Flight kitchen production is a typical form of mass catering, but has some unique features distinct from food preparation in restaurants and hotels. The time difference between food production in the flight kitchen and finally serving it on board an aircraft with limited kitchen facilities makes flight catering a high-risk food preparation operation. The complexity of the production procedures in the flight kitchen also increases the microbiological hazards associated with this type of food preparation. Major factors affecting the hygienic quality of the food are the size of the operation, the complexity of the in-flight service, the number of airlines catered for, the number of flights serviced during the day and the duration of the flights to be serviced.
Since each airline has its own specification, the management of multiple contracts increases the complexity of the planning and control. Production planning for flight caterers equates to just in time production techniques (JIT), meaning producing the necessary units, in the necessary quantities, at the necessary time (Briggs and Nevett 1995, Foskett 1995). An airline company has to decide to what extent return catering will be carried out; whether to utilise the flight kitchens of foreign airports and whether to use local suppliers. Frozen meals may be carried if an aircraft is using food from its homeland during the return leg. In general, there is a growing trend in preparing frozen meals for aircraft (Asplund 2000). Economical and production considerations as well as hygienic reasons favour frozen meals. Microbiological examination of a batch can be carried out before it is used, thus ensuring the safety of the food. Using frozen meals reduces the likelihood of the temperature reaching the critical limits within which the bacterial growth may occur.
A typical flow chart for flight kitchens is shown in Fig. 1. Flight kitchens normally use a cook-chill system for the preparation of cooked items (Kirk 1995). Cooked items are then rapidly chilled in blast chillers, according to the Association of European Airlines (AEA 1996) within 4 hours from 65°C to 10°C, and according to LSG-Hygiene Institute (1997) from 60°C to 5°C. A cold kitchen is used for the preparation of snacks, appetisers, salads and desserts. Until portioning and packing, all prepared items are kept chilled. After making up the meal trays, the trays are loaded into a trolley for the flight. If necessary, trolleys are loaded with dry ice in order to minimise the temperature rise in the aircraft galley before the food is served.
Food storage and preparation for serving takes place in aircraft galleys, which mostly have very limited space and equipment for this purpose. In common with any kitchen, a galley has to provide the following: cold storage areas, regeneration ovens, water boilers and beverage machines and the stowage of waste products. On narrow-bodied aircraft, the meals are kept chilled by using dry ice located within the trolley. Wide-body aircraft used for long-haul flights are today usually equipped with refrigerators or chiller units for trolleys (Goodwin 1995).
Chilled and frozen meals served hot must be re-heated, so that a core temperature at least 72°C is reached to destroy surviving pathogenic micro-organisms (LSG-Hygiene Institute 1997). In the 1970s, hot meal trays were transported to aircraft in hot ovens for short-haul flights and kept there until serving, the temperature of food being over 63°C (Bailey 1977). Today, a cook-chill system is mostly used, although hot served foods can still be transported hot to small aircraft if they are not equipped with ovens (Asplund 2000).
The great need for food hygiene guidelines in flight kitchens was noticed as early as the 1960s (WHO 1960, Bailey 1977). A global survey of 25 flight kitchens showed that 30% had inadequate refrigeration facilities (Mossel and Hoogendoorn 1971). Time-temperature studies of flight kitchens in the United States in 1977 revealed that the equipment used did not always keep food appropriately hot or cold in the flight kitchen or while it was transported to the aircraft (Bryan et al. 1978). In 1984, 20% of American flights were holding food at improper temperatures (Tauxe et al. 1987). A WHO working group issued recommendations for flight catering in consequence of several outbreaks associated with meals served on board (WHO, Regional Office for Europe 1977). Documentation of hygiene training and instructions dealing with good hygiene practice have since been an important part of the quality system of flight kitchens.
