- Email: [email protected]
Food hygiene in hospitals
Journal of Hospital Infection (1988) 11 (Supplement
Food hy&ene
A), 77-81
in hospitals
P. J. Wilkinson
Public Health Laboratory,
Derriford
Hospital, Plyymouth, Devon PL6 8DH
Summary:
Food hygiene in British hospitals is reviewed in the context of national trends in food poisoning and changes in food legislation. New methods of large SC le catering such as the cook-chill system are considered, and the safe operati, 6 n of such a system in a typical health district is described. The application of turrent guidelines for the microbiological quality of cookchill food is evaluated. The need for careful observance of these principles, together with appropriate microbiological surveillance of the process and the product, is demonstrated.
Introduction The general public expects prepared foods supplied for consumption at home, in restaurants or in institutions to be palatable, nutritious and safe. Statutory microbiological testing of foodstuffs has been widely practised in a number of countries in an attempt to control the quality and safety of food. The UK has few such standards, preferring to place the emphasis on hygienic procedures rather than end-product specifications (Charles, 1979). A useful summary of the statutory regulations laying down general hygiene requirements both for premises and for food handlers is given by Hobbs & Roberts (1987); the general legislation is incorporated in the Food Hygiene (General) Regulations 1970 (SI 1970 No. 1172). High standards of hygiene are particularly important in hospital catering, where meals must be provided for large numbers of patients and staff day and night throughout the year. Such catering is still often carried out in unfavourable circumstances from old kitchens under conditions of financial constraint. It can be difficult in large hospitals to provide fresh, hot food of good quality on wards situated at considerable distances from the kitchens. Factors
contributing
to food poisoning
Poor food hygiene can lead to microbial or toxic food poisoning; failures of hygiene in large catering operations may give rise to outbreaks of bacterial food poisoning, many of which have been reported in the specialist and lay press in recent years. The factors contributing to outbreaks of food poisoning in England and Wales, based on data from 1044 general and family outbreaks reported between 1970 and 1979, have been analysed by 0195-6701/88/02A077+05
Q 1988 The Hospital
SO3.00/0
77
lnfectmn
Society
78
P. J. Wilkinson
Roberts (1982), who found that most investigated incidents of food poisoning were caused by several different factors. The commonest, found in more than 60% of outbreaks, was preparation of the food more than half a day in advance of needs. Other major contributory factors were storage at ambient temperature (40%), inadequate cooling (32%), inadequate reheating (29%) and the use of contaminated processed food (19%): Undercooking of meats and poultry, not thawing frozen meat and poultry for a sufficient time, cross-contamination from raw to cooked foods and storage of hot food below 63°C also played a part. Infected food handlers did not have a significant role except in instances of food poisoning with Staphylococcus aureus. Hygiene hazard check lists based on these factors are incorporated in current guidance for the National Health Service (Department of Health and Social Security, 1986a). Salmonellosis
in hospitals
Despite clear and consistent guidance of this sort for many years, salmonellosis still occurs in British hospitals and is the commonest reported cause of food poisoning outbreaks in England and Wales. Abbott, Hepner & Clifford (1980) reviewed 522 hospital outbreaks due to salmonella infection between 1968 and 1977, which comprised between 32 and 50% of all general outbreaks each year. A more detailed analysis of 197 outbreaks between 1974 and 1977 showed that there were more in maternity, children’s and geriatric wards than in medical or surgical wards, with some extensive outbreaks in hospitals for the mentally ill and mentally handicapped. Forty-two patients (2% of those affected) died, the majority of whom were very old or very young. The source of infection was recorded in only 76 (39%) of the outbreaks and was thought to be food in 24 of them. person-to-person spread from Transmission by other routes, including patients with diarrhoea or with asymptomatic infections, and from infected staff, was incriminated in 52 outbreaks. Palmer & Rowe (1983) surveyed 55 outbreaks of salmonella infection in hospitals in England and Wales prospectively between July 1980 and July 1982. They found that food-borne infection probably accounted for only six outbreaks, but that these made up 40% of the 15 outbreaks in which there were more than five patients and staff with symptoms. They concluded that person-to-person spread accounts for most hospital outbreaks of salmonellosis, although this has been disputed by others (MacGregor & Reinhart, 1973). The Stanley’ Royd outbreak Great public concern was aroused in 1984 by a single, large outbreak of food-borne salmonellosis at the Stanley Royd Hospital, a large psychogeriatric hospital in Wakefield, Yorkshire, which became the subject
