Gastroenteritis at a University in Texas EPI Ready Case Analysis

INSTRUCTOR’S VERSION

1Centers for Disease Control and Prevention, 2Texas Department of Health, 3Baylor College of Medicine, 4Atlanta Veterans Administration Medical Center, 5City of Huntsville, Health Inspections, 6Texas Department of Health, Region 6/5S

NOTE:  This case study is based on a real-life outbreak investigation undertaken in Texas in 1998.  Some aspects of the original outbreak and investigation have been altered, however, to assist in meeting the desired teaching objectives and allow completion of the case study in less than 1.5 hours.   

Students should be aware that this case study describes and promotes one particular approach to foodborne disease outbreak investigation.  Procedures and policies in outbreak investigations, however, can vary from country to country, state to state, and outbreak to outbreak.

It is anticipated that the epidemiologist investigating a foodborne disease outbreak will work within the framework of an “investigation team” which includes persons with expertise in epidemiology, microbiology, sanitation, food science, and environmental health.  It is through the collaborative efforts of this team, with each member playing a critical role, that outbreak investigations are successfully completed.

Please send us your comments on this case study by visiting our website at https://www.cdc.gov/epicasestudies.  Please include the name of the case study with your comments.

U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES

Public Health Service

Centers for Disease Control and Prevention

Atlanta, Georgia 30333

INSTRUCTOR’S VERSION

Gastroenteritis at a University in Texas

Learning objectives: After completing this case study, the student should be able to: List categories and examples of questions that should be asked of complainants who report a possible foodborne illness to the health department. Prioritize the investigation of a suspected foodborne illness or foodborne disease outbreak. List clues that might help determine the causative agent in an outbreak of acute gastrointestinal illness. Use the descriptive epidemiology among cases, information about the setting in which the outbreak has occurred, and the results of hypothesis-generating interviews to develop hypotheses about the source of an outbreak. Discuss some of the issues associated with the collection of clinical specimens for the investigation of a suspected foodborne disease outbreak. List key areas of focus in investigating a facility implicated in a foodborne disease outbreak including interviews with food workers, observation of kitchen practices, and collection of samples. Interpret the measure of association for a case-control study.

PART I – OUTBREAK DETECTION

On the morning of March 11, the Texas Department of Health (TDH) in Austin received a telephone call from a student at a university in south-central Texas.  The student reported that he and his roommate, a fraternity brother, were suffering from nausea, vomiting, and diarrhea.  Both had become ill during the night.  The roommate had taken an over-the-counter medication with some relief of his symptoms.  Neither the student nor his roommate had seen a physician or gone to the emergency room.

The students believed their illness was due to food they had eaten at a local pizzeria the previous night.  They asked if they should attend classes and take a biology midterm exam that was scheduled that afternoon.

Question 1:  What questions (or types of questions) would you ask the student?

In recording a complaint about a possible foodborne illness, it is important to systematically collect the following information:

• WHAT is the person’s problem? (e.g., clinical description of the illness, whether a physician was consulted, whether any tests were performed or any treatments were provided)
• WHO else became ill, their characteristics (e.g., age, sex, occupation), and the nature of their illnesses (e.g., symptoms, whether any persons were hospitalized or died)?
• WHEN did the affected person(s) become ill?
• WHERE are the affected persons located? (including names and telephone numbers)
• WHY (and HOW) do they think they became ill? (e.g.,  suspected exposures, recent exposures to food, water (drinking and recreational), ice, other ill persons, children, and animals)

NOTE TO INSTRUCTOR:  Encourage individuals who have experience in receiving foodborne illness complaints for many years to share “pearls of wisdom” such as the following:

• Always collect as much information as possible from the person reporting an illness the first time contact is made; it might be difficult to talk with the person again. If the complainant cannot provide critical pieces of information, try to find out who may be able to and contact that person.  Be sure to ask the reporter how s/he can be reached in the future and if anyone else has been notified of this problem.
• Collect information on pertinent negatives as well as pertinent positives. For example, if one only records that the person’s symptoms included vomiting and diarrhea, it is difficult to know if that means there was no fever or the information was not collected.
• Collect a complete food history. Regardless of the source, complainants will often associate illness with their last meal (particularly if it was at a commercial establishment). 
  • If the etiologic agent is not known, obtain at least a 72-hour food history (i.e., all foods/beverages/meals consumed in the 72 hours prior to onset of illness).
  • For illnesses in which diarrhea is the predominant symptom (as opposed to vomiting), one should collect a 5-day food history because incubation periods for diarrheal diseases tend to be longer.
  • If the etiologic agent is known, ask about foods/beverages/meals eaten within the incubation period for that illness.
  • If more than one person is reported ill, foods/beverages/meals COMMON to all persons will be of particular interest BUT complete food histories for the appropriate time periods should still be collected.
• Remember that many illnesses that can be acquired through foods may also be acquired through other means such as water, person-to-person contact, and animal-to-person contact. Keep an open mind about possible sources and do not assume that it must be food.
• Be sure to accurately record symptoms, dates and times of the onset of illness, and dates and times of food consumption. Most people who have experienced a recent illness should be able to provide you with these answers.
• Thank the person for notifying you of their illness.

