The Foodborne Outbreak That Wasn’t
Indiana Epidemiology Newsletter
Field Epidemiologist District 2
On May 16, 2006, a local health department (LHD) alerted the Indiana State Department of Health (ISDH) of a possible gastroenteritis outbreak at an international conference held from May 9-13. Approximately 300 persons from 23 states and Mexico attended the conference. The predominant symptoms included vomiting and diarrhea. The conference was held at two adjoining facilities. Eight different meals were prepared and served at the two adjoining facilities and one local university.
A complete list of attendees, vendors, speakers, and conference activities was requested from the conference organizer. Menus for seven meals served were obtained from the two adjoining facilities. The food item choices (buffet-style meal) consumed at the off-site location were solicited from interviewees. Notification was sent via the Centers for Disease Control and Prevention (CDC) Epi-X electronic exchange and the National Foodborne Illness Listserve to all involved states.
The LHD and the ISDH developed a five-page foodborne gastroenteritis questionnaire to determine onset of illness, symptom history, and food consumption. The LHD conducted interviews within its jurisdiction. The ISDH conducted the remaining in-state interviews as well as interviews on behalf of two other states, and the CDC offered to serve as the contact for the Mexico residents.
An unmatched case control study was conducted, with 159 total interviews completed. A case was defined as any previously healthy person who attended the conference on May 9-13, 2006, or who was a contact of a conference attendee and developed diarrhea or vomiting after May 9. Sixty cases and 99 controls were identified. Symptom onset dates ranged from May 10-18 (Figure 1). The median duration of illness was 59 hours. In addition to diarrhea (95.0%) and vomiting (40.0%), other reported symptoms included nausea (66.7%), fatigue (61.7%), headache (58.3%), cramps (55.0%), chills (35.0%), body aches (33.3%), and low-grade fever (28.3%). Four were hospitalized or visited the emergency department.
The following events were analyzed to evaluate the significance of association between illness and specific conference events: 1) contact with ill people, 2) eating at other restaurants, and
3) eating any of the eight conference meals. Table 1 indicates the statistical association of each event with the likelihood of developing illness.
Figure 1: Epidemic Curve
|Contact with Ill People||1.954 (0.873 – 4.373)||0.103|
|Ate at Restaurants||0.725 (0.325 – 1.617)||0.4323|
|May 9 Dinner||0.600 (0.213 – 1.693)||0.3347|
|May 10 Dinner||2.169 (0.660 – 7.135)||0.2023|
|May 11 Breakfast||1.587 (0.551 – 4.569)||0.3918|
|May 11 Lunch||0.237 (0.068 – 0.825)||0.0237|
|May 11 Dinner||2.071 (0.458 – 9.353)||0.3442|
|May 12 Breakfast||1.689 (0.586 – 4.871)||0.3321|
|May 12 Lunch||2.305 (0.721 – 7.368)||0.159|
|May 12 Dinner||0.286 (0.100 – 0.818)||0.0196|
Environmental health specialists from the LHD visited the adjoining facilities to conduct inspections, including Hazard Analysis of Critical Control Points (HACCP), and to inquire about employee illness. The university health center reported no spike in gastrointestinal illnesses in their students, and the meal for the students included the buffet items. Therefore, the university dining hall and kitchen were not inspected. No leftovers from either facility were available for testing, but two common food preparation items were sent to the ISDH Laboratory for testing. However, since no meal was significantly associated with illness, the items were not tested.
The initial inspection of the two adjoining facilities revealed that the dishes were not properly sanitized, as the dishwashers were not heating to the required temperature. Attendees also reported water leaking on and near the food buffet and dinner tables at one of the facilities. The findings from the food inspections were determined not to be conclusive factors in the outbreak investigation. No food handlers employed by the facilities or the temporary agency utilized during the conference reported any illness.
Ten stool samples were collected and analyzed by various state laboratories for bacterial (E. coli 0157:H7, Campylobacter, Shigella, and Salmonella) and viral (Norovirus) pathogens (Table 1). Four samples tested positive for Norovirus at the ISDH Laboratory, with one of these also testing positive for Shigella. The Illinois state laboratory reported one specimen positive for Norovirus. The South Carolina state laboratory analyzed two stool specimens, with one positive for Clostridium perfringens. The Florida state laboratory analyzed two stool samples, both positive for Norovirus. All other results were negative.
The investigation confirmed an outbreak of illness among attendees at the conference. The only common exposure among the cases was attendance at the conference or their subsequent contacts.
The causative agent of this outbreak was Norovirus. Seven attendees tested positive. In addition, the 53 other attendees who were identified during the same time period had symptoms compatible with Norovirus infection. The median duration of illness was 59 hours; the wide range of illness duration could be the result of interpretation of symptoms and/or illness. One attendee reported an onset date of May 9 without an onset time, according to the questionnaire. This case was ruled out as the index case, because the only symptom reported on May 9 was nausea; case descript symptoms did not start until May 10.
Although two specimens tested positive for two bacterial agents, these were ruled out as causative agents of this outbreak. The one positive Shigella sonnei culture was identified from a contact case who attends elementary school and after-school care and, therefore, may be unrelated to this outbreak. The incubation period of the one attendee positive for Clostridium perfringens was not consistent with ingestion of a meal during the conference to the onset of symptoms; in addition, the reported symptoms were atypical of C. perfringens.
This outbreak was most likely a result of person-to-person transmission rather than foodborne. None of the conference events was statistically associated with illness, including contact with ill people, eating at other restaurants, and eating any of the eight conference meals. In addition, the epidemiologic curve most likely indicates person-to-person transmission (Figure 1). Point source outbreaks, including foodborne, have a sharp, abrupt upslope following exposure and a gradual down slope. Person-to-person outbreaks have a more gradual upslope reaching a peak, followed by a gradual down slope, as illustrated by this epidemic curve.
Although one staff member (non-food handler) was symptomatic and tested positive for Norovirus, the onset of illness for this case occurred at the peak of the outbreak (May 12). Therefore, this employee was not implicated as a source of the outbreak. Although the findings from the environmental inspections warranted corrective actions, they were not contributing factors to this outbreak. Upon further inspections, the establishments associated with this investigation became compliant with all violations.
Most Norovirus outbreaks can be prevented by adhering to the following the guidelines:
- Thoroughly wash hands with soap and water before preparing and serving food, after using the restroom, or after assisting anyone with diarrhea and/or vomiting.
- Children with diarrhea and/or vomiting should be excluded from day-care centers and schools.
- Persons with diarrhea and/or vomiting should not prepare food for others and should avoid close contact with other people while symptoms persist.
- Disinfect contaminated surfaces with a 1:10 solution of household bleach.
ISDH November 13, 2004. Retail Food Establishments Sanitation Requirements.
Title 410 IAC 7-24-122)
More information on norovirus can be found at: