The possibility of discriminating and classifying the weight of each factor involved in foodborne disease outbreaks through the use of the second-family quantification theory was investigated by applying this method to the data regarding foodborne disease collected during the year 1998. Two dependent variables were considered; the foodborne disease caused by
Vibrio parahaemolyticus and those caused by other bacteria. Independent variables considered were the date, the number of cases, the food involved, the establishment where the food was prepared, and the establishment or place where the food was consumed. The partial correlation coefficient among
V. parahaemolyticus and food, and
V. parahaemolyticus and season, were relatively high, at 0.501 and 0.303, respectively. Thus, those two variables have a strong influence in discriminating the outbreak caused by
V. parahaemolyticus from outbreaks caused by other bacteria. By restricting the type of food to seafoods, and the season to summer, the obtained category weights were 1.157 and 0.244, respectively. These results show that those two factors are strongly involved in outbreaks caused by
V. parahaemolyticus. As well, the correlation ratio was relatively high (0.440), and the hit rate obtained by analyzing the sample score data was 81.4%. The same analysis was carried out for the data concerning outbreaks caused by
Salmonella. In this case, however, a low correlation ratio such as 0.16 and a hit rate of only 72.3% were obtained. Thus, the majority of foodborne outbreaks due to
V. parahaemolyticus might be discriminated based on the analysis of the above five variables by using the second-family quantification theory, though it seems inapplicable for
Salmonella foodborne outbreaks.
View full abstract