Neighbourhood risk factors for Campylobacter infection in Danish broilers were evaluated. Campylobacter infection status of a flock was identified by PCR analysis of cloacal swab samples collected as a part of national surveillance program. Included into the study were, in total, 10,876 broiler flocks sent by 226 farms to processing plants from 2007 to 2009. A multivariable logistic regression model with autocorrelation structure was used to model the effect of exposure variables on the probability of being tested positive to Campylobacter. Results showed a significant protective effect with the absence of infected neighbours within a distance of 2 km. The analysis was adjusted for potential confounding factors. Seasonal cyclic patterns of the Campylobacter infection was accounted for by using sine and cosine function. Predicted probability maps showed a heterogeneous spatial and temporal risk of Campylobacter infection in Danish broiler.
Spatial and Spatio-temporal Epidemiology, 2012, Vol 3, Issue 4, p. 311-322