Evaluation of the safety of domestic food preparation in Malaysia.

Evaluation of the safety of domestic food preparation in Malaysia RM. Desmarchelier,1 C. Apiwathnasorn,2 D. Vilainerun,3 C. Watson,4 M.R. Johari,5 Z. Ahmad,6 & A. Barnes7 Food-handling practices were studied in 119 and 158 households, respectively, in an urban and a rural community in Peninsular Malaysia. Hazard analyses, including microbiological analysis of foods, were carried out in two households in each community and in a house that prepared food for distribution in the urban area. Kitchen hygiene was generally acceptable, although rated 'poor" in some instances in the rural area. Food prepared for lunch was usually sufficient for dinner also, the leftover items being stored at ambient temperature until required. In the house that prepared food for distribution, breakfast was prepared during the evening, stored at ambient temperature overnight, and reheated before sale the next morning. There was a local preference for cooking food at temperatures close to boiling point; this reduced the numbers of vegetative cells but not those of spores. In some stored foods the popula- tions of Staphylococcus aureus, Bacillus cereus and mesophilic aerobic bacteria increased, the last- mentioned reaching spoilage levels. Reheating reduced the populations of proliferating bacteria in most foods to acceptable levels but would not have destroyed heat-resistant enterotoxins. Because of their importance in combating acute bacterial foodborne disease, the control of the temperature and time fac- tors during the cooking and storage of food should receive special attention in education on health and food safety. Introduction Foodbome illness, one of the most widespread kinds of disease (1), remains uncontrolled in both devel- oped and developing countries. Its etiology, impact on public health, and epidemiology vary and the prob- lems associated with it in developed countries can be expected to emerge also in developing countries. In developing countries, where family income is low, the household remains a major source of human food. Food is prepared in domestic kitchens for home consumption and for sale through the informal food distribution sector. Culturally acceptable educa- 1 Tropical Health Programme, University of Queensland, St. Lucia, Queensland 4072, Australia. Requests for reprints should be sent to this author. 2 Assistant Professor, Department of Medical Entomology, Faculty of Tropical Medicine, Mahidol University, Bangkok, Thailand. 3 Sanitary Technical Officer, Department of Health, Ministry of Public Health, Bangkok, Thailand. 4 Remote area nurse, Northern Territory, Australia. 5 Microbiologist, School of Medical Sciences, Universiti Sains Malaysia, Kota Bharu, Kelantan, Malaysia. 6 Senior Lecturer, School of Medical Sciences, Universiti Sains Malaysia. 7 Associate Professor, Tropical Health Programme, University of Queensland, St. Lucia, Queensland, Australia. Reprint No. 5544 tion on safe food-handling practices is therefore important to prevent foodbome illness. In order to achieve this it is necessary to identify unsatisfactory practices and the sociocultural factors that influence them.a The "hazard analysis critical control point" approach to the control of food quality has been widely applied in the food manufacturing industry (2). It involves the systematic identification of haz- ards during the production process, the identification of critical control points for these hazards, and the design of procedures for the specific control and the monitoring of such points. This approach is also applicable to domestic food preparation where, how- ever, it has been less frequently employed (3-5). The critical points that have been identified to control the preparation of unsafe food are the targets for food safety education programmes aimed at preventing foodbome illness. The present article describes the food-handling practices identified when this approach was adopted to analyse the preparation of some com- mon foods for domestic consumption and for sale through the informal street food sector. Studied were the general domestic food manage- ment, food-handling practices, and some sociocul- tural factors associated with food-handling in a rural a Health education in food safety. Unpublished WHO document EHE/FOS/88.7, 1987. Bulletin of the World Health Organization, 1994, 72 (6): 877-884 © World Health Organization 1994 877 P.M. Desmarchelier et al. and an urban community in Kelantan, north-east Peninsular Malaysia, in order to identify