The “Foodborne Illness Complaint Worksheet” (Appendix 1) was completed based on the call.  The student refused to give his name or provide a telephone number or address at which he or his roommate could be reached.

Question 2:  Do you think the student’s complaint should be investigated further?  (VOTE)

1. Definitely
2. Probably
3. Probably not
4. Definitely not

Ideally, all reports of possible outbreaks of foodborne illness should be investigated to:

• Prevent other persons from becoming ill (either from the same food or method of food preparation),
• Identify potentially problematic food preparation practices, and
• Add to our knowledge of foodborne diseases.

Given resource constraints in many health departments, however, it might not be possible to investigate all individual cases of potential foodborne disease or investigate all cases to the same degree.  Public health workers often must choose which instances receive highest priority for investigation.  The highest priority usually should be given to outbreaks that:

• Have a high public health impact because they:
  • Cause severe or life-threatening illness, such as infection with coli O157:H7, hemolytic uremic syndrome (HUS), or botulism;
  • Affect populations at high risk for complications of the illness (e.g., infants or elderly or immunocompromised persons); or
  • Affect a large number of persons.

• Appear to be ongoing and associated with one of the following:
  • A food-service establishment in which ill food workers provide a continuing source of infection;
  • A commercially distributed food product that is still being consumed; or
• A involve a suspected adulterated food.

Clues that a follow-up investigation may not be warranted or is unlikely to be productive include:

• Signs and symptoms (or confirmed diagnoses) among affected individuals suggesting they might not have the same illness,
• Ill persons who are not able to provide adequate information for investigation including date and time of onset of illness, symptoms, or a complete food history,
• Confirmed diagnosis and/or clinical symptoms that are not consistent with the foods eaten and the onset of illness, or
• Repeated complaints by the same individual(s) with no significant findings upon investigation.

In this foodborne illness complaint, one might be a little skeptical.  The student refused to give a food history beyond the foods eaten at the pizzeria and the question about attending classes and taking a midterm exam sounds a little suspicious (i.e., as if the roommates might just want an excuse to avoid an unpleasant situation).  The fact that the student was not willing to give his or his roommate’s name, however, should not be over interpreted.  Anonymous complaints are not uncommon and do not automatically invalidate a complaint.  Complainants often request anonymity for fear of retribution.  Anonymous reports do make investigation and follow-up more challenging.

TDH staff were skeptical of the student’s report but felt that a minimal amount of exploration was necessary.  They contacted the City Health Department to determine if staff were aware of a problem.  City Health Department staff reviewed the foodborne illness complaint log to see if others had reported similar illnesses or exposures.  Although a few reports of vomiting and diarrhea had been received, no other recent complaints mentioned the pizzeria or involved students from the university.

TDH staff then made a few telephone calls.  The pizzeria, where the student and his roommate had eaten, was closed until 11:00 A.M.  There was no answer at the University Student Health Center, so a message was left on its answering machine.

A call to the emergency room of a hospital close to the university (Hospital A) revealed that 23 university students had been seen for acute gastroenteritis in the last 24 hours.  Based on the emergency room triage log, only three patients had been seen for similar symptoms from March 5-9, none of who were associated with the university.  Stool specimens from 17 students had been submitted for routine bacterial pathogens to the Hospital Laboratory on March 10, but no results were available

Around 10:30 A.M., the physician from the University Student Health Center returned the call from TDH and reported that 20 students with vomiting and diarrhea had been seen at the clinic the previous day and more were waiting to be seen that morning.  He believed only 1-2 students typically would have been seen for these symptoms in a week.

Question 3:  Do you think these cases of gastrointestinal illness represent an outbreak at the university?  Why or why not?

An outbreak is usually defined as two or more cases of a similar illness among individuals who have shared a common exposure.  The critical components of this definition are:

• Same diagnosis or symptoms and signs suggestive of same illness
• Clear association between cases, with or without a recognized common source

The association between cases includes things such as attending the same event or going to the same school or eating the same food. 

It seems likely that the cases of illness among students at the university represent an outbreak.  What is not clear is whether the outbreak is limited to the university or if the wider community is also affected.  Case finding methods to this point, utilizing a hospital near the university and the University Student Health Center, are more likely to pick up cases among students than in the community.

