Optimization of food microbiology proficiency testing

To ensure a high level of food safety, developed countries have opted for a harmonized controlling strategy, requiring billions of microbiological analyses on foodstuffs. Reliable analytical results are a cornerstone of this food safety strategy and have to be ensured by laboratory quality controls. The quality control investigated in this study is the Proficiency Testing (PT), also called inter-laboratory comparison. The general organization, statistical design and specific challenges related to the REQUASUD food microbiology PT were addressed and optimized to ensure the fitness-for-purpose of these controls. The first part of this study led to the development of tools (reference materials) allowing the organization of PT with a low number of participants. In a second time, this thesis focused on optimizing the PT samples. Technical challenges had to be dealt with, in terms of stability and homogeneity of naturally contaminated food items. Being close to routine food samples, the PT items used in this study allowed an unbiased assessment of the laboratory real-life analyses. The last part of this study was dedicated to the exploitation of the 2009-2015 PT results, a large database representing a unique opportunity to observe the laboratory analytical performances and to detect systematic errors leading to inaccurate results. Besides being a controlling tool for the assessment of the laboratory analytical capability, PT schemes turned out to be a privileged means to improve the quality of the analyses. PT providers play a key educational role in the detection, investigation and eradication of systematic errors inside the laboratories, to improve the reliability of the food microbiological analyses.(AGRO - Sciences agronomiques et ingénierie biologique) -- UCL, 2015Accès restreint02/12/2015(AGRO - Sciences agronomiques et ingénierie biologique) -- UCL, 2015Anglai