Food
hygiene
in hospitals
79
of a Committee of Inquiry (Department of Health and Social Security, 19863). The causative organism was Salmonella typhimurium phage type 49 and the probable vehicle was co$ked beef, although extensive contamination was found in the hospital kitcvens and numerous inadequacies of hygiene were recorded. Of 788 patients resident at the start of the outbreak on 26 August 1984, 355 developed symptoms; salmonellas were isolated from the faeces of 218 of these. Of the remaining 433 asymptomatic patients, salmonellas were isolated from the faeces of 81. Of 980 staff, 106 had symptoms and 51 provided faeces containing salmonellas. Twenty-five of the 884 asymptomatic staff also had salmonellas in their faeces. Thus 379 individuals, 299 patients and 80 staff, showed bacteriological evidence of infection in this outb)eak, and 19 patients died. Largely as a result of this incident, the National Health Service (Amendment) Act 1986 became law; in it, crown immunity of staff working in National Health Service premises from prosecution for contravention of any provisions of statutory food legislation was removed. This loss of crown immunity came into effect in February 1987. New methods
of catering
Against this background of operational difficulties and demands, new methods of large-scale catering in the National Health Service were under consideration. In 1970, the report of a panel set up under the auspices of the Committee on Medical Aspects of Food Policy was published, giving guidance on the nutritional and hygienic implications of the preparation and use of pre-cooked frozen foods. This report also referred to the system of pre-cooked chilled food but advised that the latter method of catering should not routinely be used to feed a community because of the enhanced risk of bacterial growth. By 1980, technological advances were considered to have made the cook-chill system a viable alternative to other systems of food preparation, provided that necessary safeguards are observed. Guidelines on pre-cooked chilled foods (Department of Health and Social Security, 1980) give specific, detailed guidance on the setting up, operation and quality control of such systems. Advisory microbiological guidelines are included in this document, which are not intended to be used for the routine testing of all batches of food and are not standards for the acceptance or rejection of any batch. Rather, they should be used when setting up a cook-chill kitchen or when establishing a new cook-chill process, to assist the responsible person in determining that a satisfactory standard has been reached. These microbiological guidelines apply to chilled food and are therefore affected by the quality of the raw material and the standards of cooking, handling, chilling and chill storage. They are not intended to be applied to the food once it has been reheated. One item of food should be taken from each batch tested and should achieve the following microbiological criteria:
P. J. Wilkinson
80
Salmonella spp. Escherichia coli Staphylococcus aureus Clostridium perfringens Total aerobic colony count plates for 48 h at 37°C
Hospital
none in less than less than less than after
incubation
25 g 10 g-’ 100 g-’ 100 g-’
of agar less than lo5 g-’
cook-chill
catering
in Plymouth,
Devon
