Analysis of farm specific risk factors for Campylobacter colonization of broilers in six European countries

This study presents on-farm risk factors for the colonization of broiler flocks with Campylobacter based on comparable data from six European countries: Denmark, the Netherlands, Norway, Poland, Spain, and the UK. The study includes explanatory variables from a large questionnaire concerning production, farm management procedures and farm conditions, climate data on mean temperature, sunshine hours, and precipitation, as well as data on Campylobacter status of broiler flocks. All together the study comprises data from more than 6000 flocks. The data were analysed using a generalized linear model. Due to a large number of parameters, some collinearity and relatively many missing values, the model was analysed by a method using all available cases at each step in the modelling process. The modelling process includes backwards elimination and forward selection. Several approaches were furthermore explored by applying different strategies for categorizing explanatory variables and for selecting and eliminating variables in the model. Despite national differences in broiler production, common risk factors for Campylobacter colonization of broiler flocks were identified across all six countries. These were generally related to inadequate biosecurity. Identified risk factors were: broiler houses older than 15 years, absence of anterooms and barriers in each house, shared tools between houses, long downtime, and drinker systems with bells or cups. Also, the risk of broiler flocks becoming colonized with Campylobacter was clearly affected by country. In descending order, broiler flocks were more likely to be colonized in Poland, the UK, Spain, the Netherlands, Denmark and Norway due to country specific factors that could not be explained by the identified risk factors or any other variables from the questionnaire. The seasonality observed for prevalence values was described by the monthly mean temperature reported in the study, i.e. the higher the temperature, the higher the prevalence of positive flocks