Salmonella outbreak associated with tiramisu dessert, Riyadh, Dec 28-31, 1997

On Jan.7, 1998, the Field Epidemiology Training Program received a telephone report from a physician at King Faisal Specialist Hospital and Research Center concerning unacquainted gastroenteritis patients who had eaten at the same restaurant in Riyadh. In addition, one patient stated that similar cases had occurred after a supper hosted by a bank on Dec. 28, 1997 and catered by the same restaurant. We started a retrospective cohort study to determine the size, extent, and possible source of the infection, and to recommend preventive measures to avoid any future foodborne outbreaks.

A case of gastroenteritis was defined as any person who developed diarrhea within 72 hours after eating any food from the restaurant between Dec. 28 and 31, 1997, or after eating food at the bank supper, or also if Salmonella enterica group D was isolated from a stool culture. In the community outbreak, all gastroenteritis cases and a random sample gastroenteritis case were reviewed in 5 hospitals and 1 polyclinic from Dec 28, 1997, to Jan 5, 1998, to locate any missing cases. Thirty-seven bank party attendees and the families of nine community cases were asked about the restaurant foods they had eaten and illnesses that followed. Bank party: Out of 90 party attendees, we interviewed a total of 37. Eighteen met the case definition; the median incubation period was 19 hours (range 5 to 23 hours). The median age for these case patients was 37 years (range 17 to 50 years). Illness ranged from 1 to 10 days. Eight women (44.4%) were hospitalized for 1 to 8 days; 6 women (33.3%) had bloody diarrhea, and 2 women (11 %) had stool samples yielding S. enterica group D.

The 26 different dishes were served buffet style from 10 pm until 1 am. Among 14 party attendees who ate tiramisu, 100% developed gastroenteritis compared to 24 % of 17 attendees who did not eat tiramisu (relative risk [RR] = 4.3, 95% confidence interval [CI] = 1.8 €“ 10). No common food was found among the four party attendees who developed gastroenteritis but did not eat tiramisu.

Community Outbreak: Of 12 persons in 7 families who shared food from the restaurant, 9 persons developed gastroenteritis. The median age for the cases was 23 years (range 10 to 38 years); 78% were female, 1 (11%) had been hospitalized, and 6 (67%) of the stool samples yielded S. enterica group D. The median incubation period was 11 hours (range 9 to 44 hours). Among the 9 patients who ate tiramisu, 100% developed gastroenteritis compared to no gastroenteritis among family members who did not eat tiramisu (RR = infinity, P-value 0.007).

Tiramisu preparation assessment: We used a thermometer to measure the internal temperature at various points in the process. The chef mixed the yolks of four raw eggs, pasteurized cream, butter, sugar and cocoa powder. He then layered this liquid dressing with commercial biscuits that had been immersed in warm coffee. The tiramisu was left at room temperature (about 23°C) for 15 minutes to cool. It was then put in the refrigerator at 5°C. The initial internal temperature of the tiramisu was 24°C; after 10 minutes at room temperature it was 22°C, and after 30 minutes in the refrigerator it reached 14°C. The eggs used were locally laid and were stored unrefrigerated at ambient temperature. Restaurant refrigerators were in good condition and operating under appropriate temperature.

On Dec. 27, the chef prepared eight pans of tiramisu equivalent to 80 individual servings (slices). He reserved two pans (20 slices) for sale at the restaurant and six pans (60 slices ) for the bank party. On Dec.28, the restaurant staff loaded all bank party food into a truck at 8:30 pm. The buffet was served from 10 pm to 1 am, approximately 4'/2 hours of time and temperature abuse. Only 90 of 200 invited guests attended the bank party, leaving much leftover food (including tiramisu) Party attendees observed the restaurant employees returning all leftover food to their truck. The following day the restaurant staff returned a leftover tart to the party organizer. From Dec. 28, 1997, to Jan. 1, 1998, 34 slices of tiramisu were sold in the restaurant. Case patients had eaten tiramisu in the restaurant from Dec 29 to Dec 31, but none had eaten tiramisu on the day of the party.

Since 1985 Salmonella serotype enteritidis has been recognized as producing an expanding international pandemic. This emerging infectious organism is capable of infecting the interior (yolk membrane) of intact eggs by transovarian transmission [1]. Freshly laid eggs are normally lightly contaminated; unrefrigerated storage of eggs will lead to multiplication of this organism inside the egg [2]. Salmonella enteritidis phage type 4 (PT4) has increased dramatically in recent years in Saudi Arabia; (it was rarely diagnosed in laying flocks before 1988) and has been isolated from intact eggs locally laid [3,4]. Although serotyping was not available, the Salmonella isolated was in the same group (group D) as the serotype enteritidis.

In this outbreak, the implicated food (tiramisu) was prepared from raw egg yolks and served without cooking. The high attack rate, severe illness, and short incubation suggest a high infective dose of Salmonella in the tiramisu. Three factors probably contributed: eggs were stored at ambient temperature in the restaurant, the egg yolks were blended together for 80 servings, and the tiramisu was left at ambient temperature for more than 4'/2 hours. Salmonella will multiply in eggs and in food at temperatures above 10°C [2]. The lack of illness among restaurant customers on Dec. 28 suggests that the 20 servings for the restaurant prepared together with the 60 servings for the party were not subjected to temperature abuse in the restaurant. We suspect that after these 20 servings were sold, the restaurant began using tiramisu left over from the 60 party servings. These were subsequently responsible for the restaurant cases in customers from Dec. 29 through 31. Because S. enteritidis is likely to contaminate eggs, institutions such as restaurants, hospitals, nursing homes and commercial food processors should use pasteurized eggs and pasteurized egg products [5]. For the family or individual, fresh eggs should be purchased refrigerated and stored refrigerated (<10°C) at all times. Thereafter, eggs and foods made with raw eggs should be well cooked (6).