With regard to food hygiene risks in airline catering operations microbiological hazards are the most important. Microbiological hazards are associated with the raw ingredients, staff and processes as well as serving on aircraft. Many flight kitchens now use the hazard analysis critical control point (HACCP) system (Gork 1993, Kirk 1995, LSG-Hygiene Institute 1997). In Europe, the European Commission (Council Directive 1993) has set the legal requirements for the food business to adopt a hazard analysis-based approach in food hygiene management. Many flight kitchens use the global quality policy described by LSG-Hygiene Institute (1997). LSG Lufthansa Service Holding AG is the biggest airline catering alliance and provides 390 million meals yearly. Their quality system consists of HACCP combined with quality requirements including standards, good manufacturing practice (GMP) and good hygiene practice (GHP).
While choosing menus for airlines, certain foods that can constitute a health hazard should be avoided as an important preventive measure. Components of aircraft meals can be placed into four risk categories: dangerous, high-risk, medium- and low-risk items (AEA 1996). Products that by nature can constitute a risk as a ready meal, either as such or due to improper heat treatment on board, are classified as dangerous items (Bailey 1977, AEA 1996). These items include dairy products containing raw milk, undercooked poultry and raw or undercooked eggs, raw meat, raw shellfish and raw fish. Neither should raw sprouts be used as components of cold meals due to known Salmonella outbreaks (Mahon et al. 1997, O’Mahony et al. 1990, Pönkä et al. 1995, Inami and Moler 1999, van Beneden et al. 1999).
Products which are intensively handled after heat treatment are classified as high-risk items. Such products include poultry and meat de-boned after cooking, stuffed eggs, cold cuts, glazing, cooked shellfish peeled after heat treatment. Medium-risk items have undergone a minimum of handling after heat treatment and include fermented and air-dried meats and sausages, stews, rice and pastas. Acidified foods (pH values below 4.6), fresh fruits that can be peeled prior to eating, canned fruits, bread and dry bakery items are considered to be low-risk items.
Food handlers are a potential source of pathogenic micro-organisms, and therefore training and practice for good personal hygiene is needed. Food handlers should have a medical examination prior to employment, and should be kept under regular medical surveillance (Bailey 1977, LSG-Hygiene Institute 1997). A person known or suspected to be suffering from a disease likely to be transmitted through food or any person afflicted with infected wounds, skin infections or sores should not be allowed to work in contact with any unpacked foods.
In order to ensure that food suppliers have implemented and maintain a sufficient control level in their production plant, flight caterers should audit their suppliers (Foskett 1995, LSG- Hygiene Institute 1997).
The official control of flight kitchens depends on the national legislation of the country where the premises are located. Flight kitchens are subject to different requirements depending on the legislation of the country concerned. The authorities responsible for controlling flight kitchen operations must have good knowledge of the special features of this type of mass catering. The need for closer co-operation between airlines, local airport health authorities and national health administrations became apparent in the 1970s, when large outbreaks were reported in connection with growing mass tourism (WHO, Regional Office for Europe 1977).
The last few decades have seen an emphasis on the global feature of flight kitchens serving international airlines. Many airline companies use standardised audit forms to perform regular hygiene audits of their suppliers (AEA 1996). The controlling authority and airline companies alike demand HACCP-based quality assurance. Non-compliance with even a single CCP means a failure to reach the AEA standard. Bacteriological results of food, drinking water and ice cubes are inspected to ensure that the buyer’s specifications are being adhered to.
A comparative study of visual inspections and microbiological sampling in high-risk premises showed that neither sampling nor visual assessment monitored the performance of the premises reliably (Tebbut 1989). A combined approach, using selective microbiological examination to support standardised inspections, was suggested for monitoring hygiene in premises preparing high-risk foods. Microbiological testing is needed within a HACCP programme for hazard identification, monitoring CCPs and verification of the HACCP programme. The microbiological control includes testing of the whole production chain. Samples are taken from food at reception, prepared food items, process lines and environment, water, ice cubes, food handlers and, finally, ready meals.
Great emphasis must be placed on purchasing food items for the flight kitchen. Microbiological hazards are linked especially with raw materials. E.g. the prevalence of Salmonella and Campylobacter may be high in meat and poultry (Uyttendale et al. 1999, Boonmar et al. 1998). Faecal contamination of vegetables may be rather common especially in non-industrialised countries (Monge and Chinchilla 1996). Contaminated raw materials increase the risk of contamination of ready meals. Listeria monocytogenes must be taken into consideration particularly when purchasing vacuum-packed ready-to-eat fish products, where the prevalence of L. monocytogenes in up to 33% and 50% of cases were found (Lyhs et al. 1998, Johansson et al.1999). Microbiological testing is needed to prove that the legal requirements as well as the customers specifications are met.