PART II – HYPOTHESIS GENERATION

TDH asked health care providers from the University Student Health Center, the Hospital A emergency room, and the emergency departments at six other hospitals located in the general vicinity of the university to report all patients with vomiting or diarrhea seen since March 5.

TDH investigators then visited the emergency room at Hospital A and reviewed medical records of the 26 patients seen at the facility for vomiting and/or diarrhea since March 5.  All but three were students at the university.  Based on these records, symptoms among the students included vomiting (91%), diarrhea (85%), abdominal cramping (68%), headache (66%), muscle aches (49%), and bloody diarrhea (5%).  Oral temperatures ranged from 98.8°F (37.1°C) to 102.4°F (39.1°C) (median: 100°F [37.8°C]).  Complete blood counts, performed on 10 students, showed an increase in white blood cells (median count: 13.7 per cubic mm [normal: 4.8-10.8 per cubic mm]).

Preliminary stool culture results from the 17 students from whom specimens had been collected did not identify Salmonella, Shigella, Campylobacter, Vibrio, Listeria, Yersinia, Escherichia coli O157:H7, Bacillus cereus, or Staphylococcus aureus.  Examinations for ova and parasites were negative.  Some specimens were positive for fecal leukocytes and fecal occult blood.

Question 4:  How might you interpret the bacterial culture results?

Several explanations exist for the negative cultures:

• Specimens may have been improperly collected or mishandled during storage, transport, or processing leading to the death of any biological pathogen present.
• Specimens may have been collected too late in the course of the patients’ illnesses (i.e., the patients were no longer excreting the pathogen in adequate numbers for detection).
• The illness may be due to some agent not tested for by the laboratory (e.g., virus or bacterial pathogen not routinely identified, parasite, preformed toxin, chemical agent).

Since the cultures were performed at the Hospital A laboratory, it would seem that transportation difficulties (and aging of specimens) would not be a large problem.  Most of the specimens were collected on March 10, shortly after the students became symptomatic.  For most infectious agents, patients would still be shedding microorganisms at that point in their illness.  We do not know about the reputability of the Hospital A laboratory or whether there may have been some temporary problems with processing the specimens.  The fact that all 17 specimens were negative suggests either the laboratory had a very big problem or, indeed, the cultures were negative for the pathogens examined.  If we assume that laboratory procedures were acceptable, it seems likely that the agent causing the illness was not detected because it was not tested for.

Question 5:  Based on the findings so far, what type of agent do you think might be causing this outbreak? (VOTE)

1. Bacteria
2. Virus
3. Parasite
4. Preformed toxin or chemical 

There are two broad classifications for gastrointestinal illnesses:

• Infections are a consequence of the growth of a microorganism in the body. Illness results from two mechanisms: 1) viruses, bacteria, or parasites invade the intestinal mucosa and/or other tissues, multiply, and directly damage surrounding tissues and 2) bacteria and certain viruses invade and multiply in the intestinal tract and then release toxins that damage surrounding tissues or interfere with normal organ or tissue function.  The necessary growth of the microorganism (for production and release of toxins) takes time; thus, the incubation periods for infections are relatively long, often measured in terms of days as compared to hours or minutes for preformed toxins.  Symptoms of infection usually include diarrhea, nausea, vomiting, and abdominal cramps.  Fever and an elevated white blood cell (WBC) count are often associated with infections.
• Illness due to a preformed toxin is caused by ingestion of food already contaminated by toxins. Although sources of preformed toxins include certain bacteria, poisonous chemicals, and toxins found naturally in animals, plants, or fungi, these illnesses most often result from bacteria that release toxins into food during growth in the food.  The preformed toxin is ingested; thus, live bacteria do not need to be consumed to cause illness.  Illness from a preformed toxin manifests more rapidly than that due to an infection because time for growth and invasion of the intestinal lining is not required.  The incubation period is often measured in minutes or hours.  Signs and symptoms depend on the specific toxin ingested.  The most common (and sometimes only) symptom is vomiting.  Other symptoms include nausea, diarrhea, and interference with sensory and motor functions (e.g., double vision, weakness, respiratory failure, numbness, tingling of the face).  Fever and an elevated WBC count are rare with the ingestion of preformed toxins.

Based on the information collected so far, it would seem that clinical findings (i.e., diarrhea, fever, an elevated white blood count, and fecal leukocytes) are most compatible with an infection.  Given that cultures for the usual foodborne bacterial pathogens and examinations for ova and parasites were negative, a virus might be the most likely causative agent.

