Food microbiology: proficiency testing scheme on a naturally contaminated fresh refrigerated food.

The major objective of the study is to assess the performance of a proficiency testing scheme of a fresh refrigerated dairy product and to compare it with those applied currently in food microbiology. The accuracy is evaluated by using the repeatability and reproducibility standard deviations (respectively referred as s(r) and s(R)). For total mesophilic bacteria, with a mean Log count of around 8 bacteria per g, s(r) was equal to 0.15 and s(R) equal to 0.32. For total and fecal coliforms, these Values were respectively s(r) = 0.19, s(R) = 0.34, and s(r) = 0.20, s(R) = 0,50, at mean Log count of about 3.76 bacteria per g. For Staphylococcus aureus at a contamination level ranging between 1.5 and 4 Log count bacteria per g, these values were s(r) = 0.54 and s(R) = 1.33. Because our samples were originally Salmonella free, spiked samples were used. The sensitivity rate of the method is 95.8 % and the specificity rate 99.1 %. In the course of the present study, a significant progressive improvement of performance among laboratories was observed.