Some flight catering companies take daily traceable counter samples from final meals representing each production batch. These are kept for up to three weeks in a freezer (LSG- Hygiene Institute 1997, Asplund 2000). In case of complaint, the respective frozen samples are tested. In order to assure the safety of meals purchased, airline companies use random sampling from final meals according to a test schedule. These samples are mostly collected on board.
Microbial biofilms that remain on surfaces after cleaning are of great concern in the food processing industry (Zottola and Sasahara 1994). Methods for measuring the efficiency of cleaning of the production environment are necessary in food premises manufacturing high-risk foods. Agar contact plates and the swabbing method can be used for hygiene control (NCFA 1987, Tebbut 1991). Commercial agar contact plates are also useful for the hygiene control of food premises (Rahkio and Korkeala 1997). They are mainly used for monitoring indicator bacteria. For specific micro-organisms, such as Listeria and Salmonella , selective enrichment and media must be chosen. The method of measuring adenosine-5’-triphosphate (ATP) bioluminescence gives results in a few minutes, thus making this system very suitable for on-line monitoring in HACCP programmes (Poulis et al. 1993, Vanne et al. 1996, De Boer and Beumer 1999). However, the ATP measured does not originate from bacteria only but the total ATP from all organic material on the surface. A significant opportunity for the future may be the provision of pathogen specificity to the ATP assays (Stewart 1997).
The legal requirement in Finland demands that a food handler having travelled outside the Nordic countries must be tested for Salmonella (Anonymous 1994). In Finland, flight kitchen food handlers are additionally screened for Salmonella once a year (Asplund 2000). Many airline companies have imposed stricter rules than the legal requirements. A Salmonella test from flight kitchen employees after travelling abroad, although not legally required, is demanded by many airlines.
Frequent microbiological tests on hands are useful to control hand hygiene. According to De Witt and Kampelmacher (1988), 8% of food handlers showed high numbers (>10 5 /hand) of Enterobacteriaceae and S. aureus on their hands. Normal hand washing resulted in a lower number of transient micro-organisms, but however, it did not lower the number of S. aureus .
Food handlers harbouring enterotoxigenic strains of S. aureus constitute a potential source of contamination of food via the hands. The primary reservoir in people are the anterior nostrils, and nares are the most consistent area from which this organism can be isolated (Williams 1963). The nasal carriage of S. aureus results easily in transfer of the bacteria to the hands.
The AEA has issued recommendations for microbiological analyses and limits for aircraft food (1996) (Table 1). Bulk items, such as hot meats, which have been portioned after heat treatment should not exceed the value of 5.0 x 10 5 cfu/g for total count and 1.0 x 10 3 cfu/g for coliforms. For items that have been handled (e.g. slicing, cutting) after heat treatment, higher values of total count and coliforms are permitted. Although the results of the total count and coliforms can be higher than the limit values, the food is not considered to be unsafe, but according to the AEA (1996) an investigation of food production practice is advised. Enumeration of total bacteria and coliforms is not considered necessary for cold meals containing raw vegetables, fruits and garnishes as well as for undercooked items, because they naturally contain high counts of these bacteria. If the AEA limits for Escherichia coli, S. aureus, Bacillus cereus, Clostridium perfringens and Salmonella spp. (Table 1) are exceeded, the food must be considered to be unsafe. Monitoring meals for indicators may reveal food processing or food handling errors but it is not advisable or valid to predict the safety of food based on these indicators alone (Tompkin 1983, Sofos et al. 1999).
Many airline companies demand stricter microbiological limits than those set by the AEA (1996). The microbiological analyses and limits used by an official food control laboratory in Finland to testing of meals served on aircraft are presented in Table 2.
