ABSTRACT
Selling fish products as fresh when they have actually been frozen and thawed is a common fraudulent practice in seafood retailing. Unlike fish products frozen to protect them against degenerative changes during transportation and to extend the product's storage life, fish intended for raw consumption in European countries must be previously frozen at -20° C for at least 24 h to kill parasites. The aim of this study was to use histological analysis to distinguish between fresh and frozen-thawed fish and to evaluate this method for use as a routine screening technique in compliance with the requirements of European Commission Regulation No. 882/2004 on official food and feed controls. Method performance (i.e., accuracy and precision) was evaluated on tissue samples from three common Mediterranean fish species; the evaluation was subsequently extended to include samples from 35 fish species in a second experiment to test for method robustness. Method accuracy was tested by comparing histological results against a "gold standard" obtained from the analysis of frozen and unfrozen fish samples prepared for the study. Method precision was evaluated according to interrater agreement (i.e., three laboratories with expertise in histopathology in the first experiment and three expert analysts in the second experiment) by estimating Cohen's kappa (and corresponding 95 % confidence intervals) for each pair of laboratories and experts and the combined Cohen's kappa for all three experts and laboratories. The observed interrater agreement among the three laboratories and the three experts indicated high levels of method accuracy and precision (high sensitivity and specificity) and method reproducibility. Our results suggest that histology is a rapid, simple, and highly accurate method for distinguishing between fresh and frozen-thawed fish, regardless of the fish species analyzed.