Between August 1985 and July 1986, a delivered meals service based on cook-chill was introduced in Plymouth Health District. Some 6000 meals for patients and staff in 23 hospitals are produced daily in the central production unit in Derriford hospital. After chill storage for up to 5 days, meals are either plated centrally and reheated on wards or distributed in refrigerated vans to the cold stores of finishing kitchens on other sites, from which they are portioned and distributed as at Derriford (large hospitals) or reheated centrally and distributed hot (small hospitals). The microbiological quality of these hospital meals has been studied in detail (Sandys & Wilkinson, 1988). Between May 1986 and August 1987, more than 2000 food items were examined at the end of chill storage. Salmonellae and Clostridium perfringens were not detected in any. On one occasion, small numbers of E. coli (less than 100 g-‘) were found in a single item of meat; this was subsequently attributed to human error in the kitchen. Staphylococcus aureus (subsequently shown not to have been a toxin producing strain) was found in two dishes, which were withdrawn; all staff were then screened and nasal carriage of S. aureus was treated. Total aerobic colony counts at 37°C for 48 h were found, with very few exceptions, to be satisfactory in all cooked foods produced strictly in accordance with the DHSS guidelines. High bacterial counts could all be attributed either to human error during food preparation or, in a few cases, to multiplication of organisms present on pre-cooked meats (chicken, ham) obtained frozen from a commercial supplier and incorporated without further cooking into meals which were then chilled. A specification of for commercial suppliers is now under microbiological quality consideration. in food Colony counts in excess of lo5 cfu g-’ were found more frequently items not produced in accordance with DHSS guidelines before chill storage. For some, such as sweet dishes (e.g., trifle, blancmange) which were heated during production but served cold, often garnished with fresh pasteurized cream or cake decorations, the strict application of DHSS guidelines for total aerobic counts was clearly not appropriate. For others such as bulk liquids (soups, gravies) which were not suitable for cryogenic chilling in shallow containers, the use of clean plastic liners eliminated contamination by surface organisms from their polythene storage vessels.
Food hygiene
in hospitals
81
The greatest incidence of high total colony counts was seen in cooked vegetables, of which 19.5% (30 out of’i54) had counts in excess of 10 cfu g-’ in the first 4 months of the study ,‘eriod. This was attributed to the process of water and ice refreshment of b! lk-steamed vegetables in large aluminium containers, prior to transfer to the chill store direct. Various attempts to improve this process met with only partial success and, since July 1987, vegetables have almost all been chilled cryogenically in the same way as other cooked dishes, with the same good microbiological results, Throughout this study period, all of the few cases of gastroenteritis in patients or staff were shown to have an explanation unrelated to the hospital food. Conclusions
Poor food hygiene in British hospitals has caused numerous outbreaks of food poisoning in the past, though this situation may improve now that NHS premises are fully subject to national food legislation. Modern catering systems such as cook-chill are efficient and have an excellent record of safety. Their introduction into the Health Service is likely to improve the quality of hospital catering, provided that the DHSS guidelines are strictly observed and that not only the setting up but also the continued operation of the process are adequately controlled microbiologically. References Abbott, J. D., Hepner, E. D. & Clifford, C. (1980). Salmonella infections in hospital. A report from the Public Health Laboratory Service Salmonella Subcommittee. Journal of Hospital Infection 1, 307-314. Charles, R. H. G. (1979). Microbiological standards for foodstuffs. Health Trends 11, 14. Department of Health and Social Securitv (1980). Guidelines on pre-cooked chilled foods. Her Majesty’s Stationery Office, Londbn. Denartment of Health and Social Securitv (1986aj. Health Service Catering &I~~Hvcliene. Her ,Majesty’s Stationery Office, London: ~ ’ Department of Health and Social Security (19866). The report of the Committee of Inquiry into an outbreak of food poisoning at Stanley Royd Hospital. Her Majesty’s Stationery Office, London. Hobbs, B. C. & Roberts, D. (1987). Legislation. In Food Poisoning and Food Hygiene, 5th edn, pp. 259-279. Edward Arnold, London. MacGregor, R. R. & Reinhart, J. (1973). Person-to-person spread of salmonella: a problem in hospitals? Lancet ii, 1001-1003. Palmer, S. R. & Rowe, B. (1983). I nvestigation of outbreaks of salmonella in hospitals. British Medical Journal 287, 891-893. Roberts, D. (1982). Factors contributing to outbreaks of food poisoning in England and Wales 1970-1979. Journal of Hygiene (Camb.) 89, 491-498. Sandys, G. H. & Wilkinson, P. J. (1988). M’ lcrobiological evaluation of a hospital delivered meals service using pre-cooked chilled food. Journal of Hospital Infection (in press).
