Taste and Safety: Is the Exceptional Cuisine Offered by High End Restaurants Paralleled by High Standards of Food Safety?

Sanch Kanagarajah; Piers Mook; Paul Crook; Adedoyin Awofisayo-Okuyelu; Noel McCarthy

doi:10.1371/currents.outbreaks.007219ac3b9a2117418df7ab629686b6

Introduction: Restaurant guides such as the Good Food Guide Top 50 create a hierarchy focussing on taste and sophistication. Safety is not explicitly included. We used restaurant associated outbreaks to assess evidence for safety.

Methods: All foodborne disease outbreaks in England reported to the national database from 2000 to 2014 were used to compare the Top 50 restaurants (2015) to other registered food businesses using the Public Health England (PHE) outbreak database. Health Protection Teams were also contacted to identify any outbreaks not reported to the national database. Among Good Food Guide Top 50 restaurants, regression analysis estimated the association between outbreak occurrence and position on the list.

Results: Four outbreaks were reported to the PHE national outbreak database among the Top 50 giving a rate 39 times higher (95% CI 14.5–103.2) than other registered food businesses. Eight outbreaks among the 44 English restaurants in the Top 50 were identified by direct contact with local Health Protection Teams. For every ten places higher ranked, Top 50 restaurants were 66% more likely to have an outbreak (Odds Ratio 1.66, 95% CI 0.89–3.13).

Discussion: Top 50 restaurants were substantially more likely to have had reported outbreaks from 2000-2014 than other food premises, and there was a trend for higher rating position to be associated with higher probability of reported outbreaks. Our findings, that eating at some of these restaurants may pose an increased risk to health compared to other dining out, raises the question of whether food guides should consider aspects of food safety alongside the clearly important complementary focus on taste and other aspects of the dining experience.

In England, several recent campylobacter outbreaks have been associated with poultry liver consumption. Following a lunch event in a hotel in Surrey in November 2013 where chicken liver parfait was served, guests reported having gastrointestinal symptoms. A retrospective cohort study showed 46 of 138 guests became unwell, with a median incubation period of two days and for 11 cases campylobacter infection was laboratory confirmed. Food item analysis identified an association between illness and consumption of roast turkey (aOR=3.02 p=0.041) or jus (aOR=3.55 p=0.045), but not with chicken liver parfait (OR=0.39 p=0.405). The environmental risk assessment did not identify non-compliance with standard food practice guidelines. This study presents a point-source outbreak of campylobacter with a high attack rate and epidemiological analysis results show that the jus or roast turkey was the likely source of infection although this could not be confirmed by the environmental assessment. Consuming the chicken liver dish was not a risk factor for developing symptoms as was initially hypothesised. Prior knowledge on the association between poultry liver food items and campylobacter outbreaks should not overly influence an outbreak investigation to ensure the true aetiology is identified and on-going public health risk is minimised.