Background: Norovirus (NoV) is commonly associated with gastrointestinal infection. It is normally transmitted person-to-person or from contaminated surfaces, although food-borne transmission is possible.
Methods: We conducted environmental, epidemiological, and microbiological investigations to ascertain the route of transmission of two linked outbreaks of NoV associated with events where food was consumed. Multivariate logistic regression was used to determine food items independently associated with infection.
Results: In outbreak A, 19 of the 26 people who completed the food questionnaire fulfilled the case definition. The highest relative risks (RR) were for chicken kebab (RR 3, 95% confidence interval (CI) 0.9-10.4), pork sausages (RR 2.1, 95% CI 0.5-9.1), pasta salad (RR 1.94, 95% CI 0.9-4.1), cheese (RR 1.6, 95% CI 0.9-2.8), and green leaf salad (RR 1.5, 95% CI 0.9-2.4). In outbreak B, 60 of the 106 people surveyed fulfilled the case definition. Green leaf salad (adjusted odds ratio (aOR) 3.2, 95% CI 1.4-9.9) and coleslaw (aOR 8.2, 95% CI 3-22.2) were independently associated with illness in the multivariate logistic regression model. NoV genogroup II genotype 6 (GII-6) was identified in cases of both outbreaks and a food handler who had prepared salads for both events.
Conclusion: Because outbreak investigations of small cohorts may not yield epidemiological association to food, most of these outbreaks may be attributed to the person-to-person transmission route. Therefore ascertainment of food-borne NoV infection may be low, underestimating the true prevalence of this route of transmission.