Watermelon and Salmonellosis Outbreak in a Handicapped Institution, Madinah, July 1998

On July 23, 1998, an outbreak of food poisoning occurred among the residents and workers of a Social Rehabilitation Center (SRC) in ALMadinah City. SRC is a governmental institution where severely mentally retarded and physically handicapped Saudis live. A retrospective cohort study was conducted to identify the source, the extent of the problem, and to suggest recommendations that would prevent occurrence of similar outbreaks in the future.

A case of food poisoning was defined as any person who ate food prepared at the SRC kitchen between July 21-23, 1998 and developed diarrhea, vomiting, or abdominal pain with or without fever between July 23-26, 1998. At the time of the outbreak, 256 Saudi handicapped residents were living at the center. The residents are divided into two sections, one for males, the other for females. A list was obtained of all residents, escorts, and other workers at the SRC in the two sections.

All members of the SRC community were interviewed whether they experienced any of the symptoms of food poisoning or not. Stool specimens and rectal swabs were taken from 27 patients who were admitted to hospitals. The summary discharges of the medical records from the hospital were reviewed to verify the reported symptoms, to determine the period of hospitalization, and to review the results of the laboratory investigations. The kitchens and the dining halls of the SRC were visited. Details on preparation of foods and drinks were collected.

To implicate a specific meal, the Attack Rate percentage (AR) was calculated for nine suspected meals served on the 3 days before the outbreak.

Among 379 persons who usually take meals at the SRC, 125 were identified as cases of food poisoning, Attack Rate (AR) =33%. Of those, 93 cases (74.4%) were admitted to hospitals, 56(44.8%) had diarrhea, vomiting, and fever, 40 (32.0%) had two of the symptoms, and 29 (23.2%) had one of the three symptoms mentioned above and also abdominal pain.

Although all meals were prepared in the same kitchen (male section), the AR in the male section was 84.5%, whereas there were no cases in the female section (AR =0.0 %).

Of all nine meals served during the 3-day period prior to symptom onset, the outbreak was associated with watermelon served at lunch on 23/07/98 (AR =42.8, 95% CI =37.2-48.5). In the male section, 148 persons had some watermelon and 125 (84.5%) had symptoms of food poisoning. Out of 125 cases, 122 (84.1%) had watermelon as a juice and 3 had watermelon slices. Salmonella enterica serogroup enteritidis was isolated from 21 patients. Symptoms started to appear five hours after eating lunch and continued to appear over a 24-hour period after eating the implicated lunch (Figure 1). The lunch was served on July 23 at 1 pm. The median incubation period was 8 hours, the Inter Quartile Range (IQR) =7-9 hours. The duration of the incubation period was associated with amount of the watermelon consumed.

Both the knife and cutting board used to slice the watermelon had been used to cut more than 120 raw chickens served the day before the outbreak. The sliced watermelon was kept at room temperature for at least 2 hours before blending at mid-day.

This outbreak documents an uncommon vehicle for salmonellosis watermelon. There are only a few reported food poisoning outbreaks due to Salmonella spp. implicating watermelon, cantaloupe, or melon [1]. Gayler et al. (1993) reported that Salmonella miami and Salmonella bareilly were responsible for two salmonellosis outbreaks associated with precut wrapped watermelon. It has been shown that interior watermelon tissue could become contaminated if Salmonella was present either on the rind of the watermelon or on the knife used for slicing [2].

It is most likely that these watermelons were contaminated while they were sliced on a dirty cutting board. Being kept at room temperature in an already warm kitchen probably favored overgrowth of bacteria. The blending of watermelon could explain the high AR observed in this study. Blending created a contaminated homogeneous liquid whereas the slices were probably only partially contaminated on their surfaces. This could also explain why those who ate water-melon slices had a relatively longer incubation period.

Having the watermelon sliced on an unclean cutting board demonstrates how inadequate hygienic practices during food handling play a major role in food poisoning outbreaks.

These inadequate practices are repeatedly implicated in literature as important causes of food poisoning outbreak [3]. The dangers are increased in food services that prepare a limited number of food items for large numbers of people. Salmonella can survive for long periods of time on or in foods not commonly implicated in outbreaks of salmonellosis, such as cheese and sliced fresh fruit. Relatively small doses of Salmonella have been shown to cause illness in an outbreak setting [4].

Meat cutting boards should never be used for preparing other food items. Routine inspection of kitchens and improved practices by food handlers is highly recommended.