Abstract
BACKGROUND: Foodborne norovirus outbreak data in Japan from 2005-2006, involving virological surveillance of all symptomatic and asymptomatic individuals, were reanalyzed to estimate the asymptomatic ratio of norovirus infection along with the risk of infection and the probability of virus shedding. METHODS: Employing a statistical model that is considered to capture the data-generating process of the outbreak and virus surveillance, maximum likelihood estimation of the asymptomatic ratio was implemented. RESULTS: Assuming that all norovirus outbreaks (n = 55) were the result of random sampling from an identical distribution and ignoring genogroup and genotype specificities, the asymptomatic ratio was estimated at 32.1% (95% confidence interval [CI], 27.7-36.7). Although not significant, separate estimation of the asymptomatic ratio of the GII.4 genotype appeared to be greater than other genotypes and was estimated at 40.7% (95% CI, 32.8-49.0). CONCLUSION: The present study offered the first explicit empirical estimates of the asymptomatic ratio of norovirus infection in natural infection settings. The estimate of about 30% was consistent with those derived from volunteer challenge studies. Practical difficulty in controlling GII.4 outbreaks was supported by the data, considering that a large estimate of the asymptomatic ratio was obtained for the GII.4 genotype.