practices and behaviours that could be used in a food safety education programme. The analyses were specifical- ly designed to target critical control points for deal- ing with contamination by and proliferation of food- borne bacteria that commonly cause illness, e.g., Escherichia coli, Staphylococcus aureus, Bacillus cereus, Clostridium perfringens and Vibrio para- haemolyticus. Materials and methods Study sites The study was performed in both an urban communi- ty, Kampung (Kg) Langgar, and a rural community, Kg Tujoh, in Kelantan State in the north-east of Peninsular Malaysia, where visits were made to 119 households (315 adults) and 158 households (594 adults), respectively. Compared with Langgar, the rural community had larger households, lower household incomes, main care-givers who had lower education levels and poorer access to amenities for good food hygiene, i.e., clean water, sanitation, cooking fuel, and refrigeration. Cooking area hygiene Cooking area hygiene was assessed by observation during the interview with the key respondent in each household, when a standard checklist of questions was used to avoid differences between interviewers. Observations were scored as 1 or 0 (corresponding to "yes" and "no", respectively) for a separate kitchen, an unclean cooking stove, food scraps, an unclean floor, uncovered food, no waste container, and evi- dence of the presence of animals or birds. Scores of 0-2, 3-4 and 5-7 were categorized as "good", "acceptable", and "poor", respectively. Hazard analysis The hazard analysis critical control point approach was used to assess the preparation of midday meals in two households (1 and 2) in Kg Langgar and in two households (4 and 5) in Kg Tujoh, and that of nasi bunkus (breakfast food) in a third household (3), also in Kg Langgar, that sold food via street vendors in this community. Each household had 10-12 occu- pants. All the households were supplied with treated town water and had access to a pour-flush toilet; cooking was carried out using a gas stove in all but household 4, which had a wood-burning stove. Only household 3 had a refrigerator, although there was also one in a coffee shop owned by and close to household 2, and fresh food, e.g., uncooked meat, was sometimes stored here. The communities were approached through local leaders and householders using methods described previously (2-5). The food handlers were requested to prepare food in their usual way and were given no specific instructions. The researcher was accompa- nied by an interpreter so that all communication was in the local dialect of Bahasa Malaysia. In most instances, both remained in the household while food was being prepared and stored, observing and recording the procedures employed and the behavi- our of the food handlers; food temperatures were measured and samples were collected for bacterio- logical analysis. However, when nasi bunkus was being prepared, the observers were unable to stay on the premises during the entire evening. In instances when rice, fish, or meat were the main ingredients, main samples were collected; for meals that consist- ed of numerous ingredients, the composite food was sampled after cooking. Bacteriological analysis Food samples were collected under sterile conditions in plastic containers using knives and spoons and were immediately placed on ice in a cooler box for transport to the laboratory, where they were stored at 4 °C for testing on the day of collection or at -20 °C for testing later. A 30-g portion, or the whole sample if it was smaller, was diluted 1:10 in sterile 0.1% buffered peptone water,b homogenized using a Waring blend- er, and further diluted as required. Foods were tested quantitatively for total mesophilic aerobic bacteria, coliforms, E. coli, coagulase-positive staphylococci, B. cereus, C. perfringens and V. parahaemolyticus (seafood only). The isolation and identification methods used were those contained in the Standards Association of Australia, AS1766,c with the follow- ing exceptions, where a different method or a mod- ification had to be used in a laboratory with minimal facilities: for E. coli, the Anderson-Baird Parker membrane method; for staphylococci, the Staphyslide latex slide coagulase test;d for B. cereus, polymixin egg yolk mannitol agar;b for C. perfringens, tryptone sulfite cycloserine agar;b and coliform bacteria, Petrifilm VRB plates.e b Oxoid Ltd, Basingstoke, England. c Australian Standard 1766: Food microbiology. Sydney, Stan- dards Australia, 1991. d BioMerieux Laboratory Reagents and Instruments, Marcy- l'ltoile, France. e Mauri Foods, Burns Philp & Co. Ltd, Australia. 