the distance to infected farms significantly increases the infection risk by serovar Enteritidis within an approximately 4 km radius and by Typhimurium within an approximately 0.3 km radius. We subsequently used these kernels to derive a model for the between-farm R0 and used it to optimise a surveillance programme that subdivided layer farms into groups at low and at high risk of between-farm transmission based on the local farm density and allowed the sampling frequency to vary between these groups. In this design, the pullet farms were always sampled one week prior to pullet distribution. Our analysis showed that the risk-based surveillance programme was able to keep the between-farm R0 of serovars Enteritidis and Typhimurium below 1 for all pullet and layer farms, using a sampling effort that was reduced by 32% compared to the currently implemented surveillance programme in Israel. The results of our study therefore indicate that local farm density is an important risk factor for infection of pullet and layer farms by Salmonella Enteritridis and Typhimurium and can be used to improve the performance of surveillance programmes.