Food hy&ene
A), 77-81
in hospitals
P. J. Wilkinson
Public Health Laboratory,
Derriford
Hospital, Plyymouth, Devon PL6 8DH
Summary:
Food hygiene in British hospitals is reviewed in the context of national trends in food poisoning and changes in food legislation. New methods of large SC le catering such as the cook-chill system are considered, and the safe operati, 6 n of such a system in a typical health district is described. The application of turrent guidelines for the microbiological quality of cookchill food is evaluated. The need for careful observance of these principles, together with appropriate microbiological surveillance of the process and the product, is demonstrated.
Introduction The general public expects prepared foods supplied for consumption at home, in restaurants or in institutions to be palatable, nutritious and safe. Statutory microbiological testing of foodstuffs has been widely practised in a number of countries in an attempt to control the quality and safety of food. The UK has few such standards, preferring to place the emphasis on hygienic procedures rather than end-product specifications (Charles, 1979). A useful summary of the statutory regulations laying down general hygiene requirements both for premises and for food handlers is given by Hobbs & Roberts (1987); the general legislation is incorporated in the Food Hygiene (General) Regulations 1970 (SI 1970 No. 1172). High standards of hygiene are particularly important in hospital catering, where meals must be provided for large numbers of patients and staff day and night throughout the year. Such catering is still often carried out in unfavourable circumstances from old kitchens under conditions of financial constraint. It can be difficult in large hospitals to provide fresh, hot food of good quality on wards situated at considerable distances from the kitchens. Factors
contributing
to food poisoning
Poor food hygiene can lead to microbial or toxic food poisoning; failures of hygiene in large catering operations may give rise to outbreaks of bacterial food poisoning, many of which have been reported in the specialist and lay press in recent years. The factors contributing to outbreaks of food poisoning in England and Wales, based on data from 1044 general and family outbreaks reported between 1970 and 1979, have been analysed by 0195-6701/88/02A077+05
Q 1988 The Hospital
SO3.00/0
77
lnfectmn
Society
78
P. J. Wilkinson
Roberts (1982), who found that most investigated incidents of food poisoning were caused by several different factors. The commonest, found in more than 60% of outbreaks, was preparation of the food more than half a day in advance of needs. Other major contributory factors were storage at ambient temperature (40%), inadequate cooling (32%), inadequate reheating (29%) and the use of contaminated processed food (19%): Undercooking of meats and poultry, not thawing frozen meat and poultry for a sufficient time, cross-contamination from raw to cooked foods and storage of hot food below 63°C also played a part. Infected food handlers did not have a significant role except in instances of food poisoning with Staphylococcus aureus. Hygiene hazard check lists based on these factors are incorporated in current guidance for the National Health Service (Department of Health and Social Security, 1986a). Salmonellosis
in hospitals
Despite clear and consistent guidance of this sort for many years, salmonellosis still occurs in British hospitals and is the commonest reported cause of food poisoning outbreaks in England and Wales. Abbott, Hepner & Clifford (1980) reviewed 522 hospital outbreaks due to salmonella infection between 1968 and 1977, which comprised between 32 and 50% of all general outbreaks each year. A more detailed analysis of 197 outbreaks between 1974 and 1977 showed that there were more in maternity, children’s and geriatric wards than in medical or surgical wards, with some extensive outbreaks in hospitals for the mentally ill and mentally handicapped. Forty-two patients (2% of those affected) died, the majority of whom were very old or very young. The source of infection was recorded in only 76 (39%) of the outbreaks and was thought to be food in 24 of them. person-to-person spread from Transmission by other routes, including patients with diarrhoea or with asymptomatic infections, and from infected staff, was incriminated in 52 outbreaks. Palmer & Rowe (1983) surveyed 55 outbreaks of salmonella infection in hospitals in England and Wales prospectively between July 1980 and July 1982. They found that food-borne infection probably accounted for only six outbreaks, but that these made up 40% of the 15 outbreaks in which there were more than five patients and staff with symptoms. They concluded that person-to-person spread accounts for most hospital outbreaks of salmonellosis, although this has been disputed by others (MacGregor & Reinhart, 1973). The Stanley’ Royd outbreak Great public concern was aroused in 1984 by a single, large outbreak of food-borne salmonellosis at the Stanley Royd Hospital, a large psychogeriatric hospital in Wakefield, Yorkshire, which became the subject