878 WHO Bulletin OMS. Vol 72 1994 Safety of domestic food preparation In Malaysia The criteria used for the acceptability of the bac- terial quality of food samples were the standards out- lined in Malaysia's Food Act. For ready-to-eat foods these are as follows: mesophilic aerobic bacteria, <3.0 x 105 per g; coliforms, <50 per g; S. aureus, <100 per g; B. cereus, <100 per g; and C. perfringens, <100 per g. Results Cooking area hygiene All the food preparation areas in the 119 households in the urban kampung were satisfactory (96%, ",good"; 4%, "acceptable"). In the 158 households in the rural area, 88% of food preparation areas were satisfactory (49%, "good"; 39%, "acceptable") and 11% were "poor". Hazard analysis Flow diagrams, showing hazards and critical control points, were prepared for each meal whose prepara- tion was observed. Examples are shown for the preparation and storage of boiled rice (Fig. 1), which was used in all households and for the fish ball soup prepared in household 1 (Fig. 2). The bacterial counts in food samples are shown in Table 1. The time and temperature relationship of food during preparation and storage was a major critical con- trol point in all households (Table 2). Household 1. In household 1 the preparation of boiled rice, fried fish and fish ball soup began at 12:00 and lunch was served at 14:00. Other food was covered and stored at ambient temperature until it was reheated for dinner at 20:00. Raw fish was pur- chased from an itinerant hawker who carried it at ambient temperature. The fried fish reached a maxi- mum temperature of 82 'C for 15 minutes and was then transferred to a plate and left under ambient conditions, where the temperature decreased from 80 'C to 33 'C over 60 minutes. The fish balls were prepared from ground raw fish, added to soup, and boiled at 99 'C for 15 minutes (Fig. 3). The soup was then left on the stove with no heat for 135 min- utes, where the temperature dropped from 98 'C to 40 °C. The raw fish had 60 coliforms/g, whereas the fried fish and fish ball soup contained only 10 and <10 coliforms/g, respectively. Household 2. In household 2 the food prepared included rice, fried fish, eggplant curry and boiled fish, all of which were served for lunch. Leftover eggplant curry and boiled fish were covered and stored at room temperature, subsequently to be Fig. 1. Hazard analysis and critical control points for preparation of boiled rice in all study households. CCP1: temperature eat promptly CCP2: time Key 0 Initial contamination likely T Contamination from equipment/utensil surfaces likely () Contamination by persons handling food likely CCP Critical control point o) Survival of vegetative bacterial cells likely 8) Inactivation of vegetative forms D Propagation of bacteria likely 4 Spores reheated before consumption at dinner. Cooking commenced at 13:00 and lunch and dinner were served at 14:30 and 20:00, respectively. The rice was boiled in JKR water (town water supply) at 102 °C for 15 minutes and was then left on the stove with no heat, its internal temperature dropping from 99 °C to 31 °C during storage for 5 hours. The raw rice had a high count of coliforms and B. cereus; boiling reduced the coliform count significantly but not that of B. cereus (Table 1). After the boiled rice had been held at 31.4 "C for 3 hours after lunch, the coli- form and S. aureus counts had increased, while the B. cereus count had decreased slightly. During prepara- tion for dinner the rice was reheated to a temperature of 71 "C for 15 minutes; this reduced the bacterial WHO Bulletin OMS. Vol 72 1994 879 P.M. Desmarchelier et al. Fig. 2. Hazard analysis and critical control points for the preparation of fish ball soup in household 2 (see Fig. 1 for explanation of symbols). counts to levels similar to those present after the initial boiling. When served for lunch, eggplant curry had acceptable bacterial counts, except for mesophil- ic aerobic bacteria. After storage at room tempera- ture (31 °C) the B. cereus count increased (Table 1). After reheating the curry to 60.2 "C and serving for dinner, the mesophilic aerobic bacteria and B. cereus counts decreased and contamination with coliforms was detected. Household 3. In household 3 the food prepared included nasi lemak, nasi berlauk, and nasi kerabu, which are common breakfast dishes consisting of boiled rice with various combinations of fish, meat curry, vegetables, and sauces. The food was prepared between 22:00 and 24:00, reheated at 5:00 the fol- lowing moming and sold between 6:00 and 10:00 at roadside food stalls. It was estimated that more than 300 servings were sold daily in the kampung. It was not possible to remain in the house during the entire food-handling period. The preparation of nasi ber- lauk, a fish or beef curry with boiled rice, was observed. The rice was prepared in a similar manner to that indicated above. The fish and beef curries were boiled at 99.0 °C and 96.8 °C, respectively, for up to 45 minutes; after ovemight storage at room temperature, the curry temperatures were 33 °C and 28 °C, respectively. Following cooking in the eve- ning, the curries, except for the beef curry, had acceptable bacterial levels. After ovemight storage at ambient temperature for 5-7 hours, the mesophilic aerobic bacteria had increased approximately ten- fold. Following reheating and storage at a roadside food stall for a further 5 hours, the counts were acceptable but the count of B. cereus had increased a hundredfold in the beef curry. Household 4. In household 4 the meals prepared included rice, siakap puteh (fried sea perch) and sayur air (boiled fish and vegetables). Preparation began at 12:50 and lunch was served at 13:50. The leftovers from lunch were served at the evening meal, but it was not possible to make any observa- tions in this regard. Food samples were collected after lunch, covered, and stored in the laboratory for 4 hours at the ambient temperature recorded in the household. During cooking, the rice, sayur air, and fried fish reached 86.9 °C, 97 °C, and 89 °C, respec- tively. Fish caught in the kampung were fried on the moming they were caught, while market fish that had been caught several days before were used for the preparation of sayur air. The local fish had 60 coliforms/g initially and <10/g after frying, and there was no increase on holding. The market fish had ele- vated mesophilic aerobic bacteria, coliform, and B. cereus counts, which were reduced during cook- ing and the preparation of sayur air; however, the mesophilic aerobic bacteria increased to 1.2 x 109/g after holding for 4 hours. Household 5. The adult residents of household 5 worked in nearby tobacco fields, returning at midday to prepare lunch, which was consumed between 14:00 and 16:00; the leftovers were usually con- sumed at 22:00 when they returned from the fields. Boiled rice, fish soup, and kukus masam manis (sweet-and-sour steamed fish) were prepared during the visit. No samples were available. The rice was cooked at a maximum temperature of 94 °C for 30 minutes over a wood fire; this reduced the meso- philic aerobic bacteria and B. cereus counts to acceptable levels at the time of serving. The raw fish, purchased from local hawkers, was imported and had a mesophilic aerobic bacterial count of 7.0 x 106/g. The fish soup and kukus masam manis had acceptable bacterial counts after cooking for WHO Bulletin OMS. Vol 72 1994880 Safety of domestic food preparation in Malaysia Table 1: Bacteriological quality of food sampled during hazard analysis in the urban and rural study areas, Malaysia Mesophilic aerobic Time Temperature bacteria Coliforms S. aureus B. cereus Food item sampled (OC) (x 1 05/g) (per g) (x 1 02/g) (x 1 02/g) Urban - household 1 Raw fish Minced fish for fish ball soup Fried fish Fish ball soup Raw rice Boiled rice Urban - household 2 Eggplant curryb Eggplant curryc Eggplant curryd Raw rice Boiled riceb Boiled ricec Boiled riced Urban - household 3 Fish curryb Fish curryc Fish curryd Beef curryb Beef curryc Beef curryd Rural - household 4 Raw rice Boiled rice Raw fish for sayur air Sayur airb Sayur airc Raw fish for frying Fried fishe Fried fishc Rural - household 5 Raw rice Boiled ricee Raw fish for kukus masam manis and soup Kukus masam manise Fish soupe 12:00 28.0 <3.0 12:00 30.0 <3.0 13:00 80.0 <3.0 12:45 98.0 <3.0 11:30 -a -a 12:30 99.0 <3.0 15:30 18:00 18.45 13:00 15:00 18:35 18:45 23:00 05:00 10:30 24:00 05:30 10:30 10:00 11:00 12:50 14:00 18:00 12:00 12:30 16:00 12:35 14:00 12:00 14:10 14:00 48.2 31.0 60.2 -a 71.6 31.4 71.0 87.0 33.0 41.0 78.0 28.0 30.0 32.0 -a 24.0 58.0 30.0 32.0 -a 30.0 33.0 86.0 33.0 74.5 50.6 8.5 4.0 1.3 >3.0 <3.0 5.6 <3.0 60 <1.0 <1.0 60 <1.0 <1.0 10 <10 10 10 20 10 3.0 x 103 >2.5 x 106 3.1 X102 3.3x 103 <10 <3.0 9.0 <30.0 >3.0 x 102 6.8 x 102 <30.0 <3.0 <3.0 29.0 <3.0 1.2 x 104 1.6 <3.0 <3.0 30.0 <3.0 70.0 <3.0 <3.0 1.1 X 102 <10 3.0 x 103 <10 <10 60 <10 <10 -a -a <10 <10 <10 <1.0 <1.0 <1.0 <1.0 <1.0 <1.0 <1.0 <1.0 <1.0 <1.0 <1.0 <1.0 <1.0 2.0 <1.0 <1.0 <1.0 <1 .0 <1.0 <1.0 <1.0 <1.0 <1.0 <1.0 <1.0 <1.0 <1.0 <1.0 <1.0 <1.0 <1.0 <1.0 <1.0 <1.0 <1.0 2.0 1.0 12.0 14.0 4.0 <1.0 <1.0 <1.0 <1 .0 <1.0 <1.0 2.0 <1.0 <1.0 1.7 <1.0 <1.0 <1.0 <1.0 <1.0 1.0 <1.0 <1.0 <1.0 <1.0 a Not tested. b Food sampled after cooking. c Food sampled after holding. d Food sampled after reheating. e Food sampled at serving. approximately 10 minutes at 94 IC and for 20 min- utes at 83-88 °C, respectively, and after standing at room temperature for 1 hour. Discussion and conclusions The communities had access to various amenities allowing the practice of safe food hygiene - ade- quate quantities of safe water, safe sanitation, and gas or wood for cooking. The hygiene standard in the food preparation areas was acceptable in most of the kitchens in both communities, although overall the urban kitchens were rated higher than the rural kitchens. Personal hygiene was not specifically studied. This is a sensitive matter and responses would have WHO Bulletin OMS. Vol 72 1994 881 P.M. Desmarchelier et al. Table 2: Temperature of food during cooking and storage and temperature duration recorded during hazard analysis Time at Time at Temperature Food a Time at <60 °C >70 °C attained on item Tmax Tma. (min) (min) (min) reheating (°C) Household 1 Rice 102 _b 5.5 140 Fried fish 82 15 6.2 25 - Fish ball soup 99 15 5 60 96 Household 2. Rice 102 15 3.3 76 71 Curry 97 15 3.5 104 60.2 Fried fish 97 15 4.4 24 64 Household 3 Rice 86.9 - - 63 - Sayur air 97 10 - 21 Fried fish 89 -c - -c - Household 4 Rice 94.7 30 0 116 - Fish soup 94 10 84.5 27 - Kukus masam manis 86 20 0 29 - a Tmax = maximum temperature. b Not relevant. c Not tested. required verification, something that was not pos- sible. The personal hygiene of food handlers is important during food preparation. However, since all the food observed was cooked, its contamination by food handlers after cooking was the more impor- tant hazard. The opportunity for contamination by food handlers before the first serving appeared limit- Fig. 3. Internal temperature of food during preparation and storage in household 2. 100 °80 0.mv 60 CL E 40 H 20 [ (Z) p c C)i v 11:00 12:00 13:00 14:00 15:00 16:00 17:00 18:00 19:00 20:00 Time of day ed, since most of the foods were semisolid or liquid and were served with clean utensils from the cooking pot to individual plates before being eaten by hand. Contamination from utensils was possible during serving but not from storage containers, since left- overs were generally stored in the original cooking pots. In household 2 there was evidence of contami- nation of the eggplant curry, in which a marked increase in coliforms was detected after reheating; this probably originated on the utensils used to stir the food. Hazards associated with poor personal hygiene were thus not highlighted, although it is emphasized that such hazards are important, espe- cially in the handling of snacks and other foods that are not heated again before consumption. Based on the local food standards for ready-to- eat foods we conclude that three of the five domestic kitchens studied served one or two meals of unac- ceptable microbiological quality. In two instances, curries were unacceptable after primary cooking, storage, and reheating. In household 2, boiled rice had unacceptable bacterial counts after cooking while, in household 4, sayur air had unacceptable counts after storage. Failure to meet the standards was mainly attributable to high mesophilic aerobic bacterial counts and, less markedly, to high counts of B. cereus, S. aureus, and coliforms. Although some of the foods were of unacceptable microbiological quality, there were no obvious organoleptic changes. WHO Bulletin OMS. Vol 72 1994 Serve 120U, n . ... 882 Safety of domestic food preparation in Malaysia The practices shown to affect the bacterial counts of the majority of the foods were cooking, storage and reheating. There was a preference in both communities for food cooked in liquids that were close to boiling point, and the temperatures and durations of cooking were sufficient to bring about significant reductions in the numbers of vegetative cells of most bacterial contaminants. Thus the counts of aerobic mesophiles, coliforms, and B. cereus in raw fish were reduced to acceptable levels when it was fried, steamed, or boiled. The boiling of rice reduced the bacterial counts to acceptable levels except in household 2. In this household the raw rice had high levels of aerobic mesophilic bacteria, coli- forms, and B. cereus, of which, after cooking, the mesophilic bacteria were sufficiently