By the next morning, March 12, seventy-five persons with vomiting or diarrhea had been reported to TDH.  All were students who lived on the university campus.  No cases were identified among university faculty or staff, students living off-campus, or from the local community.  Except for one case, the dates of illness onset were March 9-12. (Figure 1) The median age of patients was 19 years (range: 18-22 years), 69% were freshman, and 62% were female.

Figure 1.  Onset of gastroenteritis among students, University  X, Texas. (N=72) (Date of onset was not known for three ill students.

Question 6:  Based on the epidemic curve and likely causative agent, what is the likely mode of transmission and the period of interest for possible exposures related to this outbreak?

The epidemic curve (also called epi curve) is a visual display of an outbreak’s magnitude and time trend.  The horizontal axis represents the date each case became ill (i.e., the date of onset).  The vertical axis is the number of persons who became ill on each date. 

The shape of an epi curve can suggest the pattern of spread for an outbreak.  The shape of the above epi curve is suggestive of a point source outbreak (i.e., the number of cases rose rapidly to a peak and then fell off more gradually). 

In a point source outbreak, the epi curve can be used to identify the likely time period of exposure, if the incubation period for the disease is known.  The incubation periods for most viral gastroenteritis range from less than a day to about a week.  The majority of cases in this outbreak had onset of illness from March 10-12.  Counting back the minimum incubation period (< 1 day) from the earliest cases and the maximum incubation period (7 days) from the latest cases suggest that exposures of interest occurred between March 5 and 10.

TDH and City Health Department staff met with the Student Health Center physician and nurse and several administrators to learn more about the university and student body.  The following information was gathered:

The university had an enrollment of approximately 12,000 students; 2,386 students lived on campus at one of the 36 residential halls scattered across the 200+ acres of the main campus

The university used municipal water and sewage services.  Due to the large size of the campus, residential halls received water from several dozen water mains.  There had been no work on water or sewage lines in the past year nor recent roadwork or digging around campus.

The campus dining service included two cafeterias; about 2,000 students belonged to the university meal plan which was limited to persons living on campus.  Most on-campus students dined at the main cafeteria which served hot entrees, as well as items from a grill, deli bar, and salad bar.  A second smaller cafeteria on campus offered menu selections with a per item cost and was also accessible to meal plan members.  In contrast to the main cafeteria, the smaller cafeteria tended to be used by students who lived off campus and university staff.

To better understand the likely source of the outbreak, investigators undertook hypothesis-generating interviews with seven of the earliest cases reported by the emergency room and the Student Health Center; all of the cases had onset of illness on March 10.  Four were male and three were female.

Among the seven students interviewed, all but one was a freshman.  Two students were psychology majors; one each was majoring in English and animal husbandry.  Three students were undecided about their major.  The students were from seven different residential halls and all reported eating most of their meals at the university’s main cafeteria.  During the past week, all but one student had eaten food from the deli bar; two had eaten food from the salad bar, and three from the grill.  Seven-day food histories revealed no particular food item that was common to all or most of the students.

None of the students shared any classes; only one student knew someone with a similar illness (i.e., his roommate).  In the last week, none of the students had had contact with children in diapers and only the student majoring in animal husbandry had had contact with animals.

Five students belonged to a sorority or a fraternity.  Three students had attended an all school mixer on March 6, the Friday before the outbreak began; two students went to an all night science fiction film festival at one of the dorms on March 7.  Students reported attendance at no other special events; most had been studying for midterm exams for most of the weekend.

Question 7:  At this point, what is your leading hypothesis on the mode of transmission in this outbreak?  (VOTE)

1. Foodborne
2. Waterborne
3. Person-to-person transmission
4. Animal-to-person transmission

Food seems the most likely mode of transmission based on the following information:

Illness is limited to students living on campus.  The campus uses city water supplies.  If city water supplies were contaminated, one would also expect to see cases in the community.  It is possible that there are isolated problems with water and sewer lines on campus, but students from at least seven residential halls were affected, so a break would have to affect water distribution over a wide area.  (And if campus water was widely contaminated, one might expect to see illness in faculty, staff, and off-campus students who consumed water while on campus.)

Cases occurred in a number of different residential facilities and, among hypothesis-generating interviews, did not cluster by dorm rooms (i.e., roommates were not affected), or classes.  This pattern is not consistent with person-to-person spread.

No events were attended by all or most of the students who were interviewed.  A large proportion of students living on campus are part of the meal plan; most on-campus students eat at the main cafeteria.  University staff and off-campus students rarely eat at the main cafeteria.  All students in the hypothesis-generating interviews ate at the main cafeteria during the period in question and most also ate at the deli bar suggesting contaminated food or drink from this site might be the mode of transmission.  No common food items were identified through hypothesis-generating interviews.  However, viral agents are commonly transmitted through raw or poorly cooked shellfish, sandwiches, and salads.

