Outbreak caused by Clostridium perfringens infection and intoxication at a county correctional facility.
On April 16, 2012, at 8:30 a.m., the Communicable Disease/Epidemiology Unit of the Kent County Health Department (KCHD) in Grand Rapids, Michigan, received a telephone call from an employee of the Kent County Correctional Facility (KCCF). The KCCF employee reported that a foodborne illness outbreak was suspected to be taking place at the facility. The caller stated that approximately 30-50 inmates had become ill with vomiting and diarrhea after eating lunch at the facility on April 15, 2012. The estimate of inmates who were ill later increased to 250 out of 1,140 inmates as more information became available. No employees of the correctional facility were known to be ill at that time. The caller indicated that the lunch meal from the prior day was suspected to be the cause because many of the inmates complained about a foul taste and odor associated with it. That meal was served to inmates between 10:30 a.m. and 12:30 p.m. on April 15 and consisted of a chicken taco meat mixture, rice with cheese sauce, refried beans, and a flour tortilla. A garden salad with optional salad dressing, yellow cake, and powdered fruit drink were also offered. According to the caller, many people become ill within one hr after consuming the meal. Other inmates, however, developed symptoms throughout the remainder of that day and into the morning of April 16. Upon receipt of this information, KCHD organized to investigate the suspected outbreak.
Following the report of illnesses on April 16, 2012, KCHD created investigation objectives to investigate the suspected outbreak by gathering appropriate epidemiological and environmental data. The team gathering epidemiological data consisted of epidemiologists, public health nurses, and sanitarians. They developed a questionnaire using the KCCF menu for the 72 hr prior to the onset of the first report of illnesses.
Personal interviews were requested with inmates due to reported low literacy rates within that population. KCHD staff conducted those interviews at the correctional facility on April 17-19. A total of 185 inmates, including ill and well individuals, were interviewed. Questions included: sex, age, food consumed, symptoms experienced, date and onset of those symptoms, duration of illness, and whether medical care was obtained. The case definition was described as any interviewee reporting vomiting and/ or diarrhea. Data analyses were performed using the Epi Info 6 Database Analysis Program from the Centers for Disease Control and Prevention (CDC). This team also identified ill inmates who were willing to provide stool samples for laboratory analyses. Those samples were collected and submitted to the Michigan Department of Community Health laboratory for both bacterial analyses and enterotoxin identification through polymerase chain reaction (PCR).
A second work team was charged with gathering environmental data from the KCCF This team, consisting primarily of sanitarians, assessed the food preparation and service areas, investigated the history of the suspect meals, questioned employees for relevant information, and gathered food samples as appropriate. Correctional facilities are not licensed public food service operations in the State of Michigan; however, the KCHD sanitarians used the Michigan Food Law of 2000 (Public Act 92 of 2000), the 2005 Food and Drug Administration (FDA) Food Code, and generally accepted best food safety practices as guidance for conducting this investigative inspection. Food specimens were analyzed using PCR for enterotoxin source identification and incubated for plate count.
It should be noted that the Kent County Sheriff's Department also investigated the circumstances associated with the outbreak to determine if an act of intentional food adul teration had occurred. The sheriff's department and KCHD worked collaboratively to share valuable information essential to each department's respective investigation.
Of the 185 surveyed individuals who consumed lunch on April 15, 2012, 108 of them were identified as ill according to the case definition. The survey results demonstrated an overall attack rate of 58.4%. It is, however, important to acknowledge that it was not possible to interview all inmates and that sickened inmates may have been more biased toward participating in the survey than their unaffected counterparts. The actual number of sick inmates likely ranged between 250 (KCCF estimate) and 666 (projection calculated by survey attack rate). The profile of the outbreak was representative of the overall KCCF population (Table 1).
Onset of symptoms ranged from April 15 at 11:00 a.m. to April 18 at 8:00 p.m. The period of duration between exposure to the suspect meal and onset of illness ranged from <1 hr to 81 hr, with a mean onset of 9 hr and a median onset of 7 hr. The greatest frequency of illnesses occurred within 1 hr after eating the lunch meal on April 15. As illustrated by the epidemic curve (Figure 1), 29 of the 108 ill interviewees (26.9%) reported an onset of illness within 2.5 hr of when the lunch service began. No employees of KCCF or of the contracted food service company reported illness and none reported consuming the lunch meal on April 15.
