View Report Details

Management risk factors resulting in food poisoning outbreaks in the catering industry : a case-control study

Project Code: E03004

31/03/2005

Department of Epidemiology, Statistics and Public Health, Cardiff University
Palmer, S

Introduction

Most foodborne disease outbreaks in England and Wales occur in catering businesses, and the most commonly reported contributing factors are failures of food safety controls resulting in inadequate cooking or reheating, cross contamination from raw to cooked foods and inadequate refrigeration of cooked foods. In this study we have looked behind the food safety control failures to compare underlying management, operational and commercial factors in businesses where there have and have not been outbreaks. The purpose is to help explain why outbreaks occur
and to identify factors that may be targeted to reduce the risk of outbreaks in the future. We should emphasise, however, that in a case control study a risk factor indicates a statistical association between the factor and the occurrence of an outbreak of foodborne disease. It does not itself necessarily imply a causal role in the outbreak nor does its presence mean that an outbreak will occur. Several of the factors that were statistically associated with outbreak premises were included in the study on an exploratory basis, and further work is required to understand the explanation for the statistical associations.

Background

Several major reports on foodborne disease, such as the Richmond Report, the Stanley Royd Report and the Pennington Group Report have pointed to failures in management, inadequate procedures and staffing issues as contributory causes to outbreaks. However, to our knowledge few epidemiological studies have addressed these areas to try to quantify risk factors. In this study we have examined five
groups of factors related to:

Business characteristics
Staff employed and structure
Systems of management of staff
Operational practices
Unusual events and circumstances

Methods

Outbreaks of foodborne disease occurring in England and Wales in catering businesses between 1 December 2002 and 31 December 2003 were considered to be cases. These outbreaks were identified through the Health Protection Agency (HPA) surveillance system at the Communicable Disease Surveillance Centre, Colindale, and through an informal network of caterers and environmental health
officers. For each outbreak case, a control business, matched for local authority of occurrence and small medium sized enterprise (SME) status was selected from the local authority food registers. Case and control businesses were contacted by letter and telephone, and face to face interviews of the owner or manager or senior chef at the business were carried out by trained interviewers following a standard protocol.

Results

88 case businesses (90%) and 91 control businesses (93%) agreed to participate in the study.
Of the 88 outbreaks, 58 (66%) were bacterial and 44 (50%) were due to S. Enteritidis. In the 88 outbreaks there were 1,851 reported human cases.

Case businesses were more likely to be hotels, to serve Chinese cuisine and to be small or small/medium sized SMEs as opposed to micro SMEs. They were also more likely to serve dinner, snacks, offer room service and be open for 10 or more hours. They were statistically significantly less likely to offer cheaper meals. When the inter-relation between these factors was considered it appeared that serving
dinner and offering cheaper meals were independently significant.

Case businesses were more likely than controls to have multiple levels of management, employ full time kitchen staff, employ casual staff and pay head chefs and porters above the national average. They were less likely to have the owner or manager working in the kitchen. These factors tend to be closely related to one another and to the complexity of the menus offered. When this was taken into account none of these factors were independently significant.

Case businesses were more likely to provide staff accommodation and to report staff sicknesses before or at the time of the outbreak. These two factors could not be explained by association with other factors in this group.

We found associations between case businesses and several training variables. In multivariate analysis these training variables were not independently significant risk factors except for kitchen manager who received intermediate food hygiene training which was independent of the other training variables. This training variable was associated with larger SMEs and businesses employing more than 250 staff and
also businesses offering more expensive meals. Importantly, there was no indication of any protection from the training variables collected

Case businesses were more likely to use hot display cabinets for a variety of foods, and to prepare foods such as poultry from raw. They also were more likely to be supplied with fresh produce from regional level suppliers, rather than national or local sources. When these factors were examined together, two factors tended to stand out as factors of importance, the regional supply of eggs, and poultry on hot display buffets.
We considered whether the presence of food safety management systems based on the principles of Hazard Analysis Critical Control Points (HACCP) principles was a protective factor. However, it is important to acknowledge when considering the findings that the implementation of such systems was not assessed as part of thestudy and that at the time of the study the participating businesses were not legally required to have HACCP systems in place. There was no difference between case and control businesses in the maintenance of HACCP records or in verbal
approaches based on HACCP principles, nor in maintaining temperature records, cleaning schedules or staff training records.

