Staphylococcal food poisoning from a fundraiser.
On August 5, 1994, the New Mexico Department of Health (NMDOH) was notified that six persons had come that evening to the emergency room in Artesia, New Mexico, complaining of gastrointestinal symptoms including vomiting, diarrhea, and abdominal cramps. All had eaten food from a local fundraiser at approximately noon. The fundraiser had served food from 10:00 a.m. to 6:00 p.m. at a local church and had delivered plates of food to businesses and personal residences, as well. The short incubation period suggested staphylococcal or Bacillus cereus food poisoning. NMDOH investigated the outbreak to determine its cause and magnitude.
The authors conducted a cohort study among persons who ate fundraiser food to determine the number of persons who became ill and what foods and factors increased the risk of illness. The organizer of the fundraiser provided lists of homes and businesses that had received plates of foods, as well as of persons who had eaten the food at the church. The authors interviewed as many of these persons as possible by phone, using a common questionnaire that elicited demographic information, illness symptomatology, and foods from the fundraiser eaten. A case was defined by the symptoms of nausea, vomiting, abdominal cramps, and/or diarrhea within 24 hours of eating food from the fundraiser.
Food samples from the fundraiser and one stool sample collected from an ill person on August 8 were tested for Staphylococcal aureus and B. cereus contamination at NMDOH's Scientific Laboratory Division (SLD). Samples of foods that had high counts of S. aureus ([greater than] [10.sup.5] organisms/gram) and their isolates were sent to the U.S. Food and Drug Administration (FDA). Food samples were tested for the presence of preformed staphylococcal enterotoxins with an enzyme-linked immunosorbent assay (1). FDA also treated the food samples with urea and zinc acetate, which enabled the detection of preformed enterotoxin that had been altered by heat but was still biologically active (2,3).
The authors interviewed the organizer of the fundraiser and all food handlers to determine how the food was prepared, handled, and served and whether the food handlers had any cuts or sores on their hands. No specimens for culture were taken from food handlers, and the food preparation sites, which were in private homes, were not inspected.
TABLE 1 Frequency of Illness by Age Age III Total Attack Relative Risk of Illness(a,b) (years) Rate (%) (95% confidence interval) 18 2 28 7.1 referent 19-34 9 34 26.5 3.7 (0.9, 15.8) 35-59 6 35 17.1 2.4 (0.5, 11.0) 60 8 13 61.5 8.6 (2.1, 35.0) (a) The risk of becoming ill relative to the reference age group (age 18 years) is calculated by dividing the attack rate of the age group of interest by the attack rate of the reference age group. A relative risk greater than one indicates an increased risk of illness compared to the reference group. (b) Chi square test for linear trend = 9.942, p-value = 0.00162.
[TABULAR DATA OMITTED]
The authors completed questionnaires for 110 of the approximately 150 persons who had eaten food from the fundraiser. Twenty-five persons met the case definition, constituting an overall attack rate (i.e., the proportion of persons who became ill) of 23 percent.
The distribution of cases by date and time at which onset of symptoms occurred is presented in Figure 1. Symptom onset time ranged from 1:00 p.m. on August 5, the day of the fundraiser, to 3:00 p.m. on August 6: most persons began to experience symptoms during the afternoon and evening of August 5.
The median age of the 25 case patients was 41.0 years (range: 8-77 years), and 10 (40.0 percent) of these case patients were male. The median incubation period was 4.0 hours (range: 1.2-21.0 hours). Of the 25 case patients, 22 (88 percent) experienced nausea, 21 (84 percent) had diarrhea, 17 (68 percent) had abdominal cramps, 13 (52 percent) experienced vomiting, and three (12 percent) had fever. No one reported bloody stools. The median duration of illness was 1.0 day (range: 0.5-4 days). Six persons (24.0 percent) sought care at the emergency room, and one person consulted a private physician because of illness.
Food from the fundraiser was eaten throughout the day (10:00 a.m. to 8:00 p.m.) on August 5, but only persons who ate food between 11:00 a.m. and 1:59 p.m. (attack rate: 31 percent) and between 5:00 p.m. and 7:59 p.m. (attack rate: 15 percent) became ill. Attack rates are presented by age in Table 1. Very few children became ill, and the attack rate for children less than 18 years of age was only seven percent. The highest attack rate (62 percent) occurred in persons 60 years of age or older; lower rates were reported for persons aged 19 to 34 years (27 percent) and aged 35 to 59 years (17 percent) (p = 0.002, Chi square test for linear trend). Attack rates did not differ substantially by sex; the attack rate in women was 24.2 percent (15 of 62), compared with 20.8 percent (10 of 48) in men.
