Risk assessment to support guidance for norovirus outbreaks in oysters

Summary

Risk Assessment Summary

This report has been commissioned to assess the public health risk associated with consumption of raw oysters with a range of norovirus RNA levels. Norovirus infection is typically a mild, self-limiting illness, and foodborne transmission is an important route, responsible for an estimated 16% of cases of norovirus. Between 2017 and 2023 in Scotland, and between 2013 – 2022 in England, there were, in total, 1,566 reported cases of norovirus infection linked to oysters, of which 4 were hospitalised. Data is not routinely collected from Wales and was not available from Northern Ireland. Cases are more frequent during winter months. Detection of viral particles uses a polymerase chain reaction (PCR) method but this cannot distinguish between infectious and non-infectious norovirus. Cefas in England and the Marine Institute in the Republic of Ireland are the only laboratories accredited for PCR quantification of norovirus in oysters in the British Isles. Other laboratories may provide norovirus quantification, using methods comparable to the International Standard method ISO 15216. There can be variability due to differences between laboratories and the starting material, with the latter especially affecting the extraction efficiency. Estimates of the norovirus dose required to make 50% of susceptible individuals ill vary between studies from a thousand to hundreds of thousands of particles, with modelling studies suggesting it may be even lower. However, there is a broad consensus that even low doses of infectious particles can be sufficient to make susceptible individuals ill. Norovirus contaminates oysters due to human sewage discharges around oyster beds. Oysters bio-accumulate the virus in their digestive glands, with levels of norovirus significantly higher in winter months than in summer months. As part of this assessment, we compared norovirus levels in oysters sampled at retail to levels in oyster batches linked to outbreaks. Outbreak batches had significantly higher levels, with geometric mean of 874 norovirus genome copies/g compared to 24 genome copies/g in retail batches, although results varied over a broad range for both groups. There is uncertainty regarding the frequency with which oysters are consumed in the UK, but we estimate that around 4 million oyster meals are consumed each year, with an average portion size of six oysters, which weigh 30 g per oyster on average.

Conculsion

We conclude that the risk of illness is unknown in the absence of risk factors such as sewerage spills, epidemiological linkage to outbreaks, high E. coli levels or high rainfall. Should oysters be consumed without further treatment and in conjunction with scenarios such as potential human wastewater contamination from sewage spills or if the batches are linked to outbreaks, then at 1-200 norovirus genome copies/g the risk is low - medium* (rare, but does occur - occurs regularly) (*the likelihood of illness when consuming other oysters from that bed is low-medium in instances of test results below 200 norovirus genome copies per gram. This is to recognise that the likelihood may be assessed low in outbreaks with a small number of reported illnesses (or on the basis of other risk factors), or could be assessed medium in outbreaks with a large number of reported illnesses (or on the basis of other risk factors)). At 201-500 copies/g the risk is medium (occurs regularly), at 501-1,000 copies/g the risk is high (occurs very often) and >1,000 copies/g the risk is very high (events occur almost certainly) with all scenarios having high uncertainty.