Although many flight kitchens and airline companies record the microbiological quality of meals served on aircraft, only a few studies have been published. In a survey made in Bangkok in 1976, a high contamination rate of Salmonella (9%), Vibrio parahaemolyticus (3%) and S. aureus (2%) was found (Steffen et al. 1985). Monitoring Salmonella from approximately 6000 samples from flight kitchens in 40 worldwide locations showed a prevalence of 1%, and meals prepared in India and Indonesia were most frequently Salmonella positive (Munce 1986).
Between 1984 and 1986 a study was conducted on meals (567) produced by ten flight catering units at Heathrow airport, London (Roberts et al. 1989). Colony counts higher than 10 6 cfu/g were found in 24% of hot dishes. The proportion of samples exceeding 10 cfu/g for E. coli was 21%. Samples of meat products showed an incidence of C. perfringens and B. cereus of 0.2% and 3.0% respectively. Salmonella was isolated from 0.4% of samples.
The microbiological quality of food items was monitored in a Greek flight kitchen in 1992 (Lambiri et al. 1995). Salmonella was found in 1% of hot food items. Of the cold food items and desserts, 24% contained S. aureus >1.0 x 10 2 cfu/g. E. coli higher than 10 cfu/g was found in 12% and 7% of hot and cold food items, respectively. The number of Salmonella positive raw poultry samples was high (24%). Implementation of the HACCP system in 1993 followed by a new monitoring showed considerable improvement in the microbiological quality of food items (Lambiri et al. 1995).
Enterotoxin formation was tested in 47 S. aureus strains, which were isolated from aircraft-ready meals in a four-year survey in Denmark. Immunological testing showed that 51% of the isolates were enterotoxigenic (Ewald and Christensen 1987).
Since tracing of the first foodborne outbreak associated with a meal served on aircraft in 1947, 41 outbreaks have been reported altogether (Table 3). Salmonella spp., S. aureus and Vibrio spp. have been the most commonly reported agents. Thousands of flights have been involved. Approximately 9000 air passengers and crew members have been reported to have suffered from food poisoning. The number of reported deaths was 11. In consequence of a Salmonella enterica serovar Typhimurium (hereafter S . Typhimurium) outbreak via infected cold salads from Las Palmas served on charter flights, six deaths occurred in 1976 (Table 3, outbreak 21). Salmonella enterica serovar Enteritidis (hereafter S . Enteritidis) was the reason for two deaths in a major outbreak, where passengers and crew at risk on 3103 flights were reported in 1984 (Table 3, outbreak 30). Vibrio cholerae caused two deaths, in 1972 and 1992 (Table 3, outbreaks 15, 39). The causative agent remained unknown in one foodborne outbreak, which was followed by one death in 1971 (Table 3, outbreak 11).
Foodborne outbreaks traced to meals served on aircraft are most probably underreported for several reasons. The incubation period is often longer than the flight time, and passengers may be unaware of each other’s illness. Therefore recognising a cluster of foodborne illness may be difficult. When an outbreak is identified, it always gives rise to a bad reputation and great financial losses (Pakkala 1989). Therefore airline companies, just as any companies providing a food service, do not like publishing any data on foodborne outbreaks. The authorities should recognise outbreaks associated with aircraft meals. In order to prevent dissemination or recurrence of outbreaks and the incidence of health hazards, a rapid international exchange of information is also needed.
Salmonella has been the most common pathogen associated with outbreaks traced to aircraft food (Table 3). It has been reported to cause 15 outbreaks and to infect approximately 4000 people. Eight different serotypes have been identified, with S . Enteritidis being the most common, causing 6 outbreaks. Salmonella enterica serovar Typhi (hereafter S . Typhi) was the cause of two outbreaks. Typical for Salmonella outbreaks in most cases was that the dissemination of contaminated food continued for several days and many flights were involved.