Food-specific attack rates and relative risks are presented in Table 2. The relative risk of becoming ill after eating a food item is calculated by dividing the attack rate of persons who ate the food item by the attack rate of persons who did not eat the food item. Relative risks greater than i indicate an increased risk of illness. No food items other than beef tacos were associated with an increased risk of illness. Although beef tacos had only a slightly higher attack rate (23.4 percent) than other food items, all persons who became ill had eaten the tacos, and the one person surveyed who had not eaten the tacos did not become ill. The relative risk for eating beef tacos could not be calculated because of the zero value in the denominator (i.e., 23.4/0.0 = undefined).
Neither S. aureus nor B. cereus was isolated in the single stool sample that was obtained and tested. Food samples provided by the organizer of the fundraiser were refrigerated before being shipped, but had a temperature of 73 [degrees] F upon arrival at SLD. Small amounts of B. cereus organisms ([less than] 5.5 x [10.sup.4] organisms/gram) were detected in taco samples. High amounts of S. aureus (2.8 x [10.sup.6] organisms/gram) were detected in taco and spanish rice samples. A separate portion of taco meat, which had been held in reserve in a refrigerator during the fundraiser, contained no S. aureus. FDA found all staphylococcal isolates to be enterotoxigenic. Preformed staphylococcal enterotoxin was detected both in taco samples that were treated with urea and zinc acetate and in those not treated, but was not detected in the spanish rice sample.
The food for the fundraiser was prepared by four women in their respective homes. Plates of food were then assembled at the church. The taco meat and spanish rice were prepared in two batches by different women. The authors identified improper food-handling practices for one batch of taco meat and one batch of spanish rice. The organizer of the fundraiser cooked one batch of taco meat the night before the fundraiser and left the taco meat uncovered at room temperature for four hours before refrigerating it. The next morning the organizer took the cooked meat to the church where it was reheated and kept warm in an electric roaster. This batch of meat Was served first. Another woman prepared a batch of rice that she left at room temperature for four hours between frying it at 2:00 a.m. and putting it into the oven to bake at 6:00 a.m. on the day of the fundraiser. The rice was transferred to crock pots at the church and kept warm at a low temperature setting. It is not known whether this batch of rice was served first. A review of food-handling procedures revealed no other problems. Although gloves were not used, all of the food preparers and handlers denied having cuts or sores on their hands or any illnesses at the time of the fundraiser.
The menu for the fundraiser was not extensive, and most persons ate a little of each item. This made it difficult to determine, by the use of attack rates and relative risk, which food items may have been contaminated. All case patients reported eating beef tacos, and the one person who did not eat any tacos did not become ill, although she had eaten spanish rice and beans. In addition, attack rates varied by the time of day the food was eaten. This is what one would expect to see if only some of the taco meat was contaminated; persons would have to eat taco meat to become ill, but many persons eating the tacos would not become ill.
Food samples collected from the fundraiser were at room temperature when they arrived at SLD. Therefore, the laboratory results need to be interpreted with caution, although they still provide useful information. The symptoms and incubation periods observed for this outbreak were compatible with either B. cereus. or S. aureus food poisoning. All of the foods tested, however, contained either no or only low levels of B. cereus. Even if some organisms. died during. shipping, it is unlikely that any of the food samples had contained levels of B. cereus high enough to cause illness ([greater than] 500,000 organisms/gram) at the time of the fundraiser (4).
High levels of enterotoxigenic S. aureus organisms were detected in taco samples. The number of organisms detected per gram of food may have been higher or lower than the number present when the food was served; the bacteria could have either multiplied or died during shipping. However, the very high levels detected in two taco samples ([greater than] [10.sup.6] organisms/ gram), suggest that the organisms were present in some of the meat in large enough numbers to make people ill at the time it was served. In addition, heat-altered staphylococcal enterotoxin was detected in portions of a taco sample treated with urea and zinc acetate, indicating that preformed enterotoxin in the taco sample had been heated. This result suggests that preformed enterotoxin was present in at least some of the taco meat when it was reheated at the church. A sample from a portion of the second batch of taco meat, which had been held in reserve in the refrigerator and had never been reheated, did not contain S. aureus. This indicates that not all of the taco meat from the fundraiser was initially contaminated with S. aureus. The spanish rice sample did not contain preformed toxin and probably became contaminated with S. aureus as a result of being shipped in the same container when shipped from homes to church and being mixed with taco meat.
