Risk factors for L. monocytogenes contamination of dairy products in Switzerland, 1990-1999.

Our purpose was to identify the main hazards associated with the spread of Listeria monocytogenes in dairy products in Switzerland and to determine the changes in predominant serotypes of the isolates, using databases on dairy-processing and environments from the Swiss Dairy Research Station during the years 1990-1999. Overall, of 76,271 samples collected, 3722 (4.9%) were positive for the presence of L. monocytogenes. Cheese-ripening facilities had the highest proportion of positive samples (7.6%), followed by small-scale local dairies (4.4%). By sample type, the highest proportion of positive samples (9.5%) was observed in water samples used for cheese-washing, followed by cheese-surface swabs (5.0%). During the 10-year period, no positive samples were obtained from cream, ice cream, milk powder, yogurt, or fresh cheese. Of 3722 L. monocytogenes isolates, 1328 (35.7%) were serologically typeable. Serotypes 1/2a, 1/2b, and 4b accounted for 92.7% of the 1328 isolates. Until 1995, the most-prevalent serotype was 1/2b (annual proportional prevalence 39.3-72.2%)--whereas since 1996, 1/2a was the most prevalent (34.7-54.7%). During 1996-1999, serotype 1/2a increased by 88%, compared to the average of 1990-1995. In the final random-effect multivariable logistic model, the strongest predictor of a positive culture was samples from cheese-ripening plant (OR=1.54; 95% CI: 1.14, 2.08) and the second-strongest predictor was samples collected by someone who was employed by the plant (OR=1.48; 1.29, 1.71). Hard and semi-hard cheeses were more likely to be associated with serotype 1/2b and soft cheeses with serotype 1/2a.