Based on clinical findings, the descriptive epidemiology of early cases, information about the university, and hypothesis-generating interviews, investigators hypothesized that the source of the outbreak was a viral pathogen spread by a food or beverage served at the main cafeteria at the university between March 5 and 10.

Question 8:  What studies or other investigations would you undertake to explore this hypothesis?

Next steps in this investigation might include testing of stool specimens from ill persons, a controlled epidemiologic study, an environmental health assessment of the main cafeteria and deli bar (e.g., inspection of operations and interviews with staff), and collection and testing of leftover food, water, and ice from the main cafeteria and deli bar. 

PART III – STUDIES TO TEST THE HYPOTHESIS:  THE THREE-LEGGED STOOL

LABORATORY STUDIES

To explore the source of the outbreak at the university, TDH investigators initiated three lines of investigation: laboratory studies to determine the causative agent, an environmental health assessment of the university main cafeteria, and an epidemiologic study of students living on campus.

Health care providers were asked to collect stool specimens from new cases of vomiting and diarrhea.  Bacterial cultures from patients seen in the emergency rooms were to be performed at the hospital at which they were collected and confirmed at the TDH Laboratory.  Specimens collected at the Student Health Center were to be cultured at the TDH Laboratory.  Stool specimens from a sample of ill patients were sent to the Centers for Disease Control and Prevention (CDC) for viral studies including reverse transcription-polymerase chain reaction (RT-PCR).

Question 9:  What instructions would you give to health care providers for the collection of stool specimens from patients?  Include instructions on how specimens should be collected, stored, and transported.

Specimens should be collected from patients with a recent onset of illness who manifest illness typical of the outbreak, and have not received antibiotic therapy.  For bacterial cultures, rectal swabs can be used and specimens can be frozen if processing is likely to be delayed more than 48 hours.  For the viral studies, fresh stool specimens are needed.  Rectal swabs are discouraged.  Because freezing destroys the structure of viral particles, samples should be refrigerated but not frozen unless being tested by RT-PCR.

If the patient is collecting the sample, the following instructions might be helpful:

1. Place sheets of newspaper beneath the toilet seat and push them down slightly in the center.
2. Pass the stool directly onto the newspaper being careful not to contaminate the specimen with urine.
3. Scoop up the requested amount of stool and place in a clean container.

ENVIRONMENTAL HEALTH STUDIES

On the afternoon of March 12, TDH and City Health Department food safety inspectors conducted an environmental health assessment of the main cafeteria at University X and interviewed staff.  Thirty-one staff members were employed at the cafeteria of whom 24 (77%) were food workers.  Except for one employee, all dining service personnel were interviewed.

Question 10:  Given that investigators suspect a virus as the causative agent, what contributing factors are likely to have played a role in this outbreak? 

In general, contributing factors of interest in foodborne outbreaks due to a virus include the following:

• Ill food worker
• Bare-hand contact
• Inadequate handwashing
• Contaminated raw products
• Contaminated surfaces

Question 11:  What activities would you undertake during the environmental health assessment?  What key areas should be explored during interviews with the cafeteria food workers? 

 An environmental health assessment is a systematic, detailed, science-based evaluation of environmental factors that contributed to the transmission of a particular disease in an outbreak.  It differs from a general inspection of operating procedures or sanitary conditions used for the licensing or routine inspection of a restaurant or food production facility.  An environmental health assessment focuses on the problem at hand and considers how the causative agent, host factors, and environmental conditions interacted to result in the problem.

Items to cover during the environmental health assessment include:

• Identify food items served at the cafeteria (particularly the deli bar) during the implicated period.
• Outline work schedules for food workers during the implicated period and identify who was responsible for what on which days.
• Watch or reconstruct the procedures in the kitchen (in handling/preparing the food items) that happened the days it is believed the contaminated food was prepared. Try to identify any unusual occurrences or departures from routine during the implicated time period.
• Get recipes for foods served during the implicated period and identify the ingredients and sources.
• Determine whether any kitchen staff or their family members were ill at the time of the outbreak.
• Describe handwashing facilities at the cafeteria and routine hygienic practices of workers.
• Collect stool specimens from all food workers.
• Collect samples of foods from implicated period for cultures and other studies.
• Collect water and ice to test for fecal coliforms.

Investigators toured the facility and obtained a list of the foods served at the main cafeteria during the implicated period.  Cafeteria staff were questioned about their responsibilities such as the foods they handled, which meals they served, and where they usually worked (e.g., deli bar, grill).  They also were asked about use of gloves, handwashing practices, work schedule during the week before the outbreak, and if they had been ill.  None of the food workers reported being ill in the last two weeks.  The cafeteria did not have a sick food workers policy.