The stool sample taken from one of the case patients did not contain S. aureus or B. cereus. The sample was obtained approximately three days after the fundraiser, however, and the patient would probably no longer have been shedding the organisms in her stools.
Taken together, the epidemiologic, bacteriologic, and environmental evidence suggested that at least some of the taco meat was contaminated with S. aureus in sufficient amounts to cause illness after eating it, and that this was the cause of the outbreak. One batch of taco meat was left at room temperature for four hours (after cooking and before refrigeration) on the morning of the fundraiser. This four-hour period provided ample time for S. aureus to multiply and produce enterotoxin. The enterotoxin is heat resistant, and it would therefore not have been inactivated when the meat was warmed up. Previously cooked, proteinaceous food is a common vehicle for S. aureus food poisoning (5,6,7).
It is not known how S. aureus was introduced into the taco meat. None of the persons preparing or serving the food reported having cuts or sores on their hands, and no samples were taken from their hands, throats, or noses. However, staphylococci are ubiquitous, and approximately 30 percent to 50 percent of healthy human beings carry staphylococci at any one time (5). A cough, sneeze, or touch from an infected person would have been sufficient to inoculate the meat.
Differentiating the attack rates according to the times at which food was eaten revealed an interesting pattern. The attack rate was 31 percent for food eaten between 11:00 a.m. and 1:59 p.m.; it was 15 percent for food eaten between 5:00 p.m. and 7:59 p.m. No one who ate fundraiser food between 2:00 p.m. and 4:59 p.m. became ill, however. The batch of taco meat that was left out for several hours at room temperature was the first batch to be served. If this batch was contaminated with S. aureus, that would explain the positive attack rates for food eaten between 11:00 a.m. and 1:59 p.m. At some time during the day, the first batch of taco meat was finished and servers began serving the second batch, which was initially uncontaminated. This could explain why food eaten between 2:00 p.m. and 4:59 p.m. did not cause any illness. The second batch of taco meat was placed in the same electric roaster that was used for the first batch, and the same spoon was used to serve all of the meat. Therefore, it is possible that the second batch of taco meat became contaminated when placed in the roaster. It would have taken several hours for the bacteria to multiply and produce enough enterotoxin to make people ill, and this might explain why fundraiser food did not cause illness again until after 5:00 p.m.
An interesting finding was that attack rates were low for children 18 years of age or younger and were 8.6 times higher for persons 60 years of age or older. This was not just because a larger number of older persons happened to eat at the times contaminated food was served. Looking only at the times when ill persons ate (11:00 a.m. to 1:59 p.m. and 5:00 to 7:00 p.m.), which were presumably the times when contaminated meat was served, persons 60 years of age or older still had an attack rate that was 7.0 times higher than the attack rate for children. It is unclear whether a similar pattern has been observed in other staphylococcal food poisoning outbreaks, because published reports generally do not give age-specific attack rates. To study this question further, the authors reanalyzed data from an outbreak of staphylococcal food poisoning associated with contaminated turkey that occurred in New Mexico in 1986 (8). Sixty-seven cases had been identified among 162 persons interviewed (attack rate: 41 percent). In agreement with the current outbreak, persons 60 years of age or older had the highest attack rate (54.8 percent) of any of the age groups. Statistically, however, this attack rate was not significantly different from the attack rate (42.2 percent) for persons 18 years of age or younger.
This investigation indicates that an outbreak of at least 25 staphylococcal intoxications was caused by food from a fundraiser. S. aureus organisms apparently contaminated taco meat during food preparation, and violations of recommended guidelines for time and temperature in holding food items allowed the organisms to multiply sufficiently to cause illness. A previous study of 766 foodborne outbreaks in the United States from 1977 to 1982 identified the most frequent contributory factors (9). A number of those factors were identified in this outbreak. A lapse of 12 or more hours between preparation and eating, which in this case occurred for several batches of food, accounted for 25 percent of the 1977-1982 outbreaks. Holding food at ambient (or room) temperature, which also occurred in this outbreak, accounted for 21 percent of the outbreaks. There was also an indication in the present investigation that inadequate reheating may have occurred, which accounted for nine percent of the 1977-1982 outbreaks, as well as improper hot-holding temperatures, which accounted for an additional nine percent of those outbreaks.