Food
hygiene
in hospitals
79
of a Committee of Inquiry (Department of Health and Social Security, 19863). The causative organism was Salmonella typhimurium phage type 49 and the probable vehicle was co$ked beef, although extensive contamination was found in the hospital kitcvens and numerous inadequacies of hygiene were recorded. Of 788 patients resident at the start of the outbreak on 26 August 1984, 355 developed symptoms; salmonellas were isolated from the faeces of 218 of these. Of the remaining 433 asymptomatic patients, salmonellas were isolated from the faeces of 81. Of 980 staff, 106 had symptoms and 51 provided faeces containing salmonellas. Twenty-five of the 884 asymptomatic staff also had salmonellas in their faeces. Thus 379 individuals, 299 patients and 80 staff, showed bacteriological evidence of infection in this outb)eak, and 19 patients died. Largely as a result of this incident, the National Health Service (Amendment) Act 1986 became law; in it, crown immunity of staff working in National Health Service premises from prosecution for contravention of any provisions of statutory food legislation was removed. This loss of crown immunity came into effect in February 1987. New methods
of catering
Against this background of operational difficulties and demands, new methods of large-scale catering in the National Health Service were under consideration. In 1970, the report of a panel set up under the auspices of the Committee on Medical Aspects of Food Policy was published, giving guidance on the nutritional and hygienic implications of the preparation and use of pre-cooked frozen foods. This report also referred to the system of pre-cooked chilled food but advised that the latter method of catering should not routinely be used to feed a community because of the enhanced risk of bacterial growth. By 1980, technological advances were considered to have made the cook-chill system a viable alternative to other systems of food preparation, provided that necessary safeguards are observed. Guidelines on pre-cooked chilled foods (Department of Health and Social Security, 1980) give specific, detailed guidance on the setting up, operation and quality control of such systems. Advisory microbiological guidelines are included in this document, which are not intended to be used for the routine testing of all batches of food and are not standards for the acceptance or rejection of any batch. Rather, they should be used when setting up a cook-chill kitchen or when establishing a new cook-chill process, to assist the responsible person in determining that a satisfactory standard has been reached. These microbiological guidelines apply to chilled food and are therefore affected by the quality of the raw material and the standards of cooking, handling, chilling and chill storage. They are not intended to be applied to the food once it has been reheated. One item of food should be taken from each batch tested and should achieve the following microbiological criteria:
P. J. Wilkinson
80
Salmonella spp. Escherichia coli Staphylococcus aureus Clostridium perfringens Total aerobic colony count plates for 48 h at 37°C
Hospital
none in less than less than less than after
incubation
25 g 10 g-’ 100 g-’ 100 g-’
of agar less than lo5 g-’
cook-chill
catering
in Plymouth,
Devon
Between August 1985 and July 1986, a delivered meals service based on cook-chill was introduced in Plymouth Health District. Some 6000 meals for patients and staff in 23 hospitals are produced daily in the central production unit in Derriford hospital. After chill storage for up to 5 days, meals are either plated centrally and reheated on wards or distributed in refrigerated vans to the cold stores of finishing kitchens on other sites, from which they are portioned and distributed as at Derriford (large hospitals) or reheated centrally and distributed hot (small hospitals). The microbiological quality of these hospital meals has been studied in detail (Sandys & Wilkinson, 1988). Between May 1986 