An inspection of the main cafeteria food preparation area, equipment, and serving line was unremarkable.  Walk-in refrigerators and freezers were organized to prevent cross contamination and maintained at appropriate temperatures.  Food preparation surfaces were clean and appropriately situated with respect to flow of kitchen traffic.  Steam tables on the serving line heated to proper temperatures.  Other equipment (e.g., meat slicer) was clean and in good working order.

The deli bar had its own refrigerator and preparation area.  During mealtimes, sandwiches were made to order by a food worker.  Each day, newly prepared deli meats, cheeses, and condiments were added to partially depleted deli bar items from the day before (i.e., without discarding leftover food items).  While the deli bar was open for service, sandwich ingredients were not kept refrigerated or on ice.  The deli bar containers were not routinely cleaned.  The refrigerator cooled only to 47°F.

Water and ice from the cafeteria were collected to test for fecal coliforms.  Samples of leftover food were collected from the deli bar for bacterial cultures and special viral studies at CDC.  Stool specimens were requested from all cafeteria staff.

Before dinner on March 12, the City Health Department closed the deli bar.

Question 12:  Do you agree with the decision to close the deli bar?  (VOTE)

1. Definitely
2. Probably
3. Probably not
4. Definitely not

There is no solid evidence to implicate the deli bar as the source of the outbreak.  The action is based on the most likely hypothesis and circumstantial evidence (i.e., other foodhandling problems identified in the deli bar).  Because there were multiple serious problems identified at the deli bar, however, closing it down until safer practices can be assured would seem reasonable.  Furthermore, closure will be a minimal burden to the university and its students and could prevent additional cases from occurring.

EPIDEMIOLOGIC STUDIES

By the morning of March 13, one hundred and twenty-five persons with vomiting or diarrhea had been reported to TDH.  All cases were among students who lived on campus.

TDH staff undertook a case-control study to test the hypothesis that the source of the outbreak was a food or beverage served at the main cafeteria at the university between March 5 and 10.  For the study, a case was defined as vomiting or diarrhea (3 or more loose bowel movements during a 24-hour period) with onset on or after March 5 in a student who lived on campus.  Cases were selected from those reported to TDH by one of the local emergency rooms or the Student Health Center.  Controls were students who lived on campus who did not have nausea, vomiting, or diarrhea since March 5.

Investigators administered the study questionnaire by telephone from March 15-23.  Investigators asked cases and controls about a variety of exposures including foods eaten during March 5-10 and where the foods were prepared.   Thirty-six cases and 144 controls were contacted.  Cases included in the study were similar to all cases with respect to gender, age, year in college, and onset of illness.

TDH investigators tabulated the results from the case-control study.

Eating lunch or dinner at the deli bar on March 9 or 10 was statistically significantly associated with illness.  To identify the specific item(s) at the deli bar causing the outbreak, investigators reanalyzed study data only from cases and controls who ate at the deli bar during March 9-10.  (Table 1)  For these analyses, information from 28 cases and 20 controls were examined.

Table 1.  Food items eaten by students who ate at deli bar during implicated meals,* case-control study, University X, Texas

Exposure	Ill exposed/ Total ill (%)	Well exposed/ Total well (%)	Odds Ratio	95% Confidence Interval	p-value
American cheese	13/28 (46)	4/20 (20)	3.4	0.80-17.5	0.06
Swiss cheese	8/28 (29)	8/20 (40)	0.61	0.15-2.4	0.30
Ham	11/28 (39)	6/20 (30)	1.5	0.38-6.3	0.36
Turkey	15/28 (54)	11/20 (55)	0.95	0.26-3.5	0.57
Shredded lettuce	13/28 (46)	10/20 (50)	0.87	0.24-3.2	0.52
Tomato	7/28 (25)	6/20 (30)	0.78	0.18-3.5	0.50
Pickles	7/28 (25)	7/20 (35)	0.63	0.15-2.6	0.63
Mayonnaise	20/28 (71)	9/20 (45)	3.1	0.78-12.4	0.06
Mustard	10/28 (36)	9/20 (45)	0.68	0.18-2.6	0.52

*includes lunch and dinner on March 9 and lunch on March 10

Question 13:  Interpret the results in Table 1.

Interpretation: 

An odds ratio of 1.0 means that the odds of exposure among cases is the same as the odds of exposure among controls (i.e., the exposure is not associated with the disease).

