Estimating the prevalence of food risk increasing behaviours in UK kitchens.

Anna K Jones,Caroline Millman,Michael Burton,Paul Cross,Dan Rigby,Sarah J O'Brien,George-John Nychas

doi:10.1371/journal.pone.0175816

Abstract

Foodborne disease poses a serious threat to public health. In the UK, half a million cases are linked to known pathogens and more than half of all outbreaks are associated with catering establishments. The UK Food Standards Agency (FSA) has initiated the UK Food Hygiene Rating Scheme in which commercial food establishments are inspected and scored with the results made public. In this study we investigate the prevalence of food risk increasing behaviours among chefs, catering students and the public. Given the incentive for respondents to misreport when asked about illegal or illicit behaviours we employed a Randomised Response Technique designed to elicit more accurate prevalence rates of such behaviours. We found 14% of the public not always hand-washing immediately after handling raw meat, poultry or fish; 32% of chefs and catering students had worked within 48 hours of suffering from diarrhoea or vomiting. 22% of the public admitted having served meat “on the turn” and 33% of chefs and catering students admitted working in kitchens where such meat was served; 12% of the public and 16% of chefs and catering students admitted having served chicken at a barbeque when not totally sure it was fully cooked. Chefs in fine-dining establishment were less likely to wash their hands after handling meat and fish and those who worked in award winning restaurants were more likely to have returned to work within 48 hours of suffering from diarrhoea and vomiting. We found no correlation between the price of a meal in an establishment, nor its Food Hygiene Rating Score, and the likelihood of any of the food malpractices occurring.

Highlights

There are an estimated 500,000 cases of foodborne disease linked to known pathogens in the UK annually [1], and 9.4 million in the US [2]
Randomised Response Technique (RRT), designed to more accurately reveal rates of illicit behaviours, was implemented with the objectives of: 1. determining the prevalence of Food Risk Increasing Behaviours (FRIBs) amongst working chefs, catering students and the public; 2. investigating whether food malpractices are correlated with observable characteristics among the general public; 3. investigating whether food malpractices are more likely in certain types of commercial establishments (FHRS score, price band, awards won) and correlated with observable characteristics of chefs and catering students; 4. exploring the implications of the prevalence of poor practices for food hygiene and human health
The prevalence of illicit, food risk increasing behaviours are difficult to determine via direct questioning and observational studies, given their sensitive and often fleeting nature

Summary

Introduction

There are an estimated 500,000 cases of foodborne disease linked to known pathogens in the UK annually [1], and 9.4 million in the US [2]. The risk of systematic misreporting has led to the development (in other domains) of questioning techniques which induce greater truth telling and reveal more accurate estimates of the prevalence of sensitive behaviours. Methods such as the Randomised Response Technique (RRT) and the Item Count Technique [14, 17] introduce randomisation or uncertainty into the question-answer process, protecting respondents by obscuring their answer. This study focuses on poor food safety practices in kitchens It is concerned with rates of Food Risk Increasing Behaviours (FRIBs). Investigating whether food malpractices are more likely in certain types of commercial establishments (FHRS score, price band, awards won) and correlated with observable characteristics of chefs and catering students (gender, position etc); 4. RRT, designed to more accurately reveal rates of illicit behaviours, was implemented with the objectives of: 1. determining the prevalence of FRIBs amongst working chefs, catering students and the public; 2. investigating whether food malpractices are correlated with observable characteristics among the general public (gender, age, attitudes to risk, etc); 3. investigating whether food malpractices are more likely in certain types of commercial establishments (FHRS score, price band, awards won) and correlated with observable characteristics of chefs and catering students (gender, position etc); 4. exploring the implications of the prevalence of poor practices for food hygiene and human health

Materials and methods

I have served meat that is ‘on the turn’

À p1 À p2

Results

Discussion