reduced, but the coliforms, although reduced by a factor exceed- ing 104, remained above the acceptable level; the numbers of B. cereus were little changed, perhaps because of the net effect of thermal destruction of vegetative cells of B. cereus on the uncooked rice and of the simultaneous thermal shock of heat-resist- ant spores and their resultant outgrowth. Since the cooking was generally found to be satisfactory, the immediate consumption of freshly cooked food should present little hazard from foodbome bacteria requiring large infective doses. The main food eaten raw was fruit, which was more likely to present haz- ards in the form of low-infective-dose foodbome dis- eases such as typhoid fever, shigellosis, and viral diseases. In all the households it was common practice to prepare sufficient food late in the moming for both lunch and dinner. This resulted in the prolonged stor- age of cooked food at ambient temperature, even where there was access to a refrigerator. The ambient temperature in the food-storage areas averaged 30.5 °C and the duration of storage for seven of the eight observations was 3.3-6.2 hours. The foods and the environmental conditions were ideal for the pro- liferation of most foodbome pathogens and spoilage bacteria. The counts of mesophilic aerobic bacteria on curries and boiled rice increased during storage, and those in sayur air increased to 1.2 x 109/g, approaching spoilage levels. Counts of B. cereus and S. aureus increased in curries and rice, respectively. No tests were performed to detect the enterotoxins derived from these bacteria, but conditions were suit- able for their production. Stored food was not always reheated before serving. When precooked stored food was reheated, the process was less effective than initial cooking in reducing bacterial loads, since the tempeatures reached and the duration of heating were too low. This was related to local preferences, since very hot food was not acceptable and eating with the hands was common. Reheating was effective in reducing the numbers of bacteria in most of the foods studied; however, it would not have been sufficient to destroy heat-resistant enterotoxins, e.g., staphylococcal entero- toxins and the emetic enterotoxin of B. cereus, which could have been produced in the presence of higher numbers of these bacteria. A critical control point is a practice or procedure at or by which a preventive or control measure can eliminate, prevent, or minimize a hazard (2). Our observations and hazard analyses identified the fol- lowing common critical control points for the food preparations: temperature and the period of holding at a given temperature during cooking, storage, and reheating. Adequate cooking eliminates or reduces the numbers of bacterial pathogens, and storage con- ditions that retard bacterial growth prevent the pro- liferation of surviving pathogens or post-cooking contaminants. These findings are similar to those identified in other hazard analysis studies of domes- tic circumstances in tropical countries (3-5). The inclusion of critical control points in health education messages requires that practical monitor- ing processes be devised to ensure effectiveness. Cooking temperatures for semisolid or liquid food can be monitored by ensuring that the cooking liquid boils and simmers and that minimum times for cook- ing are specified. The preference for well-cooked staple food in the study communities was advanta- geous, since the required controls provided a cultur- ally acceptable product except when contamination of the raw ingredients was excessive. Solid foods were generally fried in hot oil, which should also eliminate pathogenic vegetative cells. The monitor- ing of food storage mainly involves the control of the time of storage at ambient temperature. In our study, growth was detected at storage times . 3.5. hours; further work is required to determine safe storage times either in the home or at street vending stalls for the common foods consumed in these communities and kept under local environmen- tal conditions. The critical control points identified are recom- mended as targets for behavioural changes and health education messages. However, behaviours specific to the study communities need to be considered before any intervention takes place. Studies on the percep- tions of foodbome disease in the communities and on attitudes to food storage will appear later. Acknowledgements This work was funded in part by the WHO Western Pacific Centre for Environmental Planning and Applied Sciences. We acknowledge the technical assistance of Mrs M. WHO Bulletin OMS. Vol 72 1994 883 P.M. Desmarchelier et al. Yunus and Miss N. Jamaludin; the assistance of local interpreters; and Mauri Foods, Burns Philp & Co. Ltd, for the donation of the Petrifilm VRB plates. Resume Evaluation de la salubrite de la preparation domestique des aliments en Malaisie Les cuisines domestiques sont une source impor- tante d'aliments dans les communautes a faible revenu et les aliments sont destines tant a la con- sommation des familles qu'a la vente dans le sec- teur alimentaire parallele. 