Data analyses (Epi Info 6) were utilized to evaluate the 60 food items consumed by the KCCF population during the previous 72 hr according to the menu. Odd ratios (OR) and 95% confidence intervals (CI) were calculated and p < .05 was used as a standard for significance. Interviewees were asked to indicate if they had consumed each of these 60 items. ORs for illness related to each of the food items consumed before April 15 were insignificant. Food items consumed on April 15 demonstrated statistically significant ORs indicating powerful likelihood of relationship (Table 2). The chicken taco meat mixture demonstrated a substantially greater OR than all other menu items: OR = 55.79, 95% CI (19.72, 157.83), p < .001.
During the course of the interviews, KCHD identified a subgroup of work release employees with a unique experience. This subgroup was presented with the same lunch on April 15 as other inmates, but they had heard from other inmates that there was something wrong with the chicken taco meat mixture. The offensive odor of this food item was a common comment from the interviewees. Of the 42 work release employees, only 3 reported eating the chicken taco meat mixture and only 8 (19%) became ill. It should be noted that the chicken taco meat mixture was often physically in contact with other food items on the serving tray. This contact might have transmitted infectious material and/or enterotoxins from one food item to another in the pre-prepared serving tray.
Stool specimens were collected from four ill inmate volunteers on April 16 and from two additional inmate volunteers on April 17. Based upon a recommendation by the Michigan Department of Community Health, specimens were sent to their laboratory and analyzed for Bacillus cereus and C. perfringens. All six specimens were found to be negative for B. cereus and positive for C. perfringens. Confirmatory PCR analyses detected the presence of C. perfringens enterotoxin in all six specimens.
The KCHD environmental health investigation team conducted an initial investigation at the KCCF facility on April 16, 2012, and made several follow-up visits during the subsequent two weeks. The team learned that food service operations at KCCF were contracted to a private company responsible for preparing meals, supervising kitchen trustees (inmates who are assigned to work in the kitchen under supervision), and ensuring food safety. Management staff from that company informed the KCHD environmental health team that the chicken taco meat mixed with sauce was made from a pre-packaged frozen product. According to the kitchen manager, the meat was prepared on Friday, April 13 by cooking it in steam kettles. Another individual, a kitchen trustee, reported that the chicken taco meat mixture was heated on Thursday, April 12, and that gravy leftovers from an earlier meal were added into the chicken taco meat mixture.
While this trustee's claim could not be confirmed, KCCF employees stated that it is not unusual to combine leftovers into new meals in order to conserve resources. The sources agree that the chicken taco meat mixture was brought to a simmer and then placed in large steel pans 4-6 in. deep, temporarily placed on a rack in the freezer for an undisclosed period of time, and then covered in plastic wrap and placed in the walk-in cooler. There was no indication that the temperature of that chicken taco meat mixture was recorded at that time or subsequently monitored until Sunday, April 15 when the food was removed from the cooler and prepared for lunch service by reportedly reheating it to 200[degrees]F (93.3[degrees]C) in steam kettles and then placing it in a hot holding unit. A kitchen trustee stated that the workers in the kitchen noticed that the hot holding unit did not appear to be working properly, so they transferred the chicken taco meat mixture to a pizza oven set at 150 [degrees]F (65.6[degrees]C) for hot holding. The kitchen trustee also reported that the chicken taco meat mixture had "swelled and overflowed" and a strong odor was observed when the pans were being transferred to the pizza oven. The contracted kitchen manager later reported that she checked the temperature of the pizza oven and discovered that it was holding at 90 [degrees]F (32.2[degrees]C). A number of others reported that the chicken taco meat mixture had a very offensive odor and was "bubbly" and "frothy." The food processing and handling histories for the other meal items were investigated and were found to comply with recipe directions and without apparent abuse.
The contracted food service provider was able to provide KCHD investigators with sample meals from the dates in question. As a contractual requirement, they preserved these meals, popularly known as "dead man's trays," in the cooler for several days in order to support foodborne illness investigations. While no photographs of the suspected meal from April 15 were taken, KCHD investigators did photograph a meal from the prior day that demonstrates the general presentation and appearance of meals served at KCCF It should also be noted that the meal items in the large section of the tray (beans and rice) contacted one another in a similar way as was reported from the April 15 meal (chicken taco meat mixture, rice with cheese, and beans). This sort of contact between meal items in the tray enables migration of microorganisms from one item to another.
Specimens of the chicken taco meat mixture, beans, rice, cheese sauce, and tortillas were sent to Michigan Department of Community Health Bureau of Laboratories for analyses. Cultured plate counts for C. perfringens found the rice and cheese mixture to contain 1.5 x [10.sup.7] CFU/g, the beans contained 3.7 x [10.sup.5] CFU/g, and the chicken taco meat mixture contained <10 CFU/g. Confirmatory analyses using PCR determined that the chicken taco meat mixture, rice with cheese sauce, and beans all contained C. perfringens enterotoxin.