An odds ratio of greater than 1.0 means that the odds of exposure among cases is greater than the odds of exposure among controls; the exposure may be associated with the disease if the odds ratio is statistically significantly greater than 1.0.  A p-value of less than 0.05 (or a 95% confidence interval for the odds ratio that does not include 1.0) suggests that the odds ratio is significantly greater than 1.0.

An odds ratio of less than 1.0 means that the odds of exposure among cases is lower than the odds of exposure among controls; the exposure may be protective if the odds ratio is statistically

significantly less than 1.0.  A p-value of less than 0.05 (or a 95% confidence interval for the odds ratio that does not include 1.0) suggests that the odds ratio is significantly less than 1.0.

Among persons who ate at the deli bar during the implicated meals, the odds of exposure to American cheese and mayonnaise was more than 3 times higher among cases than controls.  The odds of exposure to ham was 1.5 times higher among cases than controls.  These differences, however, could be due to chance (p-values: 0.06 for cheese and mayonnaise and 0.33 for ham).

The inability to epidemiologically implicate a specific food item could reflect the relatively small number of cases included in the analysis of food items eaten at the deli bar or inaccurate recall of foods consumed by both cases and controls.  These findings could also result if multiple foods were contaminated or if illness was caused by some factor at the deli bar that was not explored

PART IV – CONTROL AND PREVENTION MEASURES

Water and ice samples obtained from the cafeteria on March 12 were negative for fecal coliforms.  Stool cultures from the 23 food workers were negative for bacteria.

Of the 18 fresh stool specimens sent on ill students to CDC, 9 (50%) had evidence of norovirus by reverse transcription-polymerase chain reaction (RT-PCR).  Of the deli foods available from the implicated meals, only the ham sample was positive by RT-PCR for the presence of norovirus RNA.

On March 25, the university cafeteria staff member who initially refused to be interviewed agreed to talk to the investigators.  The staff member worked primarily at the deli bar.  She reported slicing ham on March 9 for use at the deli bar during lunch and dinner that day, and lunch the following day.  She also prepared and served sandwiches for these same meals.  She reported that she had worn gloves while slicing the ham and while serving sandwiches at the deli bar.  Because she wore gloves during food preparation and serving, however, she did not feel that hand washing was an important activity.

The staff member denied any gastrointestinal illness during the outbreak period but reported that her infant had been sick with watery diarrhea since March 7, two days before she prepared items for the implicated meals.  A stool sample collected from the ill infant on March 25 was positive for norovirus by RT-PCR.  The sequence of the amplified product was identical to those products from the ill students and the deli ham.

On March 26, the chief of the campus food service called TDH to find out what must be done to reopen the deli bar.

Question 14:  What actions would you recommend/require? 

At this point, one must consider short-term interventions for the control of the current outbreak and longer-term interventions that might prevent the spread of foodborne diseases at the cafeteria in the future.  At a minimum, the following actions should be undertaken:

• Throw away all leftover deli bar foods and ingredients,
• Clean and disinfect all equipment and surfaces in the deli bar,
• Remove any infected food workers,
• Educate food workers on proper food preparation procedures including avoiding bare-handed contact with ready-to-eat foods, handwashing and appropriate glove use, and monitoring hot-holding and cold-holding temperatures, and
• Change procedures at deli bar with regards to use of leftovers,
• Review, revise, or develop a sick food worker’s policy.

The first three items in the above list will prevent further spread of disease from this particular outbreak.  The last three are more focused on prevention of future outbreaks, although they may be effective in controlling the current outbreak if any food workers are infected and still

shedding virus.  It is likely that a combination of these actions will be most appropriate to improve food safety at the university cafeteria. 

Some investigators might be tempted to require all food workers to submit a stool specimen before allowing them to return to work.  Examination of stool specimens from food workers (to look for viral pathogens) will be expensive and has little benefit for the current outbreak or in preventing future problems.

NOTE:  A fundamental problem in this outbreak appeared to be the management of the deli bar and assurance that food is safely handled on a routine basis.  Because food worker turnover in this setting is likely to be high, the institution needs to take a major role in assuring:

• A safe, clean environment with adequately functioning equipment,
• Facilities which promote good food preparation practices (e.g., conveniently placed sinks with adequate soap and clean towels),
• Safe sources for foods and ingredients,
• Training of employees in proper food preparation procedures, and
• Monitoring (and correction) of employee practices and behaviors.

If this assurance function seems inadequate, the university should be asked to develop a plan to address and exercise this function. 

The health department should monitor the situation until satisfied that a safe food preparation system has been established.  A food safety specialist and/or sanitarian should return to the university cafeteria in one month and on a quarterly basis for at least the first year to make sure no further disease is occurring and assess the development and implementation of appropriate policies, the practice of good food preparation procedures by employees, and ongoing oversight by the institution, itself.

