Abstract
Background
The Solus One Salmonella immunoassay utilizes Salmonella specific selective media and automated liquid handling, for the rapid and specific detection of Salmonella species in select food types.
Objective
The candidate method was evaluated using 375 g test portions in an unpaired study design for a single matrix, instant non-fat dry milk (NFDM) powder.
Method
The matrix was compared to the United States Food and Drug Administration/Bacteriological Analytical Manual (FDA/BAM) Chapter 5 Salmonella reference method. Eleven participants from 10 laboratories within academia and industry, located within the United States, Mexico, South Africa, Germany, and the United Kingdom, contributed data for the collaborative study. Three levels of contamination were evaluated for each matrix: an uninoculated control level [0 colony forming units (CFU)/test portion], a low inoculum level (0.2–2 CFU/test portion) and a high inoculum level (2–5 CFU/test portion). Statistical analysis was conducted according to the Probability of Detection (POD) statistical model.
Results
Results obtained for the low inoculum level test portions produced a dLPOD value with a 95% confidence interval between the candidate method confirmed (both alternative and conventional confirmation procedures) and the reference method of 0.07 (−0.02, 0.15).
Conclusions
The dLPOD results indicate equivalence between the candidate method and the reference method for the matrix evaluated and the method demonstrated acceptable inter-laboratory reproducibility as determined in the collaborative evaluation. False positive and false negative rates were determined for the matrix and produce values of <2%.
Highlights
Based on the data generated, the method demonstrated acceptable inter-laboratory reproducibility data and statistical analysis.