Skip to main content

Foodborne Salmonella outbreak among enterally fed patients, May 1997

On May 7-9, 1997 (1-3 Moharram 1418) nine patients in a 180-bed hospital and a 50-bed extended care unit (ECU) in northern Saudi Arabia developed febrile gastroenteritis. All nine patients were among 22 patients who were on nasogastric tube feedings (NGT) prepared in the hospital kitchen. An epidemiological investigation was initiated to determine size, extent, possible source and factors contributing to this outbreak including possible defects in the eternal feeding.
A gastroenteritis case was defined as any patient or staff member in the hospital or ECU who developed diarrhea (more than two unformed stools per 24 hours) or had Salmonella enterica enterica isolated from a stool specimen from May 1 through May 10, 1997. Case finding included review of the medical records of all NGT patients at both facilities, of other patients present during the study period, and of microbiology records for all stool cultures done during the study period.
Only the nine cases originally reported by the hospital were found. All six NGT patients in the ECU had fallen ill (100% attack rate) between 2:30 p.m. and 7:30 p.m. on May 7. In the hospital, three out of sixteen NGT patients in different rooms were affected (19% attack rate) with onset on May 7 (one patient) and May (two patients). The median age of the case-patients was 58 years (range=15-90 years). In addition to diarrhea, all case-patients had fever (38-40°C), 67% had vomiting and 78% had hypotension (<90/60 mm Hg). All nine patients recovered from the gastroenteritis within five days. Salmonella enterica enterica Group D (non-typhi) was recovered from the stools of two patients and from the nasogastric aspirates of two other patients
All ill and well NGT fed patients had exactly the same diet administered in the same way and given at the same time for each of the three preceding meals. Accordingly, no difference in rates of illness could be demonstrated by food exposure. NGT breakfast served on May 7 (date of illness) included eggs, bread, and milk. Lunch and dinner served on May 6 included, boiled vegetables, fruit, bread, and milk, with mutton for lunch and chicken for dinner. All food items for each meal had been mixed in a common blender.
Interviews of cooks and dietitians and direct observation of the NGT food preparation revealed several key defects in the preparation, storage and distribution of the NGT foods. For the most part these foods were kept after cooking at incorrect temperatures (10 to 60°C) for extended periods (two to five hours) before being fed to patients.
Food for all patients and staff is initially prepared for cooking in the same area of the kitchen. However, NGT food and special diets are separated from the rest of the hospital food for cooking. The food is then cooked and pureed in a different area of the kitchen.
Lunch, dinner and breakfast were exposed to environmental temperatures from 28 to 33°C between cooking and arriving at the wards. ECU meals had additional exposure to temperatures up to 51°C. Although stored in a ward refrigerator prior to being served, air temperature near the food containers inside the refrigerator did not drop below 15°C up to four hours after food was cooked. Food was taken from the refrigerator and fed to patients hourly for three hours. The total time between cooking and final feeding was two to five hours.
The actual source of salmonella was not verified. However observations of cooking and preparation revealed that the cook could have deboned cooked chicken while skinning raw chicken. Unlike the chicken there was no identifiable mode of cross contamination between cooked and uncooked eggs or mutton.
Cultures taken five to 10 days after the outbreak of raw foods, the remnant water after thawing, the bloody fluid on the cutting board, equipment and utensils did not yield pathogens. Escherichia coli was recovered from a plastic cutting board and two other utensils. No enteric pathogens were isolated from any of the kitchen staff.

Editorial note:

The explosiveness of the outbreak, the clustering of illness onsets within few hours and the growth of salmonella from gastric aspirate point to a large dose of organisms and probably a short incubation period. Accordingly, the investigation focused on chicken from the previous day, eggs from breakfast of the same day, and chicken and mutton from the previous day. All three suspect meals were exposed to incorrect temperatures before and after cooking, during transportation, and serving. The fact that ECU patients had a an attack rate of more than five times that of the hospital, clearly reflects the additional temperature abuse of the ECU meals.
One of the Salmonella enterica enterica serotypes in group D, enteritidis is likely to be from eggs or chicken. Chicken was a more likely source and vehicle because we identified a clear mechanism of cross contamination between the raw and cooked chicken. Eggs are also a possible source of infection since Salmonella serovar enteriditis may produce transovarial infection from hen to egg and may survive inside eggs boiled up to eight minutes[1]. Improper washing of utensils, hands, or equipment used for preparing the liquid diet are other possible methods of cross-contamination.
The problem represented by this outbreak may be common since clusters of gastrointestinal illness occur commonly in health facilities, and may not be identified as outbreaks[2]. Salmonellosis outbreaks have consistently been the major contributor to foodborne morbidity and mortality in nursing homes. Salmonellosis was responsible for 52% of outbreaks and 81% of deaths in USA nursing homes for the period 1975-1987[3].
Foodborne nosocomial illness is perhaps the most common preventable of all hospital-acquired morbidity. With proper attention to nosocomial infection surveillance, food service practice, and maintenance of employee health, food borne nosocomial illness could become a rarity. To ensure food safety, appropriate temperatures for food must be maintained, contamination of cooked food by raw food, infected food handlers, or contaminated equipment must be avoided, and food service personnel must be educated in correct food handling procedures[2]. Enteral feeding preparation also needs to be handled carefully. A quality control program such as the Hazard analysis critical control point (HACCP)[4] needs to be established and activated in all MOH hospitals.
  1. Anglim AM, Far BM. Nosocomial gastrointestinal tract infections. In: Mayhall GC, editor. Hospital epidemiology and infection control. Baltimore: William and Wilkins; 1996. p. 196-225.
  2. Mishu B, Koehler J, Lee LA, et al. Outbreaks of Salmonella enteritidis Infections in the United States, 19851991. J Infect Dis 1994;169:547-52.
  3. Levine WC, Smart JF, Archer DL, et al. Foodborne disease outbreaks in nursing homes, 1975-1987. JAMA 1991; 266: 2105-2109.
  4. Bryan FL. Hazard analysis critical control point evaluations. A guide to identifying hazards and assessing risk associated with food preparation and storage. Geneva: WHO, 1992.