and August 1987, more than 2000 food items were examined at the end of chill storage. Salmonellae and Clostridium perfringens were not detected in any. On one occasion, small numbers of E. coli (less than 100 g-‘) were found in a single item of meat; this was subsequently attributed to human error in the kitchen. Staphylococcus aureus (subsequently shown not to have been a toxin producing strain) was found in two dishes, which were withdrawn; all staff were then screened and nasal carriage of S. aureus was treated. Total aerobic colony counts at 37°C for 48 h were found, with very few exceptions, to be satisfactory in all cooked foods produced strictly in accordance with the DHSS guidelines. High bacterial counts could all be attributed either to human error during food preparation or, in a few cases, to multiplication of organisms present on pre-cooked meats (chicken, ham) obtained frozen from a commercial supplier and incorporated without further cooking into meals which were then chilled. A specification of for commercial suppliers is now under microbiological quality consideration. in food Colony counts in excess of lo5 cfu g-’ were found more frequently items not produced in accordance with DHSS guidelines before chill storage. For some, such as sweet dishes (e.g., trifle, blancmange) which were heated during production but served cold, often garnished with fresh pasteurized cream or cake decorations, the strict application of DHSS guidelines for total aerobic counts was clearly not appropriate. For others such as bulk liquids (soups, gravies) which were not suitable for cryogenic chilling in shallow containers, the use of clean plastic liners eliminated contamination by surface organisms from their polythene storage vessels.
Food hygiene
in hospitals
81
The greatest incidence of high total colony counts was seen in cooked vegetables, of which 19.5% (30 out of’i54) had counts in excess of 10 cfu g-’ in the first 4 months of the study ,‘eriod. This was attributed to the process of water and ice refreshment of b! lk-steamed vegetables in large aluminium containers, prior to transfer to the chill store direct. Various attempts to improve this process met with only partial success and, since July 1987, vegetables have almost all been chilled cryogenically in the same way as other cooked dishes, with the same good microbiological results, Throughout this study period, all of the few cases of gastroenteritis in patients or staff were shown to have an explanation unrelated to the hospital food. Conclusions
Poor food hygiene in British hospitals has caused numerous outbreaks of food poisoning in the past, though this situation may improve now that NHS premises are fully subject to national food legislation. Modern catering systems such as cook-chill are efficient and have an excellent record of safety. Their introduction into the Health Service is likely to improve the quality of hospital catering, provided that the DHSS guidelines are strictly observed and that not only the setting up but also the continued operation of the process are adequately controlled microbiologically. References Abbott, J. D., Hepner, E. D. & Clifford, C. (1980). Salmonella infections in hospital. A report from the Public Health Laboratory Service Salmonella Subcommittee. Journal of Hospital Infection 1, 307-314. Charles, R. H. G. (1979). Microbiological standards for foodstuffs. Health Trends 11, 14. Department of Health and Social Securitv (1980). Guidelines on pre-cooked chilled foods. Her Majesty’s Stationery Office, Londbn. Denartment of Health and Social Securitv (1986aj. Health Service Catering &I~~Hvcliene. Her ,Majesty’s Stationery Office, London: ~ ’ Department of Health and Social Security (19866). The report of the Committee of Inquiry into an outbreak of food poisoning at Stanley Royd Hospital. Her Majesty’s Stationery Office, London. Hobbs, B. C. & Roberts, D. (1987). Legislation. In Food Poisoning and Food Hygiene, 5th edn, pp. 259-279. Edward Arnold, London. MacGregor, R. R. & Reinhart, J. (1973). Person-to-person spread of salmonella: a problem in hospitals? Lancet ii, 1001-1003. Palmer, S. R. & Rowe, B. (1983). I nvestigation of outbreaks of salmonella in hospitals. British Medical Journal 287, 891-893. Roberts, D. (1982). Factors contributing to outbreaks of food poisoning in England and Wales 1970-1979. Journal of Hygiene (Camb.) 89, 491-498. Sandys, G. H. & Wilkinson, P. J. (1988). M’ lcrobiological evaluation of a hospital delivered meals service using pre-cooked chilled food. Journal of Hospital Infection (in press).