The environmental health investigation also identified a number of conditions noncompliant with the 2005 FDA Food Code and best food safety practices. Monitoring and maintaining proper temperature controls in an institution are essential for reducing the risk of enteric outbreaks (Greig, Lee, & Harris, 2011). In addition to significant time and temperature control deficiencies, noncompliant conditions included, but were not limited to faulty equipment, failure to date mark food, inadequate sanitizing process for dishware, poor utensil storage, inadequate hand washing sinks, evidence of pests, and a number of minor maintenance issues. A report consisting of 23 food safety improvement recommendations was issued to KCCF and the contracted food service company as a result of these findings.
Investigators from the Kent County Sheriff's Department interviewed 20 inmates who had been assigned as trustees to work in the kitchen under general supervision from the contracted food manager. The purpose of the Kent County Sheriff's Department investigation was to determine if the food had been criminally adulterated. Their interviews with trustees did reveal information pertinent to the KCHD investigation (presented in the previous sections of this article), but did not find compelling evidence of criminal action.
According to the FDA (2012) and the CDC (2017a), C. perfringens is a spore-forming facultative bacterium located throughout the environment but found primarily in the intestines of humans and many animals. The bacteria are commonly found in raw meat products. Small numbers of the organism often are present after cooking and subsequently multiply to dangerous levels during improper cooling and storage of prepared foods. Meats, meat products, and gravy are the foods most frequently associated with outbreaks caused by C. perfringens. Illness generally is caused when sufficient numbers of the microbe are consumed and subsequently produce toxin in the intestines. The infection usually requires 8-12 hr to incubate before causing diarrhea and abdominal cramping, which subsides in approximately 24 hr. Correctional facilities and similar environments previously have been associated with these outbreaks (CDC, 2009; CDC, 2012).
Approximately 11% of foodborne outbreaks caused by C. perfringens occur in correctional facilities and 92% are related to meat and poultry (Grass, Gould, & Mahon, 2013). Cases of intoxication are rare, in part because the food becomes very offensive to the senses when this level of decay has occurred. Intoxication is typified by a rapid onset of colic and diarrhea (Heymann, 2015). The presence of vomiting (38.6% of cases) in this outbreak is also suggestive of something unusual, such as intoxication, considering that C. perfringens usually only correlates with vomiting in 9% of cases (Bennett, Walsh, & Gould, 2013).
Food and stool specimens confirmed that the outbreak of gastroenteritis at KCCF was caused by C. perfringens infection and/ or intoxication. The epidemiological investigation in this case demonstrated that the chicken taco meat mixture was the most statistically probable exposure causing the illnesses and the environmental investigation found significant abuse of this item. The laboratory analyses, however, suggested that the rice with cheese and/or the beans were the causative exposure (Table 3). CDC (2017b) provides a confirmation guideline of 1 x 105 C. perfringens organisms/g in suspect food items, which supports the case for rice with cheese and/or the beans. Due to the apparent conflict between the laboratory and statistical results, further consideration of the data was required. Two possibilities emerged for the number of C. perfringens numbers in the meat, rice with cheese, and beans.
One hypothesis suggested that-through either sampling error, laboratory error, or uneven distribution of organisms-the chicken taco meat mixture sample that was analyzed for colony count was uniquely underrepresented with viable C. perfringens organisms. The second hypothesis held that the bacteria population within the chicken taco meat mixture had either reached death phase due to gross spoilage, diminishing nutrients, and a changing pH environment, or had been diminished by the final reheating prior to service on April 15 without harming the integrity of the enterotoxin.
Vegetative spores of C. perfringens are inactivated by cooking temperatures of 131 [degrees]F (55[degrees]C) for 16.3 min to 149[degrees]F (65[degrees]C) for 0.9 min (Byrne, Dunne, & Bolton, 2006). C. perfringens enterotoxin is inactivated at 140 [degrees]F (60[degrees]C) for five min (International Commission on Microbiological Specifications for Foods, 2003). The unreliable reheating in the faulty equipment on April 15 possibly could have inactivated vegetative spores, but not the enterotoxin, and left the remaining spores with greatly decayed growth media in the chicken taco meat mixture. As a result, the chicken taco meat mixture environment contained C. perfringens enterotoxin but contained a nearly undetectable number of viable organisms. Under this second hypothesis, the high concentrations of C. perfringens organisms in the cheese/rice mixture and refried beans was caused by contamination from the chicken taco meat mixture when the items contacted one another in the serving tray.