The first widespread Salmonella outbreak connected with airline meals occurred in the early years of mass tourism on intercontinental flights from Sydney to London via Vienna in 1967 (Table 3, outbreak 7). It affected almost 400 people. Contaminated mayonnaise prepared in a Vienna flight kitchen was found to be the source of infection. A great number of shortcomings in hygiene found during the investigation of the kitchen led to it being closed for 6.5 weeks. The largest Salmonella outbreak occurred in 1976. Approximately 1800 people from several European countries fell ill as a result of eating airline food served on charter flights (Table 3, outbreak 21). Findings from ill passengers and epidemiological evidence revealed cold salads with mayonnaise prepared in Las Palmas, Spain to be the source of infection. In 1984 a widely spread Salmonella outbreak occurred, where the suspected food, appetisers, was served on 3103 flights, and 631 first class and club class passengers and 135 air crew members were affected (Table 3, outbreak 30, Burslem et al. 1990). S. Enteritidis was isolated from a great number of cold food items with aspic glaze. Two large outbreaks involving over 400 people in each were reported in the 1990s (Table 3, outbreak 36, 41). Both outbreaks involved several charter flights, the first one catered for by a flight kitchen in the Greek islands and the second one in the Canary Islands.
Eight outbreaks caused by S. aureus have been reported (Table 3). Compared to Salmonella outbreaks, only a few flights were involved. In five outbreaks cold desserts were the vehicles of infection, and in three cases hot dishes.
In the 1970s, two major outbreaks occurred. The first one broke out on three flights from Rome to the USA via Lisbon in 1973 (Table 3, outbreak 17). The flights were catered for in Lisbon. Custard dessert, bavarois, was traced to be the source of infection. High counts (10 5 to10 8 cfu/g) of S. aureus were detected in the dessert. The evidence was conclusive, because S. aureus with the same antibiogram was isolated in patients as in the dessert. The second major outbreak took place on a long-haul flight from Tokyo to Paris via Anchorage and Copenhagen in 1975 (Table 3, outbreak 19). Snacks and breakfasts were loaded onto the plane in Anchorage. Ham included in the breakfast was shown to be contaminated with the same phage type and enterotoxin-producing strain as was isolated in the patients and in inflamed finger lesion of one cook. A great number of passengers (142) and one crew member required hospitalisation during intermediate landing in Copenhagen. The onset of symptoms began very soon, 0.5-2.5 h after eating the meal in both outbreaks. The symptoms were severe, mostly with nausea, vomiting and severe abdominal cramps. A high attack rate, 56% and 57% respectively, was found in connection with both outbreaks (Table 3, outbreaks 17, 19).
Investigations of two smaller outbreaks indicated high levels of S. aureus , 10 9 and 10 6 cfu/g in eclairs and in chocolate cake, respectively (Table 3, outbreaks 23, 37). The same types were found in patients, but the possible role of food handlers being the source of infection was not investigated. Fast exchange of information between the public health agencies in the United Kingdom and the United States facilitated the rapid identification of an international outbreak, its aetiology and the food vehicle responsible for the outbreak in 1991 (Table 3, outbreak 37). It led to the withdrawal of the chocolate cake, and so the prevention of further illnesses.
Vibrio spp., V. cholerae O1, V. cholerae non O1 and V. parahaemolyticus were reported as causing six outbreaks via aircraft food (Table 3). The incubation period varied from 24 to 48 h for V. cholerae , but for V. parahaemolyticus as short as four hours incubation period was noticed. An outbreak caused by V. parahaemolyticus was reported to have broken out already during the flight, and 28 passengers were admitted to hospital (Table 3, outbreak 25).
The endemic occurrence of cholera in some Asian countries since 1961 caused the seventh cholera pandemic. It was apparently linked to V. cholerae outbreaks registered during long- haul flights from Europe to Australia in the 1970s. The gastrointestinal illness of passengers was traced to cold food loaded in Bahrain (Table 3, outbreaks 15, 16). Bahrain was experiencing an outbreak of cholera at the time. The outbreaks were caused by V. cholerae O1 in 1972 and V. cholerae non O1 in 1973 and in 1978 (Table 3, outbreaks 15, 16, 26). Contaminated cold plates were suspected to be the source of infection. Ice might also be a vehicle, because vibrios may survive for long periods in ice water.
Food prepared in a Hong Kong flight kitchen was suspected to be the reason for two outbreaks among two American tour groups to the Orient in summer 1969 (Table 3, outbreaks 8, 9). Multiple pathogenic bacteria were isolated in patients, but the gastrointestinal illness was best correlated with the isolation of non-cholera vibrios.
