Lederman, Edith R. MD*; Crum, Nancy F. MD, MPH†; Wallace, Mark R. MD†
*Parasitic Diseases Program, US Naval Medical Research Unit No. 2, Jakarta, Indonesia; †Division of Infectious Diseases, Naval Medical Center San Diego, San Diego, CA.
Address correspondence and reprint requests to Edith R. Lederman, MD, Clinical Investigation Department (KCA), Naval Medical Center San Diego, 34800 Bob Wilson Drive, Ste. 5, San Diego, CA 92134-1005. E-mail: email@example.com; firstname.lastname@example.org.
The views expressed in this article are those of the authors and do not reflect the official policy or position of the Department of the Navy, Department of Defense, or the United State Government. This outbreak was described in part during a poster presentation at the 39th Annual Meeting of the Infectious Diseases Society of America, #872.
Despite the widespread use of antibiotics in animal husbandry and stringent standards for slaughter and meat preparation, food-borne illness from meat products in the United States persists. Pathogens implicated include Salmonella species, Shiga toxin producing Escherichia coli, and Clostridium perfringens. Most consumers are aware that ingestion of incompletely cooked meat will place them at risk for food-borne illness, but few appreciate other aspects of preparation, which decrease the rate of disease transmission including deboning,1 preparing meat in small pieces,2 maintaining sufficient temperature, prompt refrigeration,2 and thorough reheating of stored food.1,3
We report an outbreak of C. perfringens food-borne illness among hospital laboratory employees participating in a Mardi Gras celebration on Fat Tuesday, 2001. This outbreak was associated with inadequate serving temperature and storage of foods served during and after the event.
We performed a case-control study to identify the causative food item(s) and organism responsible for the outbreak of gastrointestinal illness. A temporal relationship between symptom development and a Mardi Gras celebration was noted. The Mardi Gras meal was composed of traditional New Orleans fare, including jambalaya, gumbo, and king cake. The food was home-prepared and served over a period of 1.5 hours. Leftovers were distributed throughout the rest of the day and night shifts, and when consumed, they were either eaten as is or reheated in a microwave. Several hours after the initial servings, employees began to present to supervisors with a syndrome of abdominal pain and diarrhea.
The case definition used was an individual having ingested food from the abovementioned party and developing abdominal cramping and/or loose stools within the subsequent 24 hours. All employees who ingested any portion of the Mardi Gras meal were interviewed in person, and data relating to demographics, medical history, concomitant medications, foods ingested, amount of gumbo eaten (asked to estimate by 0.25 cups), and symptoms were collected using a standard survey tool. Data analysis was performed using EpiInfo 2000. Cases and controls, as well as food items eaten, were compared using the Fisher exact test, and P values and confidence intervals were calculated.
Remains of the food were examined the following morning. The gumbo was noted to "bubble and make noise," prompting a Gram stain and culture. The Gram stain yielded many gram-positive rods, and an anaerobic culture grew a pure culture of C. perfringens with a colony count of greater than 100,000/g.
Fifty-eight laboratory employees and a roommate of an employee had eaten food from the Mardi Gras party. Of the 59 subjects, 30 (50.8%) complained of gastrointestinal and/or constitutional symptoms. Symptoms included abdominal pain (80%), diarrhea (80%), gas/bloating (77%), nausea (33%), chills (7%), and fever (2%). The male-to-female ratio was 1:1, and the age of subjects ranged from 19 to 62 years (mean age of 38.2). Comorbidities included hypertension (n = 7), asthma (non-steroid-dependent, n = 3), gastroesophageal reflux (n = 3), diabetes mellitus (non-insulin-requiring, n = 1), and hyperlipidemia (n = 1). Three subjects were taking antacids; one had eaten the gumbo and developed a gastrointestinal illness subsequent to the party. One patient was taking amoxicillin and also developed the same gastrointestinal symptoms as the other ill subjects.
Twenty-nine of the 39 patients who had eaten any gumbo became ill, yielding an attack rate of 74% (Table 1). Two of the employees who became ill ate only gumbo, lending further evidence to the origin of the illness. Among those who ate the gumbo but did not become ill, all but one ate within the first 2 hours of the meal being served (Fig. 1). Those who ingested gumbo were 51.3 times more likely to develop clinical illness than those not ingesting gumbo (Fisher exact, P < 0.0001). One employee did not recall eating the gumbo but did develop loose stools. There was no correlation between the amount of gumbo ingested and development of illness (the ill ate an average of 0.90 cups, and those without illness ate an average of 0.71 cups). The average time to onset of symptoms was 9 hours, and the average time to resolution of symptoms was 22 hours. Only one employee sought medical attention; 3 employees lost 1 day of work. None were treated with antibiotics or had any sequelae. One stool sample was cultured before this investigation and was negative for routine enteric pathogens; anaerobic cultures of this specimen were not obtained.