The organisms would have found an acceptable growth media in these newly exposed items and could have multiplied substantially by the time samples of those items were submitted to the laboratory. Meanwhile, the suitability of the chicken taco meat mixture was waning and the population of viable organisms could have decreased to <10 CFU/g when the laboratory received the sample. The observations from inmates and staff regarding a strong foul smell and gas bubbles within the chicken taco meat mixture appear to support this second hypothesis.
This outbreak of foodborne illness caused by C. perfringens exhibited the characteristics of an uncommon intoxication due to the short onset of illness experienced by many of the inmates who ate the food and the testimony of foul odor and "bubbly" chicken taco meat mixture. The occurrence of nausea (67.3%) and vomiting (38.6%) may also suggest toxin ingestion. Outbreaks caused by C. perfringens intoxication may be uncommon, but it is important to recognize that individuals with limited control of their diet options may be more vulnerable. The illnesses of other inmates were more likely caused by infection in the more frequently observed manner. While the data appear to suggest conflicting causative food items, KCHD concluded that the chicken taco meat mixture was the most probable cause. The high odds ratio, history of temperature abuse, possible contamination by external ingredients, and testimony from inmates regarding strong odor and frothy appearance seem consistent with C. perfringens in the taco meat mixture.
The findings from this outbreak response demonstrate that investigators of similar foodborne illness outbreaks should recognize the possible insufficiency of bacterial colony counts from food samples for identifying the causative menu item of a foodborne illness outbreak. A full review of the environment, food history, statistical analyses, and population dynamics should be considered before developing conclusions.
Adam E. London, MPA, RS, DAAS
Julie A. Payne, MPH
Brian Hartl, MPH
Kent County Health Department
Corresponding Author: Adam London, Health Officer, Kent County Health Department, 700 Fuller NE, Grand Rapids, MI 49503. E-mail: firstname.lastname@example.org.
Bennett, S.D., Walsh, K.A., & Gould, L.H. (2013). Foodborne disease outbreaks caused by Bacillus cereus, Clostridium perfringens, and Staphylococcus aureus-United States, 1998-2008. Clinical Infectious Diseases, 57(3), 425-433.
Byrne, B., Dunne, G., & Bolton, D.J. (2006). Thermal inactivation of Bacillus cereus and Clostridium perfringens vegetative cells and spores in pork luncheon roll. Food Microbiology, 23(8), 803-808.
Centers for Disease Control and Prevention. (2009). Clostridium perfringens infection among inmates at a county jail-Wisconsin, August 2008. Morbidity and Mortality Weekly Report, 58(06), 138-141.
Centers for Disease Control and Prevention. (2012). Fatal foodborne Clostridium perfringens illness at a state psychiatric hospital-Louisiana, 2010. Morbidity and Mortality Weekly Report, 61(32), 605-608.
Centers for Disease Control and Prevention. (2017a). Food safety: Clostridium perfringens. Retrieved from http://www.cdc.gov/ foodsafety/diseases/clostridium-perfringens.html
Centers for Disease Control and Prevention. (2017b). Foodborne outbreaks: Guide to confirming an etiology in foodborne disease outbreak. Retrieved from http://www.cdc.gov/foodsafety/outbreaks/ investigating-outbreaks/confirming_diagnosis.html
Food and Drug Administration. (2012). Bad bug book: Handbook of foodborne pathogenic microorganisms and natural toxins (2nd ed.). Silver Spring, MD: Author. Retrieved from https://www.fda.gov/ downloads/Food/FoodborneIllnessContaminants/UCM297627.pdf
Grass, J.E., Gould, L.H., & Mahon, B.E. (2013). Epidemiology of foodborne disease outbreaks caused by Clostridium perfringens, United States, 1998-2010. Foodborne Pathogens and Disease, 10(2), 131-136.
Greig, J.D., Lee, M.B., & Harris, J.E. (2011). Review of enteric outbreaks in prisons: Effective infection control interventions. Public Health, 125(4), 222-228.
Heymann, D.L. (Ed.). (2015). Control of communicable diseases manual (20th ed.). Washington, DC: American Public Health Association Press.
International Commission on Microbiological Specifications for Foods. (2003). Clostridium perfringens. Microorganisms in food: Characteristics of microbial pathogens (1st ed., pp. 112-125). United Kingdom: Kuwer Academic/Plenum.
Caption: Meal served on April 14, 2012, demonstrates contact of items in large serving section. Photo courtesy of Kent County Health Department.
