Attack rates of norovirus and Campylobacter food poisonings that cannot be represented by numeric constant but by plotting attack rate against number of patients per incident in Cartesian coordinates

Abstract

　In norovirus and Campylobacter food poisonings, the frequencies of the number of patients per incident and that of the number of eaters per incident followed lognormal distribution, and their respective medians were 12~27 and 23~48 for norovirus food poisoning and 5~8 and 9~21 for Campylobacter food poisoning. The lognormal frequency distribution of eaters could be simulated by assuming that people are attracted more to a dish that has already attracted many.

　The number of patients per incident and that of eaters per incident were not necessarily inter-correlated; the frequencies of the attack rates (number of patients/number of eaters) were distributed evenly from ~0.01 to 1, that is, the attack rates of these food poisonings could not be represented by mean and standard deviation. The frequency distribution of the attack rate was nevertheless not entirely disordered: plotting the attack rate against the number of patients for individual incidents produced fingerprint-like pattern, which was repeatedly produced at prefectural and national levels.