This outbreak is classic for C. perfringens food-borne illness in its food association (stews, soups, gumbos), food preparation errors, and clinical presentation. The attack rate of 74% associated with ingestion of the gumbo is also typical of C. perfringens outbreaks.1,4 Furthermore, the attack rate for those who ate within the first 2 hours was 68% compared with 91% for those who ate after the food service had formally ended. This increase correlates with increased bacterial growth; the bacterial proliferation in the gumbo was so significant that gas production could be visualized the following morning. Only one subject's illness could not be linked directly to gumbo ingestion but may be attributed to cross-contamination of serving utensils or an unrelated coincidental gastrointestinal illness. C. perfringens must be considered early on in the evaluation of food-related gastrointestinal illnesses as routine stool cultures will not identify the organism. In addition, public health awareness regarding proper food preparation, serving, and storage should be emphasized, especially in setting of amateur mass food service (eg, company or community gatherings).
C. perfringens is a gram-positive, anaerobic spore, and toxin-forming organism found in soil5 and in the gastrointestinal tracts of healthy farm animals and humans.6,7 C. perfringens may thrive at temperatures which range from 15°C to 50°C, but it favors the mid-40°C's.5 It is the agent of gas gangrene as well as both self-limited food poisoning and necrotizing enteritis (pigbel). It can be classified into 5 toxin types (A to E) based on exotoxin production. There are 13 known toxins (α, β, ε, ι, φ, δ, λ, μ, sialidase, etc), and varying combinations of the first 4 toxins account for the 5 toxin types.8 Type A (only makes α toxin) is responsible for nearly all of food-borne illness. Only 5% of C. perfringens strains have the ability to produce enterotoxin and induce food-borne illness.9 C. perfringens requires 13 amino acids for growth; therefore, it most often proliferates in meat which naturally has a rich supply. Generation time may be as short as 8 minutes under optimal conditions.10
C. perfringens is among the top 5 most common bacterial causes of food-borne illness in the United States and may account for up to 10% to 20% of outbreaks.11 The overall incidence of C. perfringens food poisoning is declining, probably through improved food preparation practices.11,12 C. perfringens-induced gastroenteritis is usually associated with meat and meat products and is most often linked to food service establishments (cafeterias, restaurants, schools).4 Holiday meal-related cases are common because of the large quantities of precooked food that must be prepared.3 Promotion of spore formation begins in meat that is not evenly cooked (large pieces, cooked on bones, etc.) and later allowed to sit at ambient temperature. The spore-containing food is then ingested, and gastric acidity promotes germination of the spores and subsequent toxin production. The most commonly implicated food include stews, soups, and gumbos that often include meat still on the bone. In this outbreak, food servers admitted using whole chicken parts in the gumbo, did not maintain adequate serving temperatures, and left food at ambient temperature for extended periods, especially once the party had adjourned.
The C. perfringens type A produces an enterotoxin that inserts itself into intestinal epithelial cells leading to cell damage and electrolyte loss yielding the symptoms of abdominal pain and diarrhea.8 The time it takes for symptoms to develop depends upon the bacterial burden of the food and the amount ingested; the onset ranges from 8 to 24 hours (mean is 12 hours). The symptom complex is composed of nausea/vomiting and severe abdominal pain, followed by nonbloody diarrhea, sometimes accompanied by myalgias and chills; fever is rare.5
The diagnosis of C. perfringens food-borne illness may be confirmed if (1) the clinical syndrome corresponds to the above description and (2) there is microbiologic evidence, such as anaerobic stool culture (>1,000,000 cfu/g) or food culture (>100,000 cfu/g), or if the same serotype is isolated from stool and food cultures.13 Often, samples are not available for culture, or specimens are not handled properly. If the microbiologic workup does not yield an organism or if the workup is in question, the utilization of clinico-epidemiologic profiles that are pathogen-specific may be useful.14 Confirmation of cpe+ strains (the gene that encodes for enterotoxin) lends further evidence. Likewise, strain typing may be used to link cases in an outbreak when cultures are negative.7 Routine C. perfringens food-borne illness is self-limited, and only supportive measures such as fluid resuscitation are necessary.