Caption: FIGURE 1 Epidemic Curve
TABLE 1 Profile of Surveyed III Respondents Characteristics # % Gender Female 24 22.22 Male 84 77.78 Total 108 Age (year) 10-19 17 15.74 20-49 79 73.15 50-74 9 8.33 Missing information 3 2.78 Total 108 Symptoms * # % Respondents Nausea 68 67.3 101 Vomiting 39 38.6 101 Abdominal cramps 96 89.7 107 Diarrhea 94 88.7 106 Bloody diarrhea 14 16.3 86 Fever 24 29.3 85 * Onset: range = <1-81 hr, mean = 9.3 hr, and median = 7 hr. Duration: range = 1-60 hr, mean = 19.13 hr, median = 17 hr. TABLE 2 Attack Rates for Foods of Significance Consumed on April 15, 2012 Hot Cereal Breakfast Bakery Milk Sausage Biscuit III Ate 68 66 64 74 Did not eat 33 16 22 26 Total 101 82 86 100 Illness rate (%) 67 80 74 74 Well Ate 32 34 36 31 Did not eat 40 56 50 41 Total 72 90 86 72 Wellness rate (%) 44 38 42 43 Respondents 145 172 172 172 OR 2.58 6.79 4.04 3.76 95% CI 1.4, 4.8 3.4, 13.6 2.1, 7.7 1.9, 7.2 p-value .002 <.001 <.001 <.001 Chicken Cheese Flour Rice Taco Meat Sauce Tortilla Mixture III Ate 100 102 103 101 Did not eat 5 6 5 7 Total 105 108 108 108 Illness rate (%) 95 94 95 94 Well Ate 19 48 53 52 Did not eat 53 25 19 21 Total 72 73 72 73 Wellness rate (%) 26 66 74 71 Respondents 177 181 180 181 OR 55.79 8.85 7.38 5.83 95% CI 19.7, 157.8 3.4, 23.0 2.6, 20.9 2.3, 14.6 p-value <.001 <.001 <.001 <.001 Refried Beans III Ate 89 Did not eat 18 Total 107 Illness rate (%) 83 Well Ate 42 Did not eat 32 Total 74 Wellness rate (%) 57 Respondents 181 OR 3.77 95% CI 1.9, 7.4 p-value <.001 OR = odds ratio; CI = confidence interval. TABLE 3 Evidence Summary for Foods of Significance Food Type Attack OR (95% CI) p-Value Rate (%) Hot cereal 67 2.58 (1.4, 4.8) .002 Breakfast sausage 80 6.79 (3.4, 13.6) <.001 Bakery biscuit 74 4.04 (2.1, 7.7) <.001 Milk 74 3.76 (1.9, 7.2) <.001 Chicken taco 95 55.79 (19.7, 157.8) <.001 meat mixture Cheese sauce 94 8.85 (3.4, 23.0) <.001 Flour tortilla 95 7.38 (2.6, 20.9) <.001 Rice 94 5.82 (2.32, 14.6) <.001 Refried beans 83 3.77 (1.9, 7.4) <.001 Food Type Clostridium Clostridium perfringens perfringens (CFU/g) enterotoxin Hot cereal Breakfast sausage Bakery biscuit Milk Chicken taco <10 Positive meat mixture Cheese sauce 1.5 x [10.sup.7] * Positive * Flour tortilla Rice 1.5 x [10.sup.7] * Positive * Refried beans 3.7 x [10.sup.5] Positive Food Type Environmental Comments Hot cereal Breakfast sausage Bakery biscuit Milk Chicken taco Evidence of time/temperature abuse. Offensive odor meat mixture and "frothy" appearance reported. Served in contact with cheese, rice, and refried beans. Cheese sauce Served in contact with chicken taco meat mixture, rice, and refried beans. Flour tortilla Generally consumed with chicken taco meat mixture and other items. Unlikely C. perfringens media. Rice Served in contact with chicken taco meat mixture, cheese sauce, and refried beans. Refried beans Served in contact with chicken taco meat mixture, cheese sauce, and rice. OR = odds ratio; CI = confidence interval. * Cheese sauce and rice were tested together due to extensive mixing in serving tray.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||ADVANCEMENT OF THE SCIENCE|
|Author:||London, Adam E.; Payne, Julie A.; Hartl, Brian|
|Publication:||Journal of Environmental Health|
|Date:||Jul 1, 2017|
|Previous Article:||All generations need apply.|
|Next Article:||Gastroenteritis associated with rafting the Middle Fork of the Salmon River-Idaho, 2013.|