The deli bar was thoroughly cleaned.  All equipment and surfaces were disinfected.  All leftover foods and ingredients were thrown away.  The deli bar refrigerator was fixed so that it cooled to 40°F (or less).

TDH staff worked with university officials to develop and implement policies to assure safe food preparation.  Special training sessions were held with cafeteria employees to make sure they understood the policies and safe food preparation practices.  The local health department intensified its monitoring of food service activities at the university, placing a special emphasis on the oversight provided by food services management.  After implementing control measures recommended by TDH staff, the deli bar was reopened on April 1.

TDH investigators summarized the outbreak investigation in a written report and completed the CDC Form 52.13 (i.e., the National Outbreak Reporting System [NORS] report form).

APPENDIX 1

State Department of Health FOODBORNE ILLNESS COMPLAINT WORKSHEET

PERSON COMPLETING INFORMATION                    Date:  03/11/97             #:  97-076 Name:   Xavier Onassis                                                    (: (512) 555-1234 Affiliation:   o Local BOH (town): ______________    ý State DPH (division): Epi               o Other:_____________

REPORTER/COMPLAINANT Name:        Refused to provide                                                                (: (         )  _______ – ___________ Address: ___________________________________________________________________________________ Affiliation:  ý Consumer                       o Laboratory        o Local BOH o Medical Provider             o State DPH         o Other    (specify: ___________________________)

ILLNESS INFORMATION # Persons ill:         2    Symptoms:  (mark if reported for anyone):

Diarrhea: Abdominal cramps: Headache: Loss of appetite:  Burning in mouth: Onset:     ®    Earliest Latest

ý Yes  o No o Yes  ý No ý Yes  o No o Yes  ý No o Yes  ý No Date: 03/10/97 Date: 03/11/97

Vomiting: Fever: Muscle aches: Fatigue: Other symptoms: Time:  11:30 Time:  2:30

ý Yes  o No ý Yes  o No o Yes  ý No ý Yes  o No None oAM  ýPM ýAM  oPM

Nausea: Bloody stool: Chills: Dizziness:

ý Yes  o No o Yes  ý No o Yes  ý No o Yes  ý No

Duration: Less than 24 Hours    24-48 Hours   More than 48 Hours    Ongoing  Unknown

Ill Person’s Name

(

Age

Onset

Physician’s Name (if seen for this illness)

Other Relevant Information*

1

reporter (above)

refused

18 y

3/11 (2:30 AM)

drinks water from campus, no contact with children in diapers or animals

2

Refused

refused

19 y

3/10 (11:30 PM)

drinks bottled water; no contact with children in diapers or animals

3

4

*Note if individual is foodhandler. Medical attention received  (by anyone)?  o Yes  ý No  o Unknown   ®  If Yes, specify above: ­ Stool specimens submitted (by anyone)?  o Yes  ý No  o Unknown   Laboratory:  ________________________ Medical diagnosis?  ________________________________

FOOD HISTORY ® 72 hours prior to symptoms, or, if organism identified, between min and max incubation periods ® If 2 or more ill, follow above time frame for common meals (foods) only                                #                                                                           Restaurant / store where    Date & Time2        Exp3            Food(s) consumed                      purchased (name, town)         Place consumed
o B March 8         o L o D		Ate separately	University cafeteria	Same (as left) Home Other (specify):
o B March 9         o L o D		Ate separately	University cafeteria	Same (as left) Home Other (specify):
o B March 10       o L ý	2	Anchovy pizza (cheese, onions, and anchovies) and beer	University cafeteria Local pizzeria	Same (as left) Home Other (specify):
o B o L o D				Same (as left) Home Other (specify):
o B o L o D				Same (as left) Home Other (specify):
o B o L o D				Same (as left) Home Other (specify):
NOTES Student reported that he and his roommate usually ate separately at the University X main cafeteria.  Could not provide list of foods eaten.  The only shared meal was at pizzeria.
FOOD TESTING Food(s) available for testing?   o Yes  o No  ý Unknown   ®  Sent to State Lab?   o Yes  o No  o Unknown ® If Yes, specify food(s) &  sources:
PRODUCT AND MANUFACTURER INFORMATION FOR COMMERCIALLY-PROCESSED FOOD(S) Product name:  _________________________________________________     Code/lot #__________________ Expiration date: _____ /_____ /_____                 Package size/type: ________________________________________ Manufacturer:   _____________________________________________        (: (        )  ______ – __________ Address:  _____________________________________________________________________________

¹State Laboratory Institute

²Always record Time if possible; otherwise, choose B=breakfast, L=lunch, D=dinner

³Total # persons (both ill and well) who consumed indicated food(s)