Analysis of the resistance behaviour of indicator bacteria to several antibiotics in bulk milk using spatial point pattern methodology.

Between 2002 and 2004, indicator bacteria (Enterococcus, Escherichia coli) isolated from 799 bulk milk samples were tested against thirty different antibiotics for resistance. Systematic sampling over space (every eighth Styrian cattle breeding farm on an unaligned grid) was conducted within the scope of the bovine virus diarrhoea (BVD) control programme. Results of resistance testing were assigned to their respective farms by means of LFBIS (information system for agricultural and forestry enterprises: land- und forstwirtschaftlichen Betriebsinformationssystem) numbers. The distribution of resistance was evaluated using a geographic information system. To show the application of spatial point pattern analysis, results of resistance testing of Enterococcus to tetracycline and streptomycin is presented. Kernel density estimation and K functions were used to determine whether the distribution of the resistant samples was clustered, relative to isolates from samples that were sensitive to these antibiotics. Initially, spatial clustering of resistance (tetracycline and streptomycin separately) was investigated. The space-time scan statistic was subsequently used to search for space-time clusters of resistance for both antibiotics simultaneously.