The type of charity fundraiser food event that caused the 1994 outbreak is not regulated by the New Mexico Environment Department (NMED). Thus, no permits were required for this function, and the food handlers received no training or instruction from NMED. This situation is typical of regulations and practices in most states, although Wisconsin has regulations covering charity events if they occur regularly (i.e., four or more times per year) (10). In Alabama, although fairs and local festivals are inspected, as are other licensed food vendors, single charity events are not (11). Iowa requires permits for mass gatherings of more than 1,000 persons and for events run by for-profit organizations (12). Virginia exempts charity organizations but offers a short course titled "The Occasional Cook," in which enrollments of representatives from churches and civic groups - both those required to meet the regulations and those exempt - have been impressive (13).
The outbreak of staphylococcal intoxication at the fundraiser discussed in this paper indicates the risk of foodborne disease outbreaks posed by this type of event. If state regulatory agencies cannot include such events in their license and inspection requirements, then at the very least, Virginia's approach should be considered, and short courses that educate potential food preparers should be offered. Problems with time and temperature in the holding of food are the most frequent cause of foodborne outbreaks in the United States, and increased education is needed to prevent them. The 1994 outbreak in New Mexico indicates that food prepared for charity fundraisers is vulnerable to time and temperature problems. It is tragic for these fundraisers to be associated with outbreaks when their purpose is usually to help people.
Corresponding Author: Rebecca K. Olson, Ph.D, California Medical Review, Inc., 60 Spear St., Suite 400, San Francisco, CA 94105. E-mail: email@example.com.
The authors thank the following people for their assistance in investigating this outbreak: J. Smith, New Mexico Environment Department; E. Acosta, K. Reyes, and A. Rodriguez, Eddy County Health Office; P. Gutierrez, J. Miller, and L. Nims, Scientific Laboratory Division; R. Voorhees, Division of Epidemiology, Evaluation, and Planning, New Mexico Department of Health; and R.W. Bennett, U.S. Food and Drug Administration. The authors also thank L. Fehrs, of the Centers for Disease Control and Prevention, for reviewing the manuscript.
1. Bennett, R.W., and F. McClure (1994), "Visual Screening With Enzyme Immunoassay for Staphylococcal Enterotoxins in Foods: Collaboratively Study," Journal of the Association of Official Analytical Chemists Int., 77(2):357-64.
2. Bennett, R.W. (1992), "The Biomolecular Temperament of Staphylococcal Enterotoxin in Thermally Processed Foods," Journal of the Association of Official Analytical Chemists Int., 75(1):6-12.
3. Bennett, R.W. (1994), "Urea Renaturation and Identification of Staphylococcal Enterotoxin," in R.C. Spencer, S.W.B. Newsom, and E.P. Wright, eds., Rapid Methods and Automation in Microbiology and Immunology, Andover: Intercept Ltd., pp. 401-411.
4. Benenson, A.S., ed. (1995), Control of Communicable Diseases Manual, Washington, D.C.: American Public Health Association, pp. 184-189.
5. Bergdoll, M.S. (1992), "Staphylococcal Intoxication in Mass Feeding," in A.T. Tu, ed., Food Poisoning: Handbook of Natural Toxins, New York: Marcel Dekker, Inc., pp. 25-47.
6. Holmberg, S.D., and P.A. Blake (1984), "Staphylococcal Food Poisoning in the United States," JAMA, 251(4):487-489.
7. Tranter, H.S. (1990), "Foodborne Illness: Foodborne Staphylococcal Illness," Lancet, 336:1044-1046.
8. Eidson, M., G.N. graves, P. Gutierrez, J.T. McLauchlin, and L. Nims (1990), "Food Poisoning Investigation: Illness Blames on Turkey at New Mexico Buffet," J. Envtl. Health, 52(5):295-297.
9. Bryan, F.L. (1988), "Risk of Practices, Procedures and Processes That Lead to Outbreaks of Food Borne Diseases," Journal of Food Protection, 51(8):663-673.
10. Kaczmierczak, J., Wisconsin Department of Health and Human Resources (July 1996), personal communication.
11. Johnston, W., Alabama Department of Public Health (July 1996), personal communication.
12. Currier, R., Iowa Department of Public Health (July 1996), personal communication.
13. Jenkins, S., and R. Hicks, Virginia State Health Department (July 1996), personal communication.
|Printer friendly Cite/link Email Feedback|
|Author:||Sewell, C. Mack|
|Publication:||Journal of Environmental Health|
|Date:||Oct 1, 1997|
|Previous Article:||Continuing patterns of metallic-mercury exposure - alert.|
|Next Article:||Reducing exposure to electromagnetic fields: the effects of low- and high-threat risk messages on behavior change.|