The largest airline-associated-outbreak of cholera occurred in 1992 (Table 3, outbreak 39). Seventy-five of the 336 passengers who had flown from Lima, Peru, to Los Angeles became infected and one died. Epidemiological study indicated a strong association between eating a cold seafood salad and illness (Eberhart-Phillips et al. 1996). This outbreak demonstrated the potential of airline-food-associated spread of cholera from endemic areas, such as South America. The outbreak highlighted the risk associated with eating cold foods prepared in cholera-infected area. Epidemic cholera appeared in South America for the first time in the 20th century during January 1991. In 1992, the epidemic had spread to 20 countries in Latin America, and more than 600000 cases and 5000 deaths were reported (CDC 1991, CDC 1992). The year 1998 was marked by increase of nearly 100% in cholera cases on all continents (WHO 1999 b).
Vibrio parahaemolyticus caused two aircraft-meal-associated outbreaks in the 1970s (Table 3, outbreaks 14, 25). Seafood appetiser from Bangkok and a seafood cocktail from Bombay were connected with the outbreaks. V. parahaemolyticus is widely distributed in inshore marine waters throughout the world and is a well-documented and among the most common food poisoning bacteria in Japan, India and South-East Asia as well as in the United States, and is responsible for the summer peak of gastroenteritis (Zen-Yoji et al. 1965, WHO 1999 a).
Four outbreaks caused by Shigella via aircraft meals have been reported (Table 3). The first one, in 1971, was traced back to food served to charter passengers on several flights from the Canary Islands to Sweden (Table 3, outbreak 12). The food was prepared in Las Palmas, and was reported as having infected 219 passengers. The seafood cocktail served on flights was epidemiologically connected with the illness of 19 persons (Table 3, outbreak 13). A wide Shigella outbreak associated with aircraft meals on 219 flights to 24 states in the United States and to England, Germany, Japan, and Mexico in 1988 (Table 3, outbreak 33). Illness was due to the consumption of cold food items prepared in the kitchen in Twin Cities, Minnesota. Relatively low attack rates (4%) on scheduled flights, a long incubation period (1-4 days), and the dispersion of ill individuals demonstrated the difficulties in detecting a foodborne outbreak among airline passengers who live in widely scattered geographic areas. The outbreak was identified because it also involved a professional football team travelling together (Hedberg et al. 1992). Shigella has also caused several outbreaks linked to food served on other traffic vehicles, such as cruise ships (Gikas et al. 1996, Koo et al. 1996).
Clostridium perfringens has been reported to cause one outbreak, the vehicle being a hot meal (Table 3, outbreak 10). A hot dish containing turkey was epidemiologically shown to be the source of this outbreak. A total of 394 persons on eight flights had been at risk. A great number of crew members (22/62) and a few passengers suffered from gastrointestinal illness with diarrhoea as the main symptom, with the mean incubation period of 11 h, characteristic to C. perfringens .
Oysters contaminated by E. coli incapacitated 22 crew members over a period of four days in 1967 (Table 3 and 4, outbreak 6). The oysters were served on six international flights from London. E. coli showed faecal contamination, but viruses might have been involved, too. The incubation period and symptoms were similar to Norwalk-like viruses, but there was no method for virus detection at that time. The development of methods for the recovery of viruses from bivalve molluscs has proved that raw or cooked shellfish contaminated by Norwalk-like viruses was documented as being the reason for numerous outbreaks in 1990s (Chalmers and McMillan 1995, Dowell et al. 1995, Leeds et al. 1995).
One outbreak caused by enterotoxigenic E. coli (ETEC) was described in the United States in 1993 (Table 3, outbreak 40). The illness of 47 air passengers was most strongly associated with eating salad. In addition, nine passengers reported gastrointestinal illness from a different flight where the same meal was served. Investigation of a local outbreak at the same time revealed an ETEC outbreak, too. Epidemiological investigation showed the salad to be the source of infection. Carrots were found to be the common ingredient in these salads.