Type A of C. perfringens resides in the intestines of clinically normal animals. Therefore, the FDA has instituted the process of evisceration before harvest of the meat.6 Carcasses are also visually examined for fecal material and, if noted, are cleaned/trimmed appropriately. Raw food contains small numbers of C. perfringens, but the heating of food selects for spore production, and while the food is cooling, germination begins, greatly amplifying the bacterial load. "Kitchen strains" which undergo repeated episodes of heat shock (75°C, 20 minutes) are likely to sporulate and later produce toxin under the appropriate conditions.10 The toxin of C. perfringens can be destroyed by heating food to 60°C for 10 minutes.5 Recommendations for prevention of food-borne illness caused by C. perfringens are to maintain cooked food at 60°C or higher or to cool food down to 10°C within 2 to 3 hours of cooking. Reheating should achieve temperatures of 75°C or higher (this will destroy vegetative bacteria).5,6 Parsing meat into smaller pieces, as well as deboning the meat, will also facilitate quicker heating and cooling will ensure a more even temperature distribution. Eradication of the spores would require heating at temperatures of 100°C. Bleach solution with a pH of >8.0 can kill spores, and UV radiation may reduce spore-forming bacteria.15
The above report demonstrates the necessity for public awareness regarding proper food preparation and storage. As a result of this investigation, an in-service on proper food preparation and storage was conducted for all laboratory personnel in an attempt to heighten awareness and, in turn, prevent further outbreaks of food-borne illness.
1. Parikh AI, Jay MT, Kassam D, et al. Clostridium perfringens outbreak at a juvenile detention facility linked to a Thanksgiving holiday meal. West J Med. 1997;166:417-419.
2. Regan CM, Syed Q, Tunstall PJ. A hospital outbreak of Clostridium perfringens food poisoning-implications for food hygiene review in hospitals. J Hosp Infect. 1995;29:69-73.
3. Centers for Disease Control and Prevention Clostridium perfringens gastroenteritis associated with corned beef served at St. Patrick's Day meals-Ohio and Virginia. Morb Mortal Wkly Rep. 1994;c43:143-144.
4. Hook D, Jalaludin B, Fitzsimmons G. Clostridium perfringens food-borne outbreak: an epidemiological investigation. Aust N Z J Public Health. 1996;20:119-122.
5. Lund B. Foodborne illness: foodborne disease due to Bacillus and Clostridium species. Lancet. 1990;10:982-986.
6. Hollingsworth J, Kaplan B. Zero tolerance for visible feces helps FSIS fight foodborne pathogens. J Am Vet Med Assoc. 1997;211:534-535.
7. Maslanka SE, Kerr JG, Williams G, et al. Molecular subtyping of Clostridium perfringens by pulsed-field gel electrophoresis to facilitate food-borne-disease outbreak investigations. J Clin Microbiol. 1999;37:2209-2214.
8. Rood J. Virulence genes of Clostridium perfringens. Annu Rev Microbiol. 1998;52:333-360.
9. McClane BA. New insights into the genetics and regulation of expression of Clostridium perfringens enterotoxin. Curr Top Microbiol Immunol. 1998;225:37.
10. Andersson A, Rönner U, Granum PE. What problems does the food industry have with the spore-forming pathogens Bacillus cereus and Clostridium perfringens? Int J Food Microbiol. 1995;28:145-155.
11. Centers for Disease Control and Prevention Surveillance for foodborne-disease outbreaks-United States, 1993-1997. Morb Mortal Wkly Rep. 2000;49:11-16.
12. Hedberg CW, MacDonald KL, Osterholm MTO. Changing epidemiology of food-borne disease: a Minnesota perspective. Clin Infect Dis. 1994;18:671.
13. Hauschild AHW. Criteria and procedures for implicating Clostridium perfringens in food-borne outbreaks. Can J Public Health. 1975;66:388-392.
14. Hall JA, Goulding JS, Bean NH, et al. Epidemiologic profiling: evaluating foodborne outbreaks for which no pathogen as isolated by routine laboratory testing: United States, 1982-9. Epidemiol Infect. 2001;127:381-387.
15. Brown KL. Control of bacterial spores. Br Med Bull. 2000;56:158-171.
© 2004 Lippincott Williams & Wilkins, Inc.