Microbiological quality (I-III)
Meals served on aircraft were collected for a survey of microbiological quality including Salmonella analyses between 1991 and 1994 and for the specific Salmonella survey between 1989 and 1992. Samples were collected in accordance with the control programme of an airline company.
Altogether 1687 dishes were sampled for microbiological quality. This material consisted of 1012 hot and 675 cold dishes. The hot dishes comprised 112 breakfasts and 900 main dishes, and the cold dishes 273 appetisers, 168 salads and 234 desserts. Approximately half of the dishes 791 (47%) were prepared in Finland and the remaining 896 (53%) in 32 other countries.
The material for the microbiological survey, including Salmonella , consisted of 1011 hot meals and 653 cold meals. The material for Salmonella examinations consisted of two surveys. During the specific Salmonella survey, 1288 hot dishes and 923 cold dishes, such as 400 appetisers, 278 salads, 19 cheese dishes and 226 desserts, 2211 samples in all, were collected. Of the samples 1477 (67%) were of Finnish origin. The remaining 734 (33%) were prepared worldwide by flight kitchens located in 28 countries. Altogether 2299 hot meals and 1576 cold meals were tested for Salmonella in the studies.
Onboard sampling took place before serving. Hot aircraft dishes of pre-cooked food were sampled before the final re-heating on board. The samples were stored using dry ice and frozen at -20°C immediately after the flight. They were carried frozen within 1 to 3 days to the laboratory. Samples taken from Finnish flight kitchens for the Salmonella study were chilled to 4°C and transported chilled to the laboratory. For the microbiological survey between 1991 and 1994, the samples were delivered frozen to the laboratory.
Salmonella outbreak (IV)
During inspection of the Salmonella outbreak 153 food samples were taken. They consisted of 148 food samples of cooked egg products, raw chicken products, cold cuts and Swiss rolls sliced in the cold kitchen from the catering establishment. Two samples represented a batch with 1200 portions of Viennese goulash prepared on the 30th July for charter flights. Raw materials to simulate the preparation of the dishes such as Viennese goulash, fresh salad and Swiss roll served on the Rhodes flight were sampled.
Human faecal samples for Salmonella were collected in connection with investigation of the Salmonella outbreak. In the catering establishment all the staff were sampled: 118 food handlers and 44 persons from other departments. In order to be tested for Salmonella , railway and airline passengers at risk, 600 and 350, respectively, were informed by two nationwide announcements to contact their local health authorities. The total number of railway and airline passengers tested is not known, only the numbers of the positive Salmonella findings 107 and 91, respectively, have been recorded.
Hand and nasal samples of Finnish flight catering employees were collected between January 1995 and May 1997. Altogether 153 hand samples from 117 persons and 136 nose samples from 111 persons were taken. In most cases sampling was done once.
Nasal samples were taken by applying a sterile cotton-tipped swab 1-2 cm inside the anterior nares. To sample the hands, the right hand was rinsed for 10 seconds in a sterile plastic bag containing a 20 ml physiological saline solution with 1% peptone (Bacto Peptone, Difco Laboratories, Detroit, Mich.).
The methods used for the microbiological analyses of food samples are included in the collection of the Nordic Committee on Food Analysis (NCFA) and they are presented in Table 1 (I) and Table 1 (II). These methods are in common use for official control of food in the Nordic countries. The Salmonella method used was NCFA 1991 (I-III) and NCFA 1986 (IV). Before analysis, all the ingredients of the sample were homogenised. For the quantitative methods 10 g and for the qualitative methods 25 g of homogenate were used.
Faecal samples for Salmonella were examined by the method of Kelly et al. (1985). Subtyping of the Salmonella species was carried out by the National Salmonella Centre of Finland, National Public Health Institute (KTL) using the method of Kauffman (1966).
Blood agar (Blood agar base, Difco) containing 5% sheep blood and Baird-Parker agar plates (Lab M, Burry, UK) was used to isolate S. aureus from nasal swabs. For the hand samples, Baird-Parker agar was used. Wherever possible, at least five typical colonies were picked from both plates and further subjected to Gram-stain, catalase reaction and coagulase test performed with swine plasma. Gram positive, catalase positive cocci that were coagulase positive and had shown typical reaction on Baird-Parker agar plates were regarded as S. aureus .
The pH value was measured from the homogenate of 10 g of food mixed with 10 ml of distilled water (Digital-pH-Meter E 632, Metrohm, Switzerland). The water activity (a w ) was measured according to the method of the NCFA (1984) using a hygrometer (Lufft GmbH, Stuttgart, F.R.G.).
All isolated S. aureus strains (42) were characterised using DNA macrorestriction analysis. Cells were harvested from 2 ml of BHI broth (Oxoid, Basingstoke, UK) and cultures grown overnight at 37°C. In situ DNA isolation was performed as described by Maslow et al. (1993) with the modifications described by Björkroth et al. (1996). Sac I and Sma I were used for the cleavage of DNA. The samples were electrophoresed through 1% (w/v) agarose gel (SeaKem Gold, FMC BioProducts, Rockland, Maine) in 0.5 x TBE (45mM Tris, 4.5 mM boric acid, pH 8.3, and 1 mM sodium EDTA) at 14°C in a Gene Navigator system with the hexagonal electrode (Pharmacia, Uppsala, Sweden). Interpolation ramped from 0.5 s to 15 s for 18 h at 200 V was used for Sac I and from 0.5 s to 25 s for Sma I.
Enterotoxin determination was performed on the basis of the PFGE types. From each of the 32 different PFGE pattern groups, one to three isolates were tested for enterotoxin production. From the biggest type-specific groups containing isolates from different persons more than one isolate was tested. In these cases the isolates chosen were from different persons. Toxin detection was performed using the ELISA test for staphylococcal enterotoxins A, B, C and D with the test kit SET-EIA from Dr. Bommeli AG (Liebefeld-Bern, Switzerland) according to the manufacturer’s instructions.
Standardised questionnaires were sent to all 162 employees of the catering establishment and to all 350 air passengers after the flight on 1st August from Helsinki to Rhodes. Thirteen railway passengers who had suffered gastroenteritis after travelling by train on the 1st, 2nd and 3rd of August and were living in the Helsinki area were interviewed by telephone. They were asked about the time of the onset of the symptoms, the symptoms and the duration of the illness, and the food they had eaten. A wider questionnaire study could not be carried out because the names of the other railway passengers were not known.
The local health authorities inspected the catering; storage of raw materials, food preparation areas, refrigeration equipment, general cleanliness and repair of the equipment and establishment. The food preparation methods of egg sandwiches for trains and the meal served on the Rhodes flight were especially checked.
For the statistical analyses of hot and cold meals, Student’s t-test and Chi 2 -test were carried out using Statistica for Macintosh TM (Tulsa, Oklahoma). In connection with the investigation of Salmonella outbreak statistical testing of questionnaires was done by using the Chi 2 -test.