More than 3000 persons were affected on several flights from Melbourne (Table 3, outbreak 38). It was established that catering arrangements were independent, apart from a common supplier of orange juice. Surveys revealed attack rates of illness of up to 100% among orange juice drinkers, and 0% among non-orange juice drinkers. A sudden onset of severe vomiting and diarrhoea developed in between 1 and 3 days. The clinical picture was typical for a viral disease, and the presence of Norwalk-like agent was detected from faecal samples. In this case strong epidemiological evidence between gastrointestinal illness and drinking orange juice was found, although detection of the agents from the orange juice failed.
Air crews have been involved in 11 outbreaks associated with aircraft meals (Table 4). Gastrointestinal illness resulting from food poisoning is the leading cause of airline pilot incapacitation and causes an in-flight safety hazard (Beers and Mohler 1985).
Food contaminated by S. aureus served to crew during a flight led to dangerous situations in the air due to a short incubation period in connection with three outbreaks (Table 4). On the flight from Lisbon to Boston in 1982 all crew members became ill, but fortunately, the crew was still able to operate the aircraft and the plane landed safely.
Crew members eating left over food, oysters from passengers, led to foodborne illness (Table 4). A long incubation period meant that none of the crew members became ill whilst air-borne, but the crew’s illness seriously hampered the operation of scheduled services.
Salmonella infected a great number of crew (135) via appetisers with aspic glaze on (Table 4). The investigation showed that some flight crew members had been eating the same appetisers that were being served to the passengers. The airline company warned their staff of the food incident and reminded them that under airline policies crew members are expected to eat from menus that differ from those served to passengers and to each other. Crew members are also expected to eat at different times (Anonymous 1984). Incapacity of crew members emphasises the importance of providing separately produced meals for the flight crew. Airline crews should be advised of the dangers associated with food and should ensure the safest possible eating habits, especially in developing countries (Masterton and Green 1991).
The most frequent factor leading to the foodborne outbreak via airline food was insufficient refrigeration (Table 5). The next was contamination of the food by an infected food handler. Similar reasons have been shown to be important errors generally leading to foodborne outbreaks (WHO 1995).
The most prominent contributing factors reported in connection with Salmonella outbreaks were infected food handlers and inadequate refrigeration. These non-conformances often appeared together (Table 5). In four outbreaks, food handlers suffering from gastrointestinal symptoms were working in the kitchen either from ignorance or negligence. Investigation of the first outbreak of S . Typhi in 1947 revealed that the itinerant cook was a symptomless carrier and that she was found to have a previous history of typhoid fever 16 years earlier. Inadequate refrigeration with rough errors was detected in five outbreaks. The reasons were: the use of an ice box as the only means of refrigeration, mayonnaise with a pH of 5.3 stored for several days at room temperature, bulk aspic kept in an ambient temperature for three days and still used for glazing cold meals, shortcomings in the cold facilities due to flight delay and the lacking cold storage facilities on a long-haul flight. The misuse of high-risk food items such as mayonnaise and aspic glaze resulted in three outbreaks. Inadequate hygiene standards in the kitchen, such as a lack of hand washing facilities or grossly inadequate toilet facilities, were detected in three cases. Cross-contamination was discovered in two cases.
Investigation of S. aureus outbreaks has only once been traced to an infected food handler (Table 5). A cook with an inflamed finger lesion was the source of infection. The storage of contaminated ham at 10°C in the flight kitchen for 14.5 h, then at room temperature in the galley oven on an aircraft for seven hours resulted in multiplication of S. aureus and the formation of enterotoxin.
A cold dish was the vehicle in all Vibrio outbreaks. Common for the Vibrio cholerae outbreaks was that the food was purchased and prepared in cholera-affected areas. There is an inherent great risk of contamination of cold food and therefore this items should be avoided on aicraft menus in endemic areas (Table 5).
The preparation of cold food, which was also stored at an inadequate temperature by eight food handlers sick with diarrhoea (10% of staff), caused an outbreak. It was also found that the sanitation was unsatisfactory and not according to HACCP (Table 5).
A cook-chill system was not used in 1970 and a foodborne illness of passengers and crew members (Table 3) occurred on eight flights because of inadequate heat treatment and maintenance at 55°C.
Several problem areas where potential contamination could have occurred were identified in the factory producing juice. Plumbing connections were suspected (Lester et al. 1991).