11 est d'une importance capitale pour la sant6 publique de manipuler les aliments en pr6servant leur salubrit6. L'education sur la salubrit6 des aliments doit reposer sur l'identification de points de contr6le essentiels et sur la mise en oeuvre de mesures de lutte cultu- rellement appropri6es. Une 6tude visant a rep6rer les pratiques rela- tives a la manipulation des aliments constituant un danger de toxi-infection alimentaire bacterienne aigue a ete realis6e dans 277 foyers du nord-est de la peninsule malaysienne, tant en milieu rural qu'en milieu urbain. L'hygiene des cuisines do- mestiques a ete 6valu6e par observation, grace a une liste de contr6le standardisee. L'analyse des risques a 6t6 effectu6e dans deux foyers au sein de chaque communaut6, ainsi que dans un eta- blissement produisant des aliments destines a la distribution par des vendeurs des rues en secteur urbain. Les observations ont porte sur la manipu- lation des aliments a tous les stades de leur pre- paration et de leur conservation jusqu'a leur con- sommation; la temp6rature a ete mesur6e et des echantillons prelev6s. La pr6sence de germes ali- mentaires courants et de germes indicateurs a ete recherchee dans les aliments. L'hygiene des cuisines est en gen6ral accep- table, bien que m6diocre dans certaines cuisines en milieu rural. Une pratique courante observ6e dans les deux communaut6s consiste a preparer aux environs de midi une quantit6 suffisante d'ali- ments pour le d6jeuner et le repas du soir, le reste 6tant conserv6 a temperature ambiante jus- qu'a consommation. Dans le cas de l'6tablisse- ment fabriquant des aliments destin6s a la vente, le petit d6jeuner est pr6par6 le soir, conserve a temperature ambiante pendant la nuit, et r6chauf- f6 avant de le mettre en vente le lendemain matin. On remarque une predilection pour la cuisson a temp6rature proche de l'6bullition, ce qui dimi- nue le nombre de germes a l'6tat veg6tatif, mais pas celui des spores. Pendant la periode de con- servation prolong6e entre la cuisson et la con- sommation ou le rechauffage, les populations de bact6ries aerobies m6sophiles, Staphylococcus aureus et Bacillus cereus se d6veloppent dans certains aliments, la quantit6 de bact6ries a6ro- bies m6sophiles atteignant dans quelques cas une abondance telle que les aliments sont im- propres a la consommation. Le rechauffage a plus de 60 0C ramene a un niveau acceptable le nombre de bact6ries qui proliferent, mais la tem- perature atteinte ne permet pas de detruire les enterotoxines thermoresistantes. Les aliments ne sont pas toujours r6chauff6s et ceux qui sont ven- dus dans la rue sont souvent conserv6s de nou- veau a temp6rature ambiante apres r6chauffage avant d'etre consommes. Le contr6le de la temperature et de la dur6e de la cuisson et de la conservation des aliments est un point de contr6le essentiel identifie dans les analyses de risque r6alis6es dans les deux communaut6s; c'est sur ce point que les pro- grammes d'education pour la sante et pour la salubrite des aliments destin6s a lutter contre les toxi-infections alimentaires bacteriennes aigues de- vront faire porter leurs efforts. References 1. Esray SA. Food contamination and diarrhoea. World health, January-February 1990: 19-20. 2. Silliker JH et al. Microorganisms in food. 4. Appli- cation of the hazard analysis critical control point (HACCP) system to ensure microbiological safety and quality. Oxford, Blackwell, 1988: 44-61. 3. Michanie S et al. Critical control points for food prepared in households in which babies had salmo- nellosis. International journal of food microbiology, 1987, 5: 337-354. 4. Michanie S et al. Hazard analyses of foods pre- pared by inhabitants along the Peruvian Amazon River. Journal of food protection, 1988, 51: 293-302. 5. Michanie S et al. Critical control points for food prepared in households whose members had either alleged typhoid fever or diarrhoea. International jour- nal of food microbiology, 1988, 7: 123-134. 884 WHO Bulletin OMS. Vol 72 1994