Unusual events were much more likely to be reported by case businesses, as was the use of a relief manager at the time of the outbreak, or a change in menu preparation but the change of menu preparation was only of marginal significance when the association with the other two variables was taken into account.

Salmonella Enteritidis Outbreaks
Similar findings to the above were found for the subset of outbreaks caused by S. Enteritidis, although smaller numbers reduced the statistical power. In particular we found an association with Chinese cuisine and with the regional supply of eggs. We also found an association with higher wages. There was some evidence that eggs served in a hot display buffet was a risk factor, as was the use of cooling water in cooking ranges.

All bacterial outbreaks
Similar findings were seen in this dataset.

Discussion
The study, with a very high participation rate and high quality data, has produced unique insights into management risk factors.

The inter-connection between variables is highly complex. Furthermore, since multiple statistical tests were carried out in this study we would expect several false positive associations, the study findings should therefore be interpreted cautiously.

The association between outbreaks and hotels seems to be explained by the fact that hotels are more likely to serve dinner. Factors associated with more complex dinner menus such as offering more expensive meals, preparation of foods from raw and use of hot buffet displays were more common in hotels and Chinese restaurants. The presence of the owner or manager working in the kitchen did not itself appear to
be a protection once cheaper meals and business size were taken into account. We were unable to explain why full time kitchen staff were associated with case businesses. Consideration should be given to whether fulltime staff were more likely
to be expected to perform tasks for which they are not skilled.

The finding that staff sickness was a risk factor could be because staff sickness leads to foodborne outbreaks, or perhaps more probably, that staff who were cases of illness as part of the outbreak were misclassified by interviewees as ill before the outbreak. The link with providing staff accommodation cannot be explained and could be a chance finding.

An important finding was that neither formal training nor the presence of food safety management systems based on HACCP principles, although implementation of these systems was not assessed during the study, were linked to reduced risk. The lack of protection raises questions about both the appropriateness of training courses and the potential danger of businesses relying on paper proof of due diligence rather then ensuring application of sound catering principles.

The subset of S. Enteritidis outbreaks associated with Chinese restaurants can be explained in large measure by the use of regional egg supplies. Our study suggests that the quality of eggs supplied through this route should be examined.

Conclusion
Outbreaks of foodborne disease were more common in catering businesses that are larger SMEs and that cater for the higher price range, offering more services. Outbreak businesses tended to have more tiers of management, have fulltime kitchen staff, use casua l staff and offer staff accommodation. They were more likely to have supplies of eggs, poultry and meat from regional suppliers; regional supply of eggs was associated with S. Enteritidis outbreaks. They were less likely to collect food directly from the supplier. Outbreak businesses were more likely to use hot display cabinets and to reheat certain foods. They were also more likely to have a recent change of menu, or a new process or procedure or practice, or have a relief manager on at the time. Reported formal training and the presence of food safety management systems based on HACCP principles, were not related to reduced risk, although we cannot say how well these systems were followed since the study could
not assess this aspect of management.

Recommendations

Further research is needed to explore the possible relationship between occurrence of outbreaks and the employment of full-time kitchen staff and the provision of staff accommodation.
Further investigation is needed into the content and delivery of food hygiene training programmes to reinforce the ways in which training is linked to subsequent responsibilities in catering.
Further work is required to understand how approaches to HACCP control procedures are related to the actual implementation of HACCP principles in the catering business.
The catering industry should be alerted to the findings of this study. Regional workshops should be held for local authorities and other interested public health agencies in order to develop the application of the protocol for the identification of management risk factors during routine and outbreak investigations.
A working group should be established to consider the integration of key risk factors into the guidance on assessing the 'Confidence in Management' section of the inspection rating schemes, Food Safety Act 1990 Code of Practice. This group should include representatives from LACORS, FSA, environmental health practitioners and the project team.

Main File

Some of the files on this site may be in a format that your computer can't read